Abstract
When protests break out in downtown Kampala they tend to transform a fluid urban environment into bounded political camps, and myriad ambiguous concerns into comparatively clear-cut political issues. This article traces this process and conceptualises Kampala's urban politics as a politics of crystallisation: as attempts to structure highly fluid dynamics into something concrete. The article is based on ethnographic research amongst opposition activists and the police forces. Both seek to activate political boundaries and make people gravitate towards their respective side. But in line with the fluidity of urban everyday life, they also work and collaborate across these boundaries. The national regime and the opposition thus function not as permanent, stable structures, but as processes, as fields of gravity whose emergence is incited and inhibited, financed, and policed. Drawing on Gilbert Simondon's theory of individuation and AbdouMaliq Simone's work on urbanity, this analytical framework offers a dynamic reading of urban contentious politics in general, and a reinterpretation of the paradoxes of power in African politics in particular.
Hybrid regimes embody “paradoxes of power,” Tripp (2010: 9) argues; they exhibit divergent realities “depending on what processes or actors one is analyzing.” As Tripp shows, Uganda is an exemplary case in this regard. On first sight, for instance, the Ugandan state, dominated for over 30 years by the National Resistance Movement (NRM) under President Museveni, seems extremely stable. Internationally backed as a geostrategic ally of the United States, nationally supported by a wealthy business elite, equipped with a massive security apparatus (Kagoro 2014), the state institutions and the governing party are fully intertwined in an intricate political patronage system that secures the overall status quo. At the same time, however, there is a widespread sense of instability within “the political matrix in Uganda,” which is “fluid and constantly changing,” resulting in a variety of frictions within the NRM and indicating that “the biggest threats to the NRM's power emanate from those closest to Museveni” (Tripp 2010: 64, 56). Instability also emerges from outside the regime. The Kayunga riots in 2009, the Walk-to-Work protests in 2011, and the deadly clashes between opposition supporters and police during the 2016 election campaigns have manifested profound tensions and frustrations among large parts of the urban population, who feel marginalised from the supposedly ubiquitous national development and enraged by widespread corruption. And yet, to add another paradox, “corruption” and “Big Man politics” (Utas 2012) also tended to be intricately engrained in the very dynamics of popular resistance (Philipps 2016).
Africanist political sociology is shot through with such contradictions, or, one should specify: paradoxes and contradictions have become a prominent way for academics to frame African political and social dynamics. The discussion on African youth was particularly explicit in this regard, with key titles conceiving youth as Makers and Breakers (Honwana and de Boeck 2005b), Vanguard or Vandals (Abbink and van Kessel 2005), and Hooligans and Heroes (Perullo 2005), highlighting that “children and youth are extremely difficult to pin down analytically [because] they often occupy more than one position at once” (Honwana and de Boeck 2005a: 3; see also Philipps 2014). But even the foundational discussion in the 1980s and 1990s on “the African state” (e.g. Médard 1982; Bayart 1993; Chabal and Daloz 1999) versus an African “civil society” and African social movements (e.g. Harbeson, Rothchild, and Chazan 1994; Mamdani and Wamba-dia-Wamba 1995) have demonstrated profusely how divergent, heterogeneous, and sometimes contradictory the realities are that emerge from African polities in the academic disciplines that reflect on them. In that empirical and academic context, Tripp (2010: 9-10) finds it problematic that “the limitations of social science […] make embracing two contradictory realities at the same time nearly impossible, [when] that is what hybrid regimes require of us.”
In this paper, which focuses on protests and policing in Kampala, we hope to provide a theoretical-methodological framework that makes embracing contradictory realities easier. 1 Based on the work of French philosopher Gilbert Simondon (Combes 2013; Simondon 1958, 1989, 1995), the key idea is to think of realities not in terms of “objects and substances” (Lakoff and Johnson 2003: 25) – for contradictions become problematic when one “thing” also appears to be something else – but in terms of crystallisation processes: as solid forms emerging from a fluid environment. Such a perspective traces politics in dynamic terms, examining how different actors – we refer to them as “political entrepreneurs” (Tilly 2003: 34) – work on different political realities as projects: how they construe and scheme political stability or instability, how they mobilise popular support for regime sovereignty or for anti-government protests, and how they generally create contours in a fluid political field by crafting in-groups and out-groups.
