Introduction

The expression “Right to the City” (henceforth, RTC), originally coined by Henry Lefebvre (1974), has gained increasing popularity as both an analytic and a mobilizing tool (Harvey 2003; Marcuse 2009; Mayer 2012; Domaradzka 2016). Researchers interested in urban dynamics have referred to RTC as a broad framework for the interpretation of local struggles for participatory democracy, against the gentrification of the urban space, and for a more equal and widespread fruition of the urban social and physical space. As for activists, they have used the RTC idea to locate their own specific initiatives in a broader context, connecting them to other campaigns and developing on that basis some kind of collective identity. In other words, RTC might have functioned as a “master frame” (Snow and Benford 1992), or a “condensing symbol” (Gamson 1992), capable of assigning common meaning to a number of grassroots actions and struggles, that might have otherwise been perceived as quite disconnected. Referring to RTC may certainly help to bring together urban activists and/or policy makers engaged in a range of heterogeneous issues and campaigns, reinforcing solidarity and mutual understanding between them. Likewise, it may foster intellectual debate by providing foci to disparate lines of investigation and theorizing.

At the same time, adopters of the RTC concept have followed the fairly common path of over-stretching it, loading it with additional properties and qualifications, and/or undermining and reducing its original radical Marxist-infused meaning, what Merrifield (2011, p. 473) refers to as a “bourgeois re-appropriation.” The concept has also been increasingly incorporated in the language of policy makers, urban developers and international agencies, with the United Nations World Urban Forum and the World Bank adopting it in their charters to address global urban poverty (Mayer 2012; Domaradzka 2016). It is also disputable whether RTC can accurately represent recent waves of urban mobilization in their entirety; the conditions faced by urban activists from the North and the South of the world are obviously not the same, and Lefebvre’s original formulation referred to a quite different time and geographical context (Lopez de Sousa 2010; Merrifield 2011; Mayer 2012, pp. 78–80; Parnell and Robinson 2012; Domaradzka and Wijkstrom 2016, p. 304).

In this paper we don’t aim at reconstructing the debate about the development and uses of RTC. Rather, we propose an approach that enables us to separate more neatly the analytic from the mobilizing dimensions of the concept, and test its impact on collective action patterns. To this purpose, we start from the basic observation that, like any other instance of collective action, urban struggles most often develop in the context of “civic organizational fields”—by which we mean sets of actors engaged on a voluntary basis in the promotion of collective action and the production of collective goods (Diani 2015, pp. 12–13). Such fields are inhabited by a plurality of actors (individuals and organizations) with highly variable agendas, ideological and cultural models, and styles of action. Such actors may relate to each other in very different ways, ranging from sustained cooperation to competition to open hostility. Assessing the role of RTC implies, therefore, addressing two broad sets of questions. The first set refers to the relation between the expression “RTC” and other frames that are also frequently invoked to describe urban struggles and issues. In other words, what is the position of RTC-inspired frames and narratives among the broader set of cultural signifiers, used to characterize urban struggles? To what images of conflicts are references to RTC most strongly connected by activists, when they represent their initiatives? For example, are ideas attached to RTC linked to an anti-capitalist imagery or to a modernizing one? Are they most popular among actors adopting a confrontational logic or among those pursuing an incremental change of existing assets? By empirically addressing these questions, we intend to explore to what extent RTC represent a coherent way to link together and summarize a set of themes related to the use of urban space and to broader ideas about urban democracy.

The second set of questions has to do with the role of representations in shaping the structure of action fields. Does the RTC frame affect the structure of urban civic fields in significant ways? How does it fare by comparison to other systems of meaning that also circulate in the same environments? The question here is whether reference to RTC (or lack of it) affects the probability of urban actors working together and/or developing stronger connections, e.g., through their mutual identification as members of the same movement, or, in the case of organizations, through shared members. It’s also important asking whether such effect is stronger than that of other frames frequently used in reference to urban issues and struggles. In order to tackle these questions, we combine insights from social movement theory (Diani 2013, 2015), organization theory (Ahrne et al. 2016; Ahrne and Brunsson 2011; Dobusch and Schoeneborn 2015), critical urban studies (Marcuse 2009; Mayer 2012) and social network analysis (Monge and Contractor 2003). By doing so, we hope to sketch the contours of a network analytic approach to the relation between symbols and the coordination of collective action.

We illustrate our approach with evidence from one specific case from the global South, looking at the collective action field constituted by citizens’ groups and associations active on a variety of urban issues in Cape Town, South Africa. Cape Town represents a relevant setting to explore the link between symbols and the organizing of collective action, given the extreme differentiation of post-apartheid civil society. Alongside Durban and Johannesburg, Cape Town has been one of the most fertile grounds for grassroots campaigning on urban deprivation and inequality in South Africa, in some cases under the RTC label (Bond 2004; Parnell and Pieterse 2010; Morange 2011; Görgens and van Donk 2011; Mottiar and Bond 2012).

In the course of our fieldwork, conducted between 2013 and 2014, we obtained full data on 129 groups and associations located in areas from the very affluent to the very deprived. These were chosen based on an original sample of urban-focused and socioenvironmental groups based on newspaper searches and then snow-balling to identify organizations mentioned by at least three other organizations to include all organizations active locally within the highly unequal area stretching from Constantia to Lavender Hill, adding also local organizations in Gugulethu and Khayelitsha, plus all organizations active at city to national level. Interviews were conducted with representatives from each group or association, on the basis of a structured questionnaire with some open questions (for more details on the project see http://www.situatedecologies.net/archives/portfolio/ct-civnet-cape-town). Differences in levels of affluence largely overlapped with racial divides (Table 1). About one quarter of organizations in our population were based in white neighborhoods, and about half in areas which were predominantly inhabited by colored or black people. The remaining quarter were based in “mixed” areas, in which there was no dominant racial group according to the 2011 Census (see the footnote to Table 1 for further information on the classification of residential areas in terms of race). Organizations also differed quite substantially in their formal structure, with two-thirds being membership based and the rest relying on (semi)professional staff, as well as in their repertoires of action. These ranged from confrontational, grassroots contention to incremental approaches closer to a pressure group style, prioritizing lobbying to obtain specific changes on specific issues. Both distinctions refer to tensions which have been repeatedly identified within RTC campaigns (Mayer 2012). However, they take a specific, magnified form in the South African context, given the extent to which certain groups, but by no means all, representing dispossessed sectors of the population rely on disruptive techniques, partly perhaps for lack of the organizational resources that might support alternative strategies (Mottiar 2013; Mottiar and Bond 2012). While generally very critical of the political establishment, a significant minority of organizations in Cape Town also engaged in collaborations with major political actors, most notably, local chapters of ANC (about one-fourth of respondents) and Cosatu—Confederation of South African Trade Unions (one-fifth), as well as (if to a smaller extent) the Democratic Alliance.

Table 1 Basic traits of civic organizations in Cape Town
Full size table

Networks of Symbols: Framing Civic Activism in Cape Town

Social network analysis is a powerful tool to explore the connections between symbols and their clustering in broader frames. The relation between culture and social networks has long been highlighted (Pachucki and Breiger 2010). This applies not only to the general insight that relations are culturally constructed (for a discussion related to social movements, see Mische 2003). It also holds true for dual relations that may be detected between cultural elements and networks structures: on the one hand, culture may shape network structures, e.g., by activating homophilic mechanisms (for some examples referred to civic fields: Diani 1995; Carroll and Ratner 1996; Mische 2008); on the other hand, network analytic concepts and tools have been increasingly used to study the relations between cultural elements, such as concepts or symbols, as exemplified by semantic network analysis (e.g., again in reference to civic networks: Pavan 2012). In our paper, we take into account both sides of this duality. In this section, we investigate the way in which the RTC concept is linked to other symbols representing the activity of urban civic groups in Cape Town; in doing so, we identify a number of core frames that may represent in a parsimonious way the issue priorities of local civil society. In the next section, we will look at the salience of such themes in shaping network patterns; in other words, we will ask whether identification with RTC as well as with other, partially competing/partially complementary, frames, results in specific relational patterns.

Organization representatives were asked to identify, out of a list of 32 issues (that included RTC), up to three which were most important to them, and then up to five which they considered of interest. The full list of 32 issues had been assembled based on a survey of city-wide and neighborhood newspapers and interviews with key informants, including activists and scholars. We combined those two answers into a single dichotomous indicator of perceived relevance. The wording of these items was broad enough to allow us to treat them as general symbols, summarizing broad themes of potential concern to our interviewees. Data collected on participation in specific events provide more details on the specific initiatives in which these general issues are mobilized. Out of 129 respondents, about one-fifth (25) identified RTC as one of their priorities. At the same time, we also tried to identify the symbols to which RTC was mostly associated with, based on the eight issues that each organizational representative chose in the survey. A principal component analysisFootnote 1 produced four components, displayed in Table 2 along with the percentages of groups that expressed interest in each issue. Three specific issues did correlate poorly with the others, on top of attracting limited interest, and were therefore excluded from further analysis (see bottom of Table 2).

Table 2 Structure of main issues addressed by Cape Town organizations based on a principal component analysis of 32 issues (the first column reports percentage of respondents interested in an issue)
Full size table

A first set of issues, strongly related to each other, can be labeled as reflecting a concern with “global environmental justice.” It includes both an environmental sensibility for themes like biodiversity and genetically modified food (GM) and a more socially oriented agenda addressing trade, fiscal and financial issues. Another set of issues combines instead in what looks closer to an “urban conservation” agenda, as the items that load strongest on this component are classic environmental ones such as nature conservation, green commons, pollution or even climate change (that curiously does not connect strongly to the “global environmental justice” approach). There is, however, also attention for themes like urban farming and food security, which have clear social implications, and cultural heritage. The connection between heritage and environmental issues was actually central in some prominent local initiatives. For example, the campaign to protect the wetland and lake area of Princess Vlei from the commercial development of a mall drew heavily upon the cultural heritage of the everyday use of the area by people classified as colored under apartheid. Activists also emphasized the wetland’s connection to the aboriginal Khoi population of the area (Ernstson 2013; Ernstson and Sörlin 2013). The third component is the most relevant for the present paper as it includes the RTC issue among a set of items that define a clear agenda for urban change from the grassroots. It combines the quest for a rejuvenated participatory democracy, free of corruption, with an emphasis on basic services and community development, all themes present in the “rights to the city” narrative. While there is no ultimate shared definition of what falls under the RTC heading, discussions of the concept repeatedly point at some of the dimensions also evoked here: on the one hand, the right of ordinary citizens, not just the elites or the achievers, to enjoy in full the opportunities offered by the urban space; on the other, their entitlement to participate in collective decision making about urban development—i.e., to participate as equals in shaping or producing urban space—rather than accepting it as a top-down, often strongly bureaucratic and/or market-driven, process. This lies at the very heart of Lefebvre’s formulation (1974; also see Mayer 2012; Domaradzka 2016). Finally, the fourth component reflects the explicit polarization between a view of urban processes focused on what we could term “sustainable city,” reflecting an interest in issues such as alternative energy, growth reduction, public transport (as well as pollution, that also loads strongly on this component); and a view of an urban agenda in terms of “social rights” (and community development) for the weakest sectors of society (labor, female, youth, minorities). Interestingly, “global justice” also loads strongly on this component, yet correlates with the environmental element and not with the social rights one. Each of these components can be interpreted as a cognitive frame, bringing together those symbols that are perceived as closest by civic activists.

Another way of representing connections between issues/symbols is by building a matrix where the connection between two issues is given by the number of groups that identify both as relevant to their agenda. As the levels of interest across the 32 issues vary considerably, ranging from 2 to 57%, we do not use the absolute numbers in our analysis. Instead we use the Jaccard coefficient that weighs the interest expressed in each pair of issues against the overall amount of interest expressed in those issues in relation to all other 32 issues (Borgatti and Halgin 2011, p. 421).Footnote 2 “Appendix” reports the matrix of ties, measured as Jaccard coefficients, between the 32 issues analyzed. Here we only report the smaller matrix, generated by the partition of the original matrix into six different blocks, based on the principal component analysis displayed in Table 2. These consist of the “isolates” (i.e., the three issues that did not covariate with the others) plus the blocks corresponding to the four components identified in Table 2. However, we have chosen to split the component based on the polarization between sustainability and social rights into two blocks, in order to substantively enhance its interpretation. Table 3 shows the distribution of densities within and across blocks. With the obvious exception of the “isolates,” in all the other cases, the issues belonging in the same component turn out to be linked to each other in a particularly strong way, well above the average density of the matrix. These are the cells on the main diagonal, the values of which all exceed the average density of 0.100 (even if the density of the “sustainability” block only does that marginally, 0.141 vs. 0.100). If we look at the connections between the different sets of issues (given by the off-diagonal cells), we note the separation of the “global environmental justice” issues from the others (all their scores are well below 0.1). In other words, only a few groups express a sustained interest in global justice and in issues more focused on the urban dimension. To the contrary, RTC and “urban conservation” show significant degrees of overlap, suggesting quite high compatibility between the two agendas. As for sustainability and social rights, there is little overlap between the two, as also suggested by the principal component analysis in Table 2. Both sets of issues appear compatible with both the “urban conservation” frame and the RTC frame (in the sense that they have non-negligible connections), yet the social rights frame appears to be more strongly integrated to the others than the sustainability one. This interpretation is supported graphically by Fig. 1 that shows “global environmental justice” issues clustered together on the left of the graph, while the other issues cluster on the right. In the latter case the distance between the “sustainability” and the “social rights” frame is pronounced, with “urban conservation” and RTC frames located in between.

Table 3 Reduced block matrix of ties between frames using Jaccard scores (see main text for explanation and “Appendix” for the matrix showing ties between all 32 issues)
Full size table
Fig. 1
figure 1

Graph of main issues in Cape Town based on survey of 129 organizations. Legend (blocks defined by principal component analysis): black square: I. Global environmental justice; green up triangle: II. Urban conservation; light blue diamond: III. Right to the city; red down triangle: IV. Sustainable city versus social rights. Note A link between a pair of issues is shown when the strength of the connection between them exceeds the average strength of the Jaccard index for the entire network (0.100) (Color figure online)

Full size image

We can obtain a simpler representation of the relationship between the different frames if we focus on the “image matrix” of the same network, which means to let all cell values below 0.1, the average density of the network, be represented by a zero, and all above the average by one (Table 4 and Fig. 2). It clearly illustrates the fact that “urban conservation” and RTC frames operate as a bridge between quite different conceptions of urban development, one centered on the notion of sustainable growth, the other, around the protection of basic social rights for the most dispossessed groups in society. The deepest gap is, however, between the approaches that frame core issues in reference to global dynamics, and the other issues. Global issues occupy a distinctive position: apart from being the preserve of small minority of organizations, they are distinctly cohesive and poorly connected to other agendas (with the exception of the “global justice” issue, see below). As for RTC, the issues associated with this frame are certainly central both in terms of the attention they attract and in terms of their connections to other major frames, most particularly “urban conservation.” At the same time, they do not seem to occupy a dominant position in the network as opposed to other frames. In other words, they are certainly relevant but neither the only relevant frame, nor the most important one.

Table 4 Image matrix of various mobilizing frames in Cape Town
Full size table
Fig. 2
figure 2

Image matrix of connections between frames (smaller nodes indicate low internal density; a link between two frames is shown when the strength of the average connection between the issues associated with those frames exceeds 0.100, i.e., the average strength of the Jaccard index for the entire network)

Full size image

The Salience of RTC in Alliance Networks

Having assessed the plausibility of treating RTC as a distinctive frame, the next question is of course whether that frame is salient enough to affect patterns of alliances within local civil society, and how does it fare in that regard by comparison to other frames like “urban conservation” or “global environmental justice.” In other words, are civil society organizations that identify with RTC more likely to develop alliances among themselves than with other organizations? And are other frames similarly (or perhaps, more) influential in structuring alliance patterns? As such, these questions are far from new, as the role of symbols in shaping collective action fields has long been recognized, in reference to policy networks (e.g., Jenkins-Smith and Sabatier 1993) as well as to social movements (e.g., Melucci 1996). However, recent developments have focused more explicitly on the steps required to move from an aggregative to a relational view of collective action dynamics—i.e., an analysis of fields as systems of variously coordinated, if still independent, entities, rather than as disconnected populations of discrete actors.

Diani (2013, 2015) has suggested that several “modes of coordination” may be operating within each organizational field, defined by different combinations of two broad classes of mechanisms: resource allocation and boundary definition. In particular, social movements represent a peculiar mode of coordinating collective action that implies sustained dense exchanges of resources and relatively tight symbolic boundaries between agents. This view of social movements strongly resonates with approaches, increasingly popular among organization scholars (Den Hond et al. 2015), which see them as a distinctive form of “emerging social order.” Contributing to this insight, while building further on it, some organizational analysts have suggested a terminological and conceptual shift, which aims to preserve both the distinctiveness of organizations as a particular way of coordinating human behavior, while overcoming rigid dichotomies between organization and other forms of coordination, such as networks or institutions. These scholars have proposed to move the focus from “organizations” to “organizing” (Ahrne et al. 2016; Ahrne and Brunsson 2011) or “organizationality” (Dobusch and Schoeneborn 2015), i.e., to the processes through which forms of coordination are created. Such processes may not always display all the defining properties of organizations, yet perform some of their functions of coordination and building identity, which has allowed observers to speak of “partial orders” (Ahrne and Brunsson 2011). This is the case of social movements that rely on continuous informal negotiations between their activists to allocate resources since strategies and tactics are not decided in hierarchical fashion (even though hierarchy may well emerge out of repeated interactions, see, e.g., Diani 2003; Ernstson et al. 2008). The definition of “membership” is furthermore based on mutual recognition between agents, and also the other dimensions, associated with formal organization by Ahrne et al. (2016, p. 95)—the setting of rules, the monitoring of members and the sanctioning of non-complying behaviors—are all, in social movements, dependent on informal interactions.

Looking at the Anonymous hacker community, Dobusch and Schoeneborn (2015) have pointed at the role of symbols in enabling even very loose collectivities to develop some capacity to act in a coordinated way. Along compatible lines, social movement theorists have emphasized the role of symbolic production (“framing”: Benford and Snow 2000) in building bridges between multiple agents, thus facilitating collective action. While the bulk of this literature has focused on micro dynamics such as individual recruitment (Snow et al. 1986), a few have looked at how shared master frames may facilitate the integration of civic organizational fields (Carroll and Ratner 1996; Di Gregorio 2012; Diani 1995; Saunders 2013). Exploring this dynamic is crucial to current debates about RTC because the existence of a number of groups and associations adopting this frame to represent their agenda does not necessarily imply the presence of a “RTC movement” in the absence of some degree of coordinated interaction between the agents committed to that particular frame of urban issues; nor can we speak of a “RTC movement” if those very agents fail to develop some degree of solidarity and collective identity.

Accordingly, in what follows we explore the impact of the RTC frame on alliance patterns and thus on the informal organizing of a collective action field. We ask in particular whether identification with the RTC frame identifies sectors of civic fields that stand out for the particular density of the ties between their incumbents. In other words, are there clusters of organizations that interact heavily with one another and that also share a particular symbolic frame? In our empirical analysis, we consider organizations associated with a frame if they have identified as important at least two of the issues that load at level 0.3 or higher on the component corresponding to that frame (Table 2). The size of these non-mutually exclusive groups varies considerably, from “global environmental justice” (7% of the population) and “sustainable city” (9%) via “social rights” (31%) to “urban conservation” (57%) and RTC (60%). In what follows we refer to an inclusive notion of alliance. We equate it to resource exchanges in general, without differentiating between those that also imply deeper personal connections between members and those that do not (what Baldassarri and Diani 2007 denoted “social bonds” and “transactions,” respectively).

Figure 3 reports cooperation between organizations in Cape Town, regardless of whether they identify with RTC or not. Among the organizations that adopt RTC as a frame are organizations like the Princess Vlei Forum (114_PVF), that coordinated the homonymous successful campaign to stop the building of a shopping mall on undeveloped land that had been widely used by the local communities for recreational purposes; the organization coordinating the “Right to know” campaign, an initiative aimed at stopping the Protection of State Information Bill, also dubbed the Secrecy Bill, discussed by the SA parliament from 2010, and to enhance rights to access to public data (108_R2 K); as well as the local chapter of Abahlali baseMjondolo (607_ABM), a shack dwellers’ organization most prominent in Durban (Pithouse 2009; Mdlalose 2014), and one of those that most explicitly adopted the reference to RTC as a frame for their activism (Mathivet and Buckingham 2009; Mayer 2012).

Fig. 3
figure 3

Graph of resource exchanges among civic organizations in Cape Town (organizations represented as small black squares do not identify with RTC; bigger nodes do)

Full size image

Data show that the large majority of organizations that do identify with RTC (58 out of 78, about three quarters) are actually connected into one single main component (Fig. 4). At the same time, however, even a superficial inspection of the graph suggests the presence of substantial exchanges between organizations across the cleavage that might have been created by this particular type of framing. This is confirmed by the overall distribution of ties both within and across the sets of organizations that do and do not identify with RTC. Differences between subgroups turn out to be non-significant (Table 5). We are actually facing an “amorphous” organizational model (Breiger 1979, p. 30) in which no relational pattern based on RTC is discernible, and ties evenly distribute across the two groups. It seems safe to conclude that the RTC frame shows no salience within the network, i.e., no structuring power when it comes to alliance building.

Fig. 4
figure 4

Graph of resource exchanges among organizations identifying with RTC as an encompassing frame

Full size image
Table 5 Density of resource exchanges by identification with the Right to the city (RTC) frame
Full size table

Two qualifications are in order though. First, although a density score in the region of 0.015–0.020 may appear small, it is not so much outside the “norm” identified by the (still rare) available studies of networks of voluntary organizations. For example, a study of civic networks in Glasgow (0.023) and Bristol (0.015), where the size of networks was highly comparable to the one in Cape Town (124 and 134 vs. 129), found similar densities (0.023 and 0.015, respectively: Diani 2015, p. 73). Developing and sustaining alliances is a demanding task and the capacity of each voluntary organization in that regard is limited (Diani 2015; Knoke 1990). Second, the fact that RTC does not operate as a factor structuring alliance building does not mean that organizations identifying with this frame play a limited role in local civil society in Cape Town. They are a significant force there, not just because of their number but also because of their involvement in a number of major local campaigns. Table 6 reports data on twelve campaigns that run between the late 2000s and early 2010s. The two most popular campaigns, attracting almost one quarter of surveyed organizations (“Right 2 Know” and “Stop the Mall at Princess Vlei”), resonate strongly with RTC core ideas; several other campaigns are also close to some basic tenets of the RTC approach. At the same time, it should be noted that those campaigns were similarly popular among organizations not adopting the RTC to characterize their agenda.

Table 6 Involvement in campaigns by organizations with variable identification with the Right to the city (RTC) frame
Full size table

While RTC does not shape civic networks, some of the other frames identified in the first section of this paper actually do. We conducted the same analysis using the autocorrelation routine in the software program Ucinet to search for significant homophilic or center-periphery dynamics. We did that for three networks: the one consisting of all resource exchanges; the one consisting of “strong ties” only (i.e., multiple ties in which resource exchanges overlap with ties created by activists on a personal basis, and ties of solidarity toward the other organization); and the one in which a tie was posited if two organizations shared identification with the same social movement.

Table 7 reports our findings for each network. More precisely, it reports the levels of significance for differences in tie density between organizations that identify and do not identify with a certain frame. Similarly to what we have just shown for the alliance (“any tie”) network, reference to RTC also shows no salience for the “strong ties” and the “shared movement identity” networks. The same applies to the “sustainable planning” frame and also largely to the “social rights” one (where the distribution of “any tie” shows significant differences, but against the expected direction; organizations that are not interested in “social rights” are more likely to be connected, while the opposite holds for those who are committed to them). In contrast, we find significant effects for the other two frames, “global environmental justice” and “urban conservation,” at all relational levels. As only nine organizations were committed to the former, we can concentrate on the latter that proved very popular among Capetonian organizations. The graph shown in Fig. 5 and the densities reported in Table 8 (right section) illustrate the distribution of ties across organizations that identify with the “urban conservation” frame (brighter nodes) and those that do not (smallest, black nodes). Table 8 in particular suggests a “multiple caucus” structure (Breiger 1979), with two clusters of actors that are significantly connected among them, but with markedly less exchange between clusters. The capacity of the “urban conservation” frame to structure local civil society seems to also depend significantly on persisting racial divides. Data on the racial composition of the neighborhoods in which the different organizations are based suggest that the “urban conservation” frame is significantly more popular among organizations located in more affluent and white-dominated neighborhoods (Table 9). This gives some support to the claim that white class-based interests in post-apartheid South Africa has become, at least partially and within civil society, organized around quite narrow nature conservation and ‘green issues’ (Cock 2006).

Table 7 Salience of different frames over different types of civic organizational networks (levels of significance reported, where NS means “not significant”)
Full size table
Fig. 5
figure 5

Graph of resource exchanges among civic organizations in Cape Town (organizations represented as small black squares do not identify with the “urban conservation” frame; the size of nodes is scaled to their degree centrality)

Full size image
Table 8 Density of resource exchanges by identification with various frames
Full size table
Table 9 Identification with the “urban conservation” frame by residential area of organization (p < 0.001)
Full size table

Finally, the organizations identified with “urban conservation” and “global environmental justice” frames are the only ones to display the relational multiplexity that Diani (2015) associates with a social movement mode of coordination: they are significantly more connected among themselves both in terms of generic resource exchanges and in terms of the stronger ties created by activists’ multiple involvements in several organizations that characterize processes of boundary definition (Diani 2015, chs. 1 and 4). They are also linked by common identification with specific social movements. However, they do not stand out for an explicit adversarial approach to collective action. Organizations identifying with the “global environmental justice” frame are as likely as other actors to identify specific opponents, whether public or private. As for those that adopt the “urban conservation” frame, they are even significantly more reluctant than the rest of the organizations we surveyed to identify opponents (12 vs. 18%). This suggests a mild tendency toward a “consensus movement” mode of coordination, in which actors coordinate at multiple levels yet do not identify any social or political opponent, preferring to mobilize in reference to broadly shared values and goals (Diani and Bison 2004). Needless to say, the RTC frame does not show a social movement mode of coordination because organizations that mobilize on RTC issues do not have significantly more ties among each other than what they have to the rest of local civil society.

Conclusions

In this paper we have illustrated an approach to explore how frames shape alliance patterns between multiple agents in a specific organizational field. More specifically, we have looked at a specific symbol, Right to the city, or RTC, in a specific context, civil society in Cape Town, hugely divided across racial and class lines. We have done so in two steps. First, we have looked at the semantic network consisting of a set of symbols (we would not go as far as claiming “all symbols”) through which civic actors characterize their main issues. Through a combination of standard data reduction and network analytic techniques, we have located the set of symbols to which RTC is most strongly correlated. Broad theoretical arguments about the nature of RTC and its combining aspirations to the democratization of urban politics with the full use of the urban space by ordinary citizens have largely been supported by our data. As a broad frame, RTC has proved to be a popular shared representation of urban issues in Cape Town linking to issues of local democracy, housing, and access to public and basic services. At the same time, RTC has not been the only significant frame among Capetonian civic organizations, as “urban conservation” has also attracted comparable attention.

As a second step of our analysis, we have explored the extent to which symbols and frames have shaped alliance patterns within civic organizational fields, what we have called their “salience.” In that regard, the RTC frame does not seem to have any impact on the choices of alliances in Cape Town, while other frames do, most significantly, “urban conservation.” Alliances are more likely to be found not only between organizations that share it, but also between organizations that do not adopt it. This suggests a significant divide along “urban conservation,” which was not the case for the RTC frame. As we have just seen, this divide is at least partially embedded in race cleavages: the urban conservation frame is most popular among organizations located in areas with a dominant white population, which under apartheid were areas classified as for whites only. It is also worth asking what might lie behind the fact that organizations distant from the “social rights” frame are quite cohesive, while those identifying with it are not. Class dynamics and thus a general lack of resources in terms of money, time to meet and organize, etc., might well be behind this difference, given the notorious difficulty for the most dispossessed sectors of the urban population to organize beyond the boundaries of their specific community and to engage in broader alliances. This leaves the organizations working on behalf of the urban poor and working class, several with a focus on social rights, quite isolated, as they cannot sustain longer-term alliances.

Altogether, RTC is certainly playing an important role in civil society campaigns. Many organizations identify with it, and many campaigns are conducted which are consistent with its basic tenets. At the same time, it does not produce a distinctive relational pattern within the broader civic organizational field; moreover, organizations identifying with RTC are also engaged in campaigns on issues which are relatively distant from RTC, such as nature conservation or assistance to migrants (see Table 6). It seems therefore difficult to speak of a distinctive “RTC movement.” Still, in a society so deeply divided along race and class lines, the limited salience of the RTC frame in shaping alliances might perhaps be an asset. It could be seen as a master frame capable of bringing together different sectors of civil society in a way that “urban conservation,” still shaped by racial divisions, fails to do. It would not correspond to a specific mode of coordination, but would rather provide a ground for different types of actors to meet. Indeed, after the conclusion of our fieldwork in 2014 there have been broader-based efforts under the call of “Reclaim the City” (see, e.g., Hendricks 2016), emerging out of the active Ndifuna Ukwazi organization, which was part of our survey. Other examples include mobilizations that have linked broader notions of urban rights and sanitation (see, e.g., McFarlane and Silver 2017).

It is worth closing by asking to what extent our findings, no matter how partial, contribute to our understanding of RTC initiatives on a broader scale. After all, the South African case stands out at the very least for its extremely deep levels of inequality and the persisting relevance of racial divisions. Still, the configuration of symbols and relations we have charted in Cape Town seems to present several analogies with similar processes in other parts of the world. Mapping the re-emergence of urban mobilizations in Poland in the early 2000s, Domaradzka and Wijkstrom (2016) noted that while RTC had certainly played a significant role in bringing together some local initiatives and broadening their agenda, it did not represent by any means the entirety of recent Polish urban struggles, nor did it provide a covering frame for them. Our own study, while quite different in scope, method and geographical context, seems to point in the same direction.