Urban civility or urban community? A false opposition in Richard Sennett’s conception of public ethos

The agonistic conception of public ethos

Sennett argues that the experience of strangeness is unsettling, frightening or even painful. Contact with what is unpredictable, unfamiliar, puzzling and unintelligible is a challenge to the comfortable illusion of being in control and all-knowing – an illusion that Sennett claims is characteristic of immaturity. We can only become adults in a meaningful sense of the word by opening up to the contingent and strange, to the complexity of social life and by dealing with difficult social interactions. The passage from adolescence to this new, possible adulthood, whose freedom lies in the ‘acceptance of disorder and painful dislocation’, depends on ‘a structure of experience that can only take place in a dense, uncontrollable human settlement – in other words, in a city’ (Sennett, 1970a: xvii). However, although modern complex and overwhelming cities offer this possibility, at the same time, the unsettling nature of dealing with urban difference constantly seduces us to seek cover by constructing a defensive identity that seeks to avoid personal change or transformation through disorder and otherness.

In his first and most empirical book, Families against the City (1970b), Sennett explores how the rapid growth and increasing complexity of cities like Chicago at the end of the nineteenth century, as a result of migration and industrialization, produced this fear of exposure to diversity and as a consequence a phenomenon he refers to as the ‘intense family’. The perception of urban confusion and social disorder led to the cultivation of the nuclear, close-knit family as a refuge from society, among others by means of the withdrawal from a rich social life outside of the domestic sphere. This type of family is associated by Sennett with a type of inward-turning life-form in the city, namely the middle-class suburb.

Suburbs meet the need of these families to avoid the experience of difference and social ambiguity, given that these are places of intense order and social sameness, among others by a rigid separation of functions. ² Sennett sees this whole development as disastrous for the cultivation of public ethos:

Intensive family life in America sapped a generation’s interest in participating in dissimilar kinds of contact experiences in the city … Indeed, it might best be said of city life during the past twenty-five years of suburban growth that the scale of life has become too intimate, too intense. (1970a: 73, 82)

Characteristic of this purified identity on a collective scale is that this sense of togetherness is defined by homogeneity, by a desire to be among those who are the same. Sennett takes this to be a dangerous development for a number of reasons – most importantly, that it tends to suppress deviance within, and that these communities tend to become entangled in a violent escalation with other communities. As a consequence of a lack of tolerance for a certain amount of disorder, each manifestation of social tension and disharmony is perceived as an absolute threat to the vital texture of the community. The paradox that Sennett uncovers here in 1970 – namely, that the orientation towards social harmony itself is responsible for disruptive socio-political conflict – is not unfamiliar since the work of Chantal Mouffe, James Tully and others from the 1980s onwards in the tradition of agonism.

Sennett’s radical solution in The Uses of Disorder to the tendency towards purified community is a restructuring of the city in such a way that people are forced to deal with social and cultural others in an arena-like social situation. He proposes establishing so-called ‘survival communities’, namely communities that are mixed in terms of class, ethnicity, culture and taste and in which citizens have to work together to establish and maintain a number of public goods and facilities, such as schooling, zoning and renewal (1970a: 141).

Although Sennett refers to this social cooperation as a type of ‘community’, namely, a survival community, he stresses the fact that these communities ought not to be dictated by an ideal of social harmony. Instead, hostile feelings and racial fears should be expressed openly, not be muted because otherwise ‘cities will continue to burn’ (p. 147). It is only by confronting social conflict in the city, in ‘a mode of ongoing expression … pushing men to say what they think about each other’ that we are able to live together with difference in ‘a more civilized and mature’ manner (pp. 181, 150). The problem of community is that ‘aggression is denied outlets other than violence’. The implicit theory of emotion that is in the background here is that ‘bottling up hostile aggressiveness’ towards others leads to violence, while ‘expressive outlets’ for these feelings make society more civil and stable (pp. 178–81). In a much later book, namely The Conscience of the Eye, the agonistic ideal is also present, as well as the idea that ‘mutual verbal conflict’ works as a ‘release-value, dissipating energies that could otherwise break out in violent form’ (1990: 198).

Central to Sennett’s vision is the idea that differences should interact. His rejection of internal walls within the city that produce sealed communities, completely separate in terms of space and socio-cultural relations is persuasive. The ability to deal with diversity is a necessary skill for a diverse polity and a skill that needs to be practised. Furthermore, completely separate communities might promote negative stereotypes. ³ The problem, however, is that his vision of the city includes the demand that differences constantly ought to interact (even the family as a comfort zone is critically scrutinized), that these interactions ought to be significant ⁴ (that is, not superficial), and that community-orientated initiatives are a threat to the stability of society. Hence the orientation is not so much towards the creation of permeable borders within the city, but making sure that borders themselves disappear to make room for diversity as an individualistic affair. I will present three main arguments against this view, but for now – by way of a prelude – I just want to point out two internal problems with it.