The authors thank Anna Baral, Elísio Macamo, Mats Utas, Sverker Finnström, Anders Sjögren, and the lively and critical voices at the Nordic Africa Institute in Uppsala (where the paper was first presented) for their valuable input. Many thanks also go to the two anonymous reviewers at Africa Spectrum.
We will elaborate this perspective in three steps. The first section outlines the fluid urban environment as a political context, taking the case of Kisekka Market, a market for spare car parts in downtown Kampala. At Kisekka, shopkeepers, hustlers, police spies, and middlemen are oftentimes unaware of who supports the opposition and who the government. While all are fated to interact with one another, politics remains within the realm of the intangible: less a matter of fixed and known identities than of dynamic and unknown identifications (see Brubaker and Cooper 2000). From this ambiguous context, however, highly unambiguous phenomena can emerge; Kisekka Market, for instance, has long been known (and stigmatised) for its protests and riots. In the second part of this article, we delineate how this transformation happens, or more generally, how political entrepreneurs attempt to turn a highly fluid environment into something concrete by sparking processes that engender political alignments. The third section looks at police, arguing, maybe unconventionally, that the police work through similar dynamics as Kisekka Market. With their fluid borders and vast networks, police forces extend way into the civilian population and sometimes even overlap with the very opposition networks that they seek to control. Moreover, since even the core of police is fraught with uncertainty and instability, the police are evidently part of what we call the “metastable city” – and to police it, they need to remain flexible and fluid while also creating or maintaining the impression of being a bounded state institution. Politics, from this perspective, is only secondarily about the conflict between a government and an opposition. Prior to this, and more fundamentally, it is about crafting concrete realities, groups, and movements that gather momentum and followers in a context of contingency and uncertainty (see Philipps 2016; Simone 2001; Utas 2012).
This idea draws heavily on the philosophy of Gilbert Simondon (1958, 1989, 1995; see also Combes 2013), currently gaining popularity as a strong inspiration for the work of Gilles Deleuze. 2 Simondon's primary concern is to understand how phenomena emerge as distinguishable and tangible, a process he calls “individuation.” This process, he argues, is best understood as an interplay of innumerable factors that are intangible in their diversity and heterogeneity, but gradually develop into definite forms. Specific ideas, for example, emerge from relations between our subconscious and our consciousness that cannot be clearly defined (Lakoff and Johnson 2003); artefacts emerge from a variety of contexts (cultural, material, historical, technological, etc.) that are never conceivable in their entirety (Burke 2002); and, to reiterate Simondon's (1995: 31) key metaphor, solid crystals emerge from a metastable solution that is initially fluid and intangible. African cities have widely been described in terms of similarly intangible relations by authors like Diouf (1996, 2003), Mbembe and Nuttall (2004), Pieterse (2011), and especially AbdouMaliq Simone (1998, 2001, 2003, 2004a, 2004b, 2008). Simone gives a captivating depiction of what one may call the metastable city: a space of potentiality and spontaneous formations where economic pressures require that large numbers of individuals eking out a living take chances and spontaneously follow cues, intuitions, social ties, and dynamics without knowing where these dynamics may take them. In this context, uncertainty is ubiquitous – as a constraint but also as a resource. As Simone (2008: 22) describes in the case of Bepanda Market in Douala, Cameroon, uncertainty provides
We also relied on Combes's (2013) English citations of Simondon, translated by Thomas LaMarre.
much room for dissimulation[] much room for making things seem as if they are real when they are not, or making them real simply through the sheer mobilizing of money, interest, or support on the part of those schemed or part of making a scheme.
The metastable city can thus be characterised as a site of crystallisation, where vague and fluid ambiguities can be turned into concrete realities that people work and align with to probe into their potentiality (Cooper and Pratten 2015: 1; see also Newell 2012). 3 In capital cities, such alignments easily gain an explicitly political character, for it is here where political power is continuously staged and attacked, solidified and undermined, through incessant negotiations of reciprocity and solidarity with the broader urban public (see Bekker and Therborn 2012; Christensen and Utas 2008; Diouf 1996, 2003; Utas 2012). To illustrate this rather abstract idea, we enter Kampala's former Kisekka Market as an exemplary site of metastability.
As a theoretical side note, this inverts how we tend to think about hybridity (see Albrecht and Moe 2015): while hybridisation implies, quite controversially, that previously bounded entities turn into a less distinct mix, the idea of crystallisation sees the indistinct mix as a universal primary condition from which intelligible entities, i.e. realities, unfold in indeterminate ways.
The Metastable City: The Case of Kisekka Market
A place of multiple interactions towards uncertain outcomes, of diverse activities and heterogeneous realities, Kisekka Market illustrates key features of the metastable city – all the more since it is also associated with Kampala's most unambiguous protests and riots (see Goodfellow 2013: 6-7). Up until 2014, when the market was demolished in order to be rebuilt, it is said (though this has not been confirmed) that major incidents of political violence were concentrated or spread from here. In 2007 and 2008, riots erupted over the lease of the Kisekka Market land to a retired army colonel (see Goodfellow and Titeca 2012: 267). In September 2009, the Kayunga riots that spread throughout Kampala seemed to have sprung from Kisekka Market (Baral and Brisset-Foucault 2009; Branch and Mampilly 2015: 122; Golooba-Mutebi 2011: 10-11; Human Rights Watch 2010a), and since the massive 2011 Walk-to-Work protests (see Branch and Mampilly 2015), opposition politicians have frequently aroused militant support and sparked anti-government protests when appearing in the vicinity of the market. 4 It is in that context that Kisekka Market had come to be known as, and stigmatised as, the “hotbed of rioters and demonstrations in Kampala's central business centre” (Mukisa 2014) – although, to reiterate, these riots were only a tiny fraction of what was going on at Kisekka Market, only a snippet of its diverse incidents, coincidences, and potentials. 5
On the riots in Kampala and the institutionalisation of “noise” in Uganda's urban politics, see Goodfellow (2013).
Anna Baral's insights from her ethnographic fieldwork at Kisekka Market have been helpful in this regard, all the more as she problematises rather than retraces the emergence of the market's stigma as a hotbed of rioters.
Kisekka Market's metastable character also bespeaks the ways in which the past routinely resurfaces in the present. Nowadays centrally located in downtown Kampala, the area of Kisekka Market used to be a swamp in the Kingdom of Buganda. As all land in Buganda, it was the king, the kabaka, who owned the land; still today, he is widely considered the Baganda's landlord, an issue that we shall return to below. The market emerged in the late 1980s as an informal trading spot for spare car parts, many of them stolen, exchanged, and sold through opaque networks associated with Uganda's magendo underground economy (see Prunier 1983). In the early 1990s, 26 row houses were built to accommodate the ever-increasing workforce, estimated to be “at least 10,000 people” in 2014 (Mukisa 2014), in a narrow space between Nakivubo Channel and Nakivubo Road. Since this space was inaccessible by car, the trades at Kisekka Market relied on middlemen. Hundreds, some say thousands, of brokers would intercept customers arriving by car at Nakivubo Road. They would negotiate a price and get the demanded spare part from inside the market, keeping the difference between the price they charged the customer and what they paid the shop owner. These could be substantial sums, given the high value of the traded goods and the ambiguity of prices. Customers, too, could profit from good deals if they had the right connections and a working knowledge of how Kisekka Market functioned. Much depended on being in the right networks, on being at the right place at the right time, and on being cunning enough to tilt the trades to one's advantage. Such an aura of potentiality was not restricted to the hawkers, brokers, shopkeepers, dealers, mechanics, flows of customers, and hundreds of food vendors: many youth from the city's outskirts would roam the market in search of random possibilities, action, and income, and interlocutors repeatedly emphasised that Kisekka Market attracted Kampala's sharpest and toughest hustlers, widely known as bayaye (see Frankland 2007), looking not only for individual deals but for projects, ideas, and formations that could yield something. 6 As one informant vaguely put it, “I won't be paying anything, and I will be making something. […] So when something just comes up, you're already tuned.” The second “something” refers here to riots and protests.
Interviews with Julius, 5 April 2014, Wabigalo, Kampala; Marc, 9 March 2014, Kazo, Kampala. The same goes for other markets and similar public spaces (see Frankland 2007). The names of all of Joschka Philipps's interviewees and informants have been anonymised.
The riots usually responded to previous political events in Kampala: arrests of opposition politicians such as Kizza Besigye and Erias Lukwago, election results that were suspected or proven to be fraudulent, controversial legislative decisions and executive measures (see Goodfellow 2014). Though each instance certainly featured specificities, market vendors described a recurrent pattern of rioting over the past years: a small group would block Nakivubo Road, chant slogans, and sometimes ignite car tires, growing into a sizeable crowd that anticipated the arrival of police. Confrontations ensued between stone-throwing and sometimes looting protesters and anti-riot police using teargas, rubber bullets, batons, and water cannons. Such stand-offs could last several hours, as Kisekka Market's dense architecture made it difficult for police to clear the space of protesters. With an audience of journalists filming the scene and onlookers from the surrounding “arcades” – multilevel buildings with additional shops that feature open facades and balconies – the confrontation between protesters and police was embedded in observation. Since the 2009 Buganda riots, which spread throughout the capital city and to other urban areas in central Uganda (Branch and Mampilly 2015: 122), there was an anticipated potentiality that the upheavals could proliferate beyond the confines of Kisekka Market, both spatially and in the sense of building political momentum as a media phenomenon, through newspapers, television, and radio shows, through the radio trottoir (Ellis 1989), and through the Internet. During Uganda's Walk-to-Work protests in 2011, for instance, online video snippets of the police's violent arrest of opposition leader Kizza Besigye became an important factor for the movement gaining further momentum through an increasing focus on police brutality and the right to protest (see Baker 2015; Kagoro 2015a; Goodfellow 2013, 2014). 7 Layer by layer, to draw on the crystallisation metaphor, the scope and significance of the event could thus evolve and change.
The incident happened not at Kisekka Market, however, but at the Mulago Hospital roundabout.
Despite the recurrent patterns of rioting, their origins usually remained opaque. Indeed, amongst all interviewees and informants for this research – including opposition activists, police spies and officials, dealers at Kisekka Market, and investigative journalists – consensus was restricted to the claim that the opposition had regularly given out money to spark riots, and that the government had paid money to impede them. Some argued that a clique in Kisekka Market's Block A was in touch with the opposition; others hinted that the Bangawa group, a Baganda youth organisation based in and around Kisekka, were behind the large majority of riots happening at Kisekka. Most said the middlemen started the riots, but many others argued that rioters were mobilised from elsewhere and had no relation to Kisekka Market. Several interviewees argued that the driving force behind the riots was Ganda ethnicity, while others emphasised urban youth unemployment and outrage against government impunity. Throughout the research, several of these claims could be partly verified, while none of them could be decisively refuted. Ambiguity persisted, and started at the individual level of who is who, and who is for whom at Kisekka. According to an anonymous police officer, there were between 500 and 1,000 police spies among the 10,000 workers at Kisekka Market, and this was widely acknowledged by the dealers and brokers themselves. 8 “Government spies are there,” one shopkeeper at Kisekka said. He went on:
Interview with anonymous police officer, 22 March 2014, Kampala.
The people who came to work focused on getting money are there, everybody is there, [but] we don't see – you don't see where is which one or what”. 9
Interview with Mohamed, 29 March 2014, Kisekka Market, Kampala.
Relationships between shopkeepers and middlemen also oscillated between trust and mistrust. “They are all about survival,” one shopkeeper argued of the middlemen, “they work for you but they can also destroy you,” citing examples where looting and rioting middlemen had stolen from shopkeepers they had just dealt with hours before. 10 In short, appearances could rarely be taken at face value, and the crucial concern in interactions was not necessarily what was real, but what could crystallise into some beneficial reality to pursue.
Interview with Ian and Daniel, 30 March 2014, Kisekka Market, Kampala.
Ambiguity also pertained to the one thing that seemed beyond doubt at Kisekka Market: allegiance to the kabaka. Though almost every interviewee at Kisekka Market was quick to emphasise that “love for the kabaka is strongest here in Kisekka,” that “you can be killed for saying anything even slightly degrading about the king here” because “we see him as superior in all aspects,” 11 the kabaka also remained, as an informant would remark, “a mystery person.” 12 Gerald, the informant, said most youth knew little about the king since he did not talk much in public (see also Brisset-Foucault 2013b: 6). “Whatever we, the youth, know about the king, and whatever we believe in, stays what we believe in,” he said. Since the previous kabaka, Mutesa II, was exiled under the Obote regime, and the current kabaka, Mutebi II, was hindered by police from entering “his” territory of Kayunga District in 2009, the kabaka, to many, epitomised the oppressed dignity of the Kingdom of Buganda, which was continuously put down and ridiculed by a corrupt central government (on the same narrative, see Mutibwa 1992, 2008). Love for the kabaka, then, could be more “political” than “ethnic,” more concerned with marginalisation and government corruption than with Ganda culture, and a vehicle of expression for multiple concerns, emotions, and positions. 13 This was evidenced during the 2009 Kayunga riots, which followed the police's interception of the kabaka at Kayunga District: surprisingly, a substantial number of rioters, possibly even the majority, were non-Baganda. 14 Side by side with Baganda youth, they would chant slogans, loot, and attack military forces while some of their peers stopped and beat people who could not “walk like a Muganda” – meaning those who could not identify their lineage within the Baganda clan system. In the midst of highly heterogeneous and contradictory concerns and energies, along with a carnival of disparate forces, ethnicity emerged as an overarching conceptual frame to designate what is largely understood today as an ethnic riot, in which “the Baganda took to the streets of Kampala to protest the police action” (Baker 2015: 378; emphasis added).
Interviews with Saadiq, 22–24 March 2014, Kisekka Market, Kampala.
Interview with Gerald, 22 August 2013, Makindye, Kampala.
For a strikingly similar case in Guinea-Conakry, see Philipps (2013a, 2013b).
According to numerous interviews with police, less than half of the arrested suspects were Baganda. For instance, conversation with Moses Kafeero, who, at the time, was regional police commander of the Kampala Metropolitan Police, 16 December 2015, Kampala.
In a context where such highly contradictory dynamics can coalesce into such consistent representations, Simondon urges analysts not to sever the event from its representation, but to understand the event as a seed, around which further layers of representation crystallise. Since these layers may alter the form and significance of the initial event quite considerably, they are often deliberately crafted to pursue specific political agendas (see Philipps 2016). The NRM, for instance, benefitted tremendously from depicting the riots as an instance of “ethnic violence” to undermine Ganda opposition vis-à-vis the central government. As Goodfellow (2014: 13) notes, the government
lost no time in capitalizing on [the violence]. Officials declared that the riots had been planned by the Kingdom's leaders, and began a clampdown on public space, arresting journalists accused of inciting the violence and closing CBS radio [Buganda's main media institution], which was taken off air for a full year. 15
See also Baral and Brisset-Foucault (2009), Brisset-Foucault (2013a).
This leads us to outlining the politics of crystallisation in Kampala.
The Politics of Crystallisation
Though the concept is new, politics of crystallisation have been widely discussed in terms of framing, brokerage, boundary activation, and polarisation, most explicitly in McAdam, Tarrow, and Tilly's (2001) works on “dynamics of contention” (see also McAdam and Tarrow 2011; Snow et al. 1986; Tilly 2003, 2004). Even if the crystallisation approach is situated in quite a different theoretical tradition, it analyses similar phenomena – for instance, how political entrepreneurs craft political groups and representations, and how these groups and representations serve as reference points in periods of contention. One key difference is that the crystallisation approach focuses specifically on how these contours emerge from a context of political opacity (Vigh 2015). It describes a process whereby a highly fluid environment becomes solidified into something concrete, whereby ambiguity is transformed into tangible realities. To Simondon (1958, 1989, 1995), this “individuation” process is universal: anything that is perceived as a concrete entity – a group, idea, or institution – is in fact the product of intangible and diverse relations. These intangible relations, in turn, only become perceptible through the things they produce. A plant, for instance, emerges from a relational process between sunlight and minerals in the ground. The relation between the sun and the minerals is usually intangible, but emerging from a seed, the plant absorbs these intangible relations and transforms them into a concrete form. An “individuated” entity thus “absorbs into itself the relation that gave rise to it, thus obscuring it” (Combes 2013: 16). Simondon therefore challenges us to approach anything – any event, phenomenon, or artefact – with the question of how it emerges from relations. It is in this sense that Simondon provides an analytical method as much as a theory, in which individuation constitutes a universal process of becoming that can be retraced from a metastable milieu, through an induced seed, to an indeterminately crystallising phenomenon. Thus, as much as Kampala's urban protests may seem particularly amenable to Simondon's analytical framing, it needs to be emphasised that the crystallisation metaphor originally relates to a much broader philosophical claim and easily applies to quite different phenomena as well. The process of things going viral on the Internet, for instance, is another important case in point (see Nahon and Hemsley 2013). 16
For the overall theoretical argument on social movements, protests, and riots in African Studies, see Philipps (2016).
The crystallisation approach advocated in this paper highlights the contingency and the unpredictability of such processes. Simondon's analogy to the natural sciences notwithstanding, the approach also emphasises the importance of human agency (see also Emirbayer and Goldberg 2005; Emirbayer and Goodwin 1994; Emirbayer and Mische 1998). 17 In short, in a political context where it is particularly uncertain how things will play out, political actors will be all the more intent to