Firstly, Sennett wants social diversity without local communities. The question is whether it is possible to have one without the other. The urban theoretical thesis of ‘critical mass’ is that a city – compared to typical rural settlements – provides a fertile breeding ground for social deviance and eccentricity by virtue of a social concentration of some kind (Hannerz, 1992: 201 ff.). Birds of a feather can flock together here, and as such they are able to cultivate a life-form that resists pressures to conform to stable standards of ‘normality’. These types of concentrations, such as residential concentration, seem not an inherent threat to a diverse city but its very precondition: difference needs a chance to grow, diversity needs to be cultivated, cultural particularity needs social groups. To argue otherwise would imply signing on to the monological subject, a subject that is able to hang on to dissenting views and practices simply on its own. ⁵ Besides the socio-psychological conditions, residential concentration also makes certain ethnic institutions economically viable, such as specific shops and services that cater to the needs of an ethnic community. This enhances further a particular cultural variant and therefore the experience of urban diversity (Bolt et al., 1998).

Secondly, the question is whether Sennett’s proposals really would work in favour of urban peace. Although empirical research has established the notion that, under certain conditions, social conflicts can have an integrating and stabilizing power (Coser, 1956; Simmel, 1955), the bold suggestion that expression of cross-cultural or interethnic hostility in itself leads to stable cities is questionable. What is open to discussion is Sennett’s idea that this type of expression will work as an ‘expressive outlet’ and so will relieve strained relationships. This idea is referred to by some theorists of emotion as the ‘hydraulic model’ of emotion (Solomon, 1993: 162). This model argues that poisonous emotion can be weakened by discharging it, as in the therapeutic theory of primal screaming. The idea is that the expression of emotion is caused by the emotion that precedes it and at the same time releases its force. However, ‘not all emotional expression is a matter of “discharge” and “release”… Even limiting our attention to straightforward expressive actions, it is clear that they often have both the intention and the effect of intensifying, not eliminating, the emotion’ (Solomon, 1993: 164). So the ideal of ‘letting it all out’ is by no means a safe strategy for a successful intercultural city.

The puzzling thing about Sennett’s work on city life and public ethos is that four years after writing The Uses of Disorder, Sennett seems much more sensitive to this last point. For in The Fall of Public Man (Sennett, 1974), he argues precisely against the idea that the direct expression of one’s personal feelings, thoughts and dispositions towards others is part of a desirable public ethos. Here, he tends to stress the importance of privacy, formality and ritualized behaviour based on ‘the comprehension that, because every self is in some measure a cabinet of horrors, civilized relations between selves can only proceed to the extent that nasty little secrets of desire, greed, or envy are kept locked up’ (1974: 5). The two notions of civility, that are scattered throughout Sennett’s work, remain disconnected, even though he presents his work on the subject of the urban public sphere as ‘parts of a whole’ (1974: 44). ⁶ We will deal with this rather different vision of civility in the next section. For the sake of conceptual clarity I will refer to that mode of civility as ‘side-by-side civility’ and to the agonistic version as ‘agonistic civility’.

City life as a theatrum mundi

According to Sennett in The Fall of Public Man, something threatens to get lost in the Western, urban public sphere under the pressure of the upcoming ideal of authenticity – an ideal that he traces the beginnings of in late modernity, roughly the end of the eighteenth and the beginning of the nineteenth century. ‘Authenticity’ in this context means that citizens, individually or collectively, perceive themselves as ‘having’ a unique personality that they have to take seriously in order to be true to themselves. The collective manifestation of this preoccupation occurs when a group of citizens living in the same urban neighbourhood understand themselves as being part of a fixed, reified ‘we’. Here a desire for a shared intimate sphere arises and, as a consequence, the ability to deal with strangers decreases.

This clearly resonates with the notion of ‘purified identity’. Yet the norm that figures in the background is not the ideal of passionate interaction but, instead, is ‘public man’, defined, before his ‘fall’, by a crucial distance towards the self and, as a consequence thereof, by the ability to deal with strangers on a day-to-day basis within the public social sphere. According to Sennett, this crucial distance can only be cultivated via a role play that characterizes the public sphere as it should be, a role play that Sennett believes was characteristic for eighteenth-century Paris and London. This role play is so essential to public life that Sennett refers to a functioning public sphere as a worldly theatre, a theatrum mundi (1974: 34 ff.). This worldly theatre, in which we play our different segmented roles, instead of being serious about some presumed core identity that remains the same in different contexts, is under threat from the ethos of authenticity. As a consequence, civilized relations between citizens are undermined.

Being civilized here implies a cultivation of the playful character of being in public: ‘wearing a mask is the essence of civility’ for ‘masks permit pure sociability, detached from the circumstances of power, malaise and private feeling of those who wear them’ (Sennett, 1974: 264). Instead of the more confronting, in-your-face interaction that defines survival communities, here the emphasis is on the importance of social reserve, formality, role-playing and valuing social distance: ‘Civility is treating others as though they were strangers and forging a social bond upon that social distance’ (Sennett, 1974: 264). ⁷ The social reserve, the typical urban blasé attitude, is not simply a way to keep others at a safe distance but to keep them at a distance for the sake of living together.

The fundamental different conception of civic virtue – social reserve versus confrontation, formal ritual versus emotional expression, the mask versus in-your-face agonism – does not lead to a critical difference in valuing urban community. In both cases Sennett does not leave much room for legitimate clustering around ethnicity, class or lifestyle. Agonistic civility requires intensive and continuous cross-cultural contact, which is evidently hard to combine with ethnic clustering (that provides some relief in this regard). Side-by-side civility requires an anonymous and open public sphere, to which the ‘tyrannies’ of local intimacy, according to Sennett, pose a constant threat.

The only room left for ‘community’ is in the shape of the ‘survival community’, which implies a stretching of the meaning of ‘community’ beyond all recognition. ⁸ Most real-life situations of community formation in contemporary cities are characterized by the perception of some shared property that people positively identify with, particularly ethnicity, class, race or lifestyle (Gans, 1961). That does not imply the thesis that an ethnic or cultural or racial concentration equals a ‘community’. The concentration can be accidental, or the consequence of involuntary clustering based on practices of discrimination, what theorists sometimes refer to as ‘segregation’, or it can be based on patterns of choice that have nothing to do with ethnic of cultural solidarity but, for instance, with the desire to live in close proximity to family members (Wimmer, 2004: 16). ⁹

Community, as I will understand it, refers to a cluster of people that consider themselves and each other as part of a social group on the basis of a shared good, such as a language or culture, without this belonging being the direct result of individual choices or certain achievements. I will not explore this working definition of community here in depth. ¹⁰ At this point it is just worth emphasizing that belonging in this regard implies a process of identification that consists both of self-identification and of identification by others. But even though self-identification should be central to this definition of community, these social attachments cannot be understood as the result of an individual decision or choice, contrary to, for instance, one’s membership of a tennis club. An individual cannot simply choose to grow up in a particular culture, or to be of a specific nationality or ethnic descent. What is characteristic of a social attachment is precisely that it always, to a certain degree, escapes individual understanding and personal control. This does not imply that a social attachment to a cultural or religious group is unavoidable, one’s fate (Kymlicka, 1989: Chapter 4). Yet a community cannot be understood as the product of some social contract, nor the belonging to it as an individual decision (Young, 1990: 46). If I refer to the moral importance of ‘voluntariness’ of residential clustering, I refer to this notion of self-identification, not to a logic of naïve decisionism.

Living in an inner-city ghetto, with high levels of material deprivation and crime, is typically not the effect of voluntary choices in this sense, but rather of institutional racism, racist harassment and white flight to higher-status white neighbourhoods (Massey and Denton, 1993; Phillips, 2010). Citizens who are caught up in such slum dwelling, are trapped. From this perspective, living together with ‘one’s own kind’ offers not much that is positive but is a reminder of being socially excluded, of being at the bottom of society. These circumstances do not provide a fertile breeding ground for a positive sense of community. The sense of a ‘shared good’ is missing. To the contrary, ‘a looming sense of danger and a high propensity for violent interpersonal conflict sow seeds of distrust in ghetto neighbourhoods, making it difficult for a broad sense of community to form or be maintained’ (Shelby, 2007: 139). ¹¹ This type of forced and violent living together is what needs to be rejected, prevented and resisted. ¹²

Elizabeth Anderson argues that the distinction between ‘voluntary clustering’ and ‘involuntary segregation’ is problematic in the case of African-Americans, given the widespread and institutionalized anti-black racism that is still present in the USA. Instead of focusing on the justification of residential concentration, we ought to give priority to equality in social relations, among others by means of certain types of affirmative action (Anderson, 2010). However, I will argue that urban communities often do have a positive value for minorities, such as immigrants, that goes beyond the negative function of being a mere refuge from the larger hostile social context. The central question is whether such communities lead to bad citizens and lack of public ethos; an argument that Sennett embodies and that lies at the basis of minority ethnic dispersal policies in many Western countries, such as housing diversification and housing allocation measures (e.g. quota systems) (Bolt et al., 2010). ¹³

What is most problematic about Sennett’s notion of community in this regard is that the reference to a shared good, which implies a certain level of social unity or commonality, is resolutely rejected as being ‘dangerous’. ‘We’ is a ‘dangerous pronoun’ for the reason that ‘the desire for community is defensive, often expressed as rejection of immigrants or other outsiders’ (Sennett, 1969, 1998: 147, 138). However, to reduce community to the moment of contra-identification or ‘boundary maintenance’ (Barth, 1969) does not do justice to the fact that ethnic, cultural, religious and national communities, in general, also define themselves by an internal process of identification (with a shared homeland or culture or language) instead of solely being concerned with a flag-waving identity and the construction of ‘outgroups’. Very often this sense of a perceived ‘shared good’ is present, not just ‘shared outsiders’ (Jenkins, 2008a: Chapter 2; Jenkins, 2008b: Chapter 8). More importantly, although the formation of an ethnic identity inevitably implies the sense of a boundary between ‘us’ and ‘them’, the ‘them’ can be – and often is – perceived as part of a shared co-operative scheme or a shared cultural-linguistic identity that constitutes an overarching sense of ‘we’ (Eriksen, 1993: 28, 116).

What is strong and convincing about Sennett’s notion of side-by-side civility, however, is the fact that it is based on some of the structural elements of modern city-life, namely, its anonymous character and the social reserve that it incorporates, referred to by Georg Simmel as a blasé attitude and by Louis Wirth and others as the typical ‘indifference’ of city-dwellers (Simmel, [1903] 1995; Tonkiss, 2003; Wirth, [1938] 1995). The idea of taking seriously the existing culture of urban social reserve and uncovering its potential for living with cultural diversity is powerful and original. It constitutes an accessible type of public role because it is supported by the broad public sphere of many contemporary cities, although Sennett believes that this support is under threat by the ‘fire urban planners play with so carelessly when they talk about building a sense of community at a local level in the city, as opposed to reawakening meaningful public space and public life in the city as a whole’ (1974: 309). The last quote clearly shows the general opposition of ‘community’ and a ‘meaningful public space’ that Sennett works with.

In Sennett’s work, the relationship between agonistic civility and this side-by-side model remains unclear and, what is more important, the demon – urban community – that they expel as a threat to either intensive cross-cultural interaction or an anonymous public sphere does not get a fair chance. My goal here is to set the record straight with regard to the formation of urban communities, not just as a comment on Sennett but more importantly to question the current commonplace of regarding ethnic enclaves as proof of urban disintegration and failing citizenship. In doing so, I will also shed some light on the first issue by discussing these two modes of civility during the three arguments against Sennett’s anti-communitarianism.

Urban community and bridging

The basic premise of Sennett’s rejection of urban communities is that citizens are either committed to the wider public good or they are simply self-absorbed, exclusively focused on their local interests and problems. It is as if a sense of local togetherness and solidarity based on ethnic or cultural features inevitably crowds out a civic relation with others, whether these others are members of other communities within the city or whether they are simply perceived as other citizens tout court. However, there simply is no anthropological or sociological evidence to suggest that people either live in private or in public words and that complex identities in this regard are an impossibility.

Citizens, generally, participate in a range of different networks, some based on weak, civic ties but some on stronger emotional bonds (Robinson, 2006: 51 ff.). This incorporates the likelihood that some networks are emotionally denser than others and that it makes more sense to talk of degrees of privacy or publicness on a continuum, rather than a dichotomy between ‘public’ and ‘private’. It is possible, of course, that a social group sharing a neighbourhood is closer to the ‘private’ pole on the continuum, for instance, by thinking and acting according to the NIMBY maxim. But such a limited and self-interested outlook is by no means a necessary consequence of urban villages or ethnic clustering. One could only suggest otherwise by implying that people’s social and political bonds are by definition singular and exclusive – a very questionable suggestion.

A recent empirical study by Robert Putnam shows that Sennett’s suggestion in this regard is unsubstantiated and misguided. Putnam argues that, to put it in simple terms, living within a relatively homogenous ethnic community increases not only a sense of trust in members of the perceived in-group, what Putnam refers to as bonding social capital, but also tends to increase trust in the members of so-called out-groups, what he describes as a rise in bridging social capital (Putnam, 2007). ¹⁴ In other words, if people experience a sense of belonging by being part of a particular social network that generates norms of reciprocity and trustworthiness (that is, social capital), their attitudes towards those groups that are not a direct part of that network are more likely to be positive compared to citizens that are more or less isolated in social terms.

What Putnam’s study shows is that exposure to difference without a sense of safe belonging, however implicit and taken for granted that sense may be, tends to undermine both in-group and out-group trust. What he thereby rejects is a dominant fundamental assumption of much theorizing, ‘namely that in-group trust and out-group trust are negatively correlated’ (2007: 143). In order for a wider, community-transcending public ethos to be possible, we need exposure and contact with others, but not at the expense of in-group identifications and the sense of shared experiences, as Sennett claims. Instead, ‘bonding social capital can … be a prelude to bridging social capital, rather than precluding it’ (2007: 165).

In a recent book, Sennett tries to sideline the implications of Putnam’s study by suggesting that it ‘profiles attitudes more than actual behavior’ and that people should not act upon these attitudes of unease: after all, ‘in everyday life … we are constantly obliged to deal with people we fear, dislike or simply don’t understand’ (Sennett, 2012: 4). But this is missing the point forever that higher levels of trust in dealing with ethnic strangers are preferable to fear or unease and that ethnic clustering can (paradoxically maybe) facilitate this.

For a long time, a general connection has been made between residential concentration and lack of integration. This connection has lead to policies of housing diversification and quota systems. ¹⁵ For Sennett, it is an ‘open question’ whether ‘forced mixings’ would ever work, but he affirms the relationship between integration and racial or class dispersal (1974: 297). This connection, however, has recently been challenged by empirical research that is in line with Putnam’s findings (Arbaci and Malheiros, 2010; Bolt, 2009; Bolt et al., 2010; Drever, 2004; Harrison et al., 2005; Musterd, 2003; Phillips, 2010). Here it is argued that the possible positive effects of residential concentration are often overlooked and that ethnic mixing does not necessarily lead to more integration. In fact Bolt et al. argue, though with some qualifications, that ‘ethnic concentrations do not hamper integration but facilitate it’, especially for newly arrived immigrants:

Living in a neighborhood where social and cultural practices are familiar reduces the initial stress of having to operate in a culture that is foreign, and provides a buffer against feelings of social alienation. In addition, ethnic communities may play an acculturative role in helping their members adjust and fit in the host society. (Bolt et al., 2010: 177–8) ¹⁶

What these empirical findings show is the either-or choice that Sennett presents is misleading. Residential clustering does not necessarily pose a threat to a meaningful public sphere. Instead it could facilitate wider integration into a diverse society. Iris Young has argued convincingly for the normative possibility and acceptability of affinity groups in the urban public sphere on the basis of the premise that citizens are able to be both citizens of the wider city or nation and at the same time to cultivate an affinity with a local group that one shares ethnic or cultural traits with (Young, 2000: 196–235). To be clear, Young does make a conceptual (and current) ¹⁷ distinction between voluntary clustering and segregation, that is, ethnic or racial concentration that is the consequence of discrimination and exclusion. Segregation, for a number of reasons (some of which are obvious), is rejected by Young. But the solution to segregation does not have to be mixing, as some argue (Trappenburg, 2003). Besides the legal fight against racism in these cases, serious re-investments in these neighbourhoods ought to take place that in many cases are characterized by long-term neglect (Massey and Denton, 1993; Sampson, 2009).

Now, if bonding and bridging do not exclude one another, the question is what the bridging ties ought to live up to from the normative point of view. I want to suggest that it is good enough that these bridging ties take on the form of weak, civic ties. They do not have – and normally will not have – the same emotional intensity as local attachments to a particular urban community. Sennett’s side-by-side civility is quite valuable in this regard. Not only is it institutionally supported by metropolitan conditions, but it also constitutes a morally acceptable way of dealing with diversity. After all, it constitutes a type of mild tolerance that is able to accommodate a whole range of cultural and ethnic differences. The typical urban ethos of social reserve and social distance is a condition for the possibility of a wide band width of acceptable social diversity.

There are good reasons why the city is a welcoming place for ‘the other’– for migrants, single people, homosexuals, feminists, artists, etc. The city is the place where peripheral groups can settle, where cultural dissidents and newcomers can feel accepted in their otherness. In the city, otherness and uncommonness, instead of being domesticated, have a real chance of growing into a critical mass. This is not despite, but thanks to, a certain level of social distance.

In a way, this urban ethos approaches a kind of indifference-to-difference. I want to stress that this ethos is morally acceptable provided that it does not degenerate into an indifference to the basic needs of others, to their fate, the kind of indifference that the German journalist Günter Wallraff was confronted with many times in his disguise as ‘Ali’ (Wallraff, 1988). That type of indifference must be firmly rejected for the reason that it entails a violation of basic respect and civic care (Van Leeuwen, 2010).

My main point here is that this qualified side-by-side civility should not be thought of as irreconcilable with urban enclaves and ethnic clustering. To claim otherwise is to construct a false opposition between urban community and civility.

Urban community and empowerment

Sennett often describes the darker side of the sense of empowerment, namely, what steers a community on a self-interested course while disregarding any claims by the wider political context. An example in this regard is his interpretation of the Forest Hills casus: a middle-class, mostly Jewish section of Queens, New York, that was confronted in the beginning of the 1970s with the influx of black families through a housing project that New York City planned to build in the area. Sennett describes how, during the protests against these plans, a self-absorbed and self-interested sense of community developed without any regard for the needs of the black families in question – they simply had to go and live somewhere else (1974: Chapter 13). This is indeed an example of a community becoming uncivilized. However, I disagree with the tendency of Sennett to take such bad examples of urban communities and to draw general conclusions in regard to urban community formation.

One recurring line of community critique concerns the presumed a-political character thereof. According to Sennett, such groups are caught up in ‘a refusal to deal with, absorb, and exploit reality outside the parochial scale … converting claustrophobia into an ethical principle’ (1974: 310). Instead of focusing on the larger political realities that they are dependent on and trying to effectively influence the decision-taking processes that might affect them, he claims that community formation makes people incompetent to achieve this. They have in fact cut themselves off from the wider political institutions and in the process made themselves more vulnerable to become politically passive:

the more people are plunged into these passions of community, the more the basic institutions of social order are untouched … People in struggling to be a community become ever more absorbed in each other’s feelings, and ever more withdrawn from an understanding of, let alone a challenge to, the institutions of power. (Sennett, 1974: 309–10)

Secondly, if members of a cultural or ethnic group are on the receiving end of ‘discrimination, deprecating stereotypes, exclusion, and comparative disadvantage’, according to Iris Young, ‘the neighborhood clustering of the group can serve as an important source of self-organization, self-esteem, relaxation, and resistance’ (Young, 2000: 216–17). Young is right to defend such relative separation within urban areas if its purpose is mutual aid, building a sense of togetherness in an indifferent or even hostile wider culture and trying to achieve a more equal and just city. ¹⁹

Raymond Rocco describes in great detail how such communities can have this empowering effect. Rocco studied Latino communities in an area of south-east Los Angeles known as the ‘hub cities’ (Rocco, 1999: 257). What his research uncovered was the positive role forms of association that support the development of strong identities can play in promoting more inclusive and responsive forms of democratic governance and citizenship. One case he studied involves an informal network of Latino immigrants, mainly women from Mexico, Peru, El Salvador and Nicaragua. Not only did these woman help each other with exchanging child care, washing and ironing and developing shared strategies for coping with low incomes, but also a sense of being excluded grew that laid the foundation for the moral consciousness that they had ‘the right to make certain claims on society and political institutions’ (1999: 260). One example of such a claim was better local health care, for which a number of these women visited council meetings to make their voice heard.

Urban community can have an empowering impact and eventually lead to legitimate claims on the wider public. In this case, the claims are not geared towards the exclusion of ‘others’ who venture onto their ‘turf’, as in the Forest Hills case. The way the arguments are framed in the cases that were investigated by Rocco lead him to the conclusion that what is at stake is not simply narrow self-interest, but an appeal to a more equal distribution of recourses within the larger social setting of which these social groups are a part. The notion that community leads to depoliticized withdrawal is not supported by this kind of community research. On the contrary, the formation of rather thick patterns of solidarity formed the basis onto which the new claims to citizenship emerged. In this case, it is not only ethnicity but also gender that seems to explain the empowering effect of community, since the network discussed by Rocco involved women from different countries (Mexico, Peru, etc.).

As soon as citizens become politically active, however, the agonistic concept of citizenship becomes relevant. In the political arena conflict is inevitable and in this context it is not enough to cultivate a sense of detached respect, a relaxed live-and-let-live attitude, and to celebrate social reserve and anonymity. Yet I doubt that the expression of hostile feelings is, as Sennett claims, key here. More important in this regard is the ability and disposition to make a fundamental distinction between a political adversary and a social enemy – hence the importance of Chantal Mouffe’s work on agonism. In order to make this distinction and to move from the political to politics proper, a sense of a ‘common symbolic space’ is crucial (Mouffe, 2000: 13). Mouffe stresses the importance of a shared allegiance to certain principles in this regard (2000:102–3), but the reciprocal recognition of being a legitimate part of the same city and sharing an equal ‘right to the city’ (Mitchell, 2003) can be morally important as well. This is where the urbanites of the Forest Hills casus fall short. The boundaries of their community, during the protest against the housing project, became the boundaries of the moral community itself. Representatives of the ‘outside’ world were reduced to enemies that had to be resisted at all costs.

So there is a place for agnostic civility within the urban frame, but not quite in the general sense that Sennett envisaged in his work. Herman Van Gunsteren makes a conceptual distinction between the ‘public-political’ sphere and the ‘public-social’ sphere: ‘The first one is that of political and administrative decision-making about the (adjustment of the) organization of society. The second sphere relates to the interaction between people, legal as well as natural persons, in (semi)public places, such as the street’ (Van Gunsteren, 1998: 138). Agonistic civility is more appropriate for the public-political sphere, while side-by-side civility is better suited for the broader public-social sphere. Citizens who have to operate in both need both modes of civility. But many citizens are not politically active in the above sense. That is why I think agonistic civility is less basic than side-by-side civility.

Agonistic civility has been very influential in recent urban theory (Amin, 2002; Sandercock, 2003; Wood and Landry, 2008). Although this approach in general does not rely on the contested theory of emotion that Sennett’s interpretation is based on, it remains a rather demanding strategy for urban civility. According to the agonistic model, the acknowledgement of difference ought to be present in daily encounters in order to engage in passionate, but civil, debate concerning stereotypes, value systems and ways of life. It requires actively engaging the ideas and values of your partner to interaction, passionately rejecting those that you think are worth rejecting, listening carefully (even to views that seem offensive) and making an effort to defend your take on things without slipping into ideological cramps. This is far from a modest type of civility, given that urbanites are confronted with ethnic and religious others on a daily basis in a whole range of different contexts. They have often learned to deal with this diversity by cultivating a mild live-and-let-live attitude, a certain affective distance and mutual reserve, towards the difference (or ‘identity’) in question. I believe that this stance of side-by-sideness can be a morally acceptable and, more importantly, a practicable way of dealing with urban diversity, while agonism seems more fit for particular occasions, sites and institutions that we consider to be part of the political sphere in a more limited sense of the term, that is, in which matters of public concern are debated.

Urban community and well-being

The argument to allow voluntary ethnic and cultural clustering is not just based on the fact that these local ties that bind can facilitate, instead of simply hamper, bridging ties of civic trust towards ethno-cultural others. And it is not just that these urban communities can play an empowering role by providing a basis for improving local social justice. A third argument for allowing voluntary urban clustering is based on the fact that ethno-cultural communities within larger, and often quite anonymous, urban areas can provide a sense of belonging. Although cities are generally hospitable to difference because of a wide breadth of acceptable cultural diversity, the sense of belonging that they provide is typically not based on thick patterns of recognition and solidarity. ²⁰ Sennett certainly admits this. The later Sennett describes urbanites as being merely ‘lightly engaged’ through a ‘loosening of strong social connections’ (Sennett, 1990: 126–7; 1994: 355–70). Fran Tonkiss provides an analysis of the typical ‘urban solitude’ as an effect of this urban reserve (Tonkiss, 2003).

But although solitude at times can be inspiring and valuable, as Sennett points out (1994: 274), living with a sense of belonging, however implicit and natural this feeling may be to many of us (especially privileged groups), is experienced by most people as crucial to their well-being. Yet, in Sennett’s case, cultural or ethnic belonging is rejected, for although it might be tempting to feel more at home here than somewhere else (here, in this neighbourhood, in this ethnic community), uprootedness and painful interaction are the real human condition and at the same time the price and the solution for urban peace.

Sennett’s characterization of Simone Weil’s stress on the importance of rootedness – according to Weil ‘perhaps the most important and least recognized need of the human soul’ (Weil, [1952] 2002: 40) – as ‘romantic-reactionary sentimentality’ is indicative of Sennett’s own position in this regard. This is not the place to discuss Weil’s philosophy. ²¹ My point is that a sense of belonging – ‘roots’ – is simply rejected with the claim that ‘difference, discontinuity and disorientation’ ought to be our moral framework (Sennett, 1990: 228, 226). Characteristic in this regard is Sennett’s article ‘The Foreigner’ (1996). Whereas ‘pluralism’ is presented in terms of ‘a segregating game’, a ‘self-enclosure in ethnicity’ imprisoning people ‘in a Balkanized, unequal city of differences’, Sennett argues that migrants and, more generally, urbanites should become and stay strangers, for ‘uprooting has a positive moral value’ (1996: 191, 193, 186). We have to be estranged from ourselves and what is familiar and embrace the creative process of self-transformation, because only a ‘disturbed sense of ourselves’ will prompt us ‘to turn outward toward each other’ (Sennett, 1994: 374). The maxim of displacement becomes: ‘Look in the mirror and see someone else’ (Sennett, 1996: 194). Have a nice day!

This celebration of the status of the uprooted foreigner, being estranged from himself and his socio-cultural context, strikes me as frivolous. This type of urban cosmopolitanism seems simply a lifestyle for the privileged, bourgeois elites. To what extent is this rejection of ‘roots’ and the celebration of ‘individual liberty’ based on a privileged position where belonging is already a lived reality, though implicit and taken for granted? What Sennett seems to take no notice of in his plea to make ‘people become truly cosmopolitan’ (2001: 1) is the fact that people – and migrants are especially vulnerable in this regard – can suffer from a lack of cultural and social integration, from a loss of feeling at home. This suffering, in some cases, even leads to mental illness.

There is a fairly large literature on the so-called ‘ethnic density effect.’ This effect refers to a phenomenon whereby adverse mental health outcomes appear among members of minority groups who live in neighbourhoods where they comprise a relatively small proportion of the population. The current debate among social scientists is not so much whether this effect exists, but what the mechanisms behind it are (Pickett and Wilkinson, 2008; Whitley et al., 2006). The first empirical study to show convincing evidence for this effect dates from 1939: Faris and Dunham concluded that ethnic minorities in Chicago had ‘greater chances for mental breakdown and personality disorganization, especially when a person is living in an area not primarily populated by members of his own group’ (Faris and Dunham, quoted in Whitley et al., 2006: 376–7). Since then there has been a growing body of evidence, ‘fairly consistent across time, national boundaries and different ethnic groups, that ethnic group density affects mental health’ (Pickett and Wilkinson, 2008: 326).

Furthermore, the ethnic density effect has been found for a variety of mental health outcomes, including neurotic symptoms, incidence of schizophrenia, deliberative self-harm and suicide. There are also studies that found a positive effect of ethnic density on physical health (Smaje, 1995). The required degree of density to exert a protective effect on health and the mechanisms behind the density effect are still unclear. Some possible candidates for the latter are the adverse effects on mental health of stigma/racism and the beneficial effects on health of social capital. ²² However, what is key to my critique of Sennett is the existence of the effect itself. This clearly shows that ‘displacement’ and ‘uprootedness’ have a sinister side that Sennett either ignores or – and this is less likely – is unaware of.

To present the argument in favour of urban clustering in terms of avoiding mental illness might seem like an over-dramatization. But the empirical data are indicative of the point I am trying to make. Urban communities can have a protective function by providing a sense of belonging and cultural orientation in the midst of the larger urban environment. This explains the well-established fact that immigrants generally attach great importance to living in urban areas in which a sizeable number of co-ethnics live (though a certain mix is often preferred to complete homogeneity) (Bolt et al., 1998; Bowes et al., 1997; Ehrkamp, 2005; Phillips et al., 2007; Van Ham and Manley, 2009). ²³

Some theorists, though in a somewhat different context, make a distinction between a positive and a negative justification for certain types of recognition of identity. ²⁴ The positive formulation argues for recognition of identity in terms of the contribution thereof to the formation of a healthy and intact sense of self, while the negative justification focuses on the damage to identity that a denial of respect brings about. In many cases the negative formulation is more powerful for the reason that a norm often only becomes visible when it is being violated (Margalit, 1998). The third argument in favour of allowing urban clustering around culture of ethnicity can be considered a negative justification thereof.

It would not be enough, however, to allow urban communities without establishing and cultivating institutions that, in addition to a sense of belonging, offer sufficient opportunity for challenging interactions with ethnic, cultural, religious and socio-economic ‘others’ that are part of the city or the nation. ²⁵ I speak of ‘challenging interactions’ for the reason that dealing with those that are different on some social axis can involve a level of insecurity or even unease. In that sense, ‘well-being’ cannot be the only guide to urban living together.

But here we should accept that interaction always entails risks, and that the alternatives are much less attractive. One alternative is to think of the city as a collection of tectonic plates, where internally homogenous and hard-edged pieces only occasionally come into contact, accompanied by violent eruptions destroying a larger sense of loyalty and citizenship. This is the city Sennett fears and rightly so. The other alternative, however, is the city as an inhospitable place to a sense of belonging – be it from the perspective of radical agonism or from the perspective of cultivating public anonymity at all costs, as a kind of fetish. The city should neither be a collection of different planets nor a place of anti-communitarian hybridity. Let us focus on making the city work as a place of roots and of wider and more challenging connections, a place where the notion of living together has many different layers. That type of city, where there is both a sense of urban belonging and challenging interaction, seems the best we can aim for in an age of migration and urbanization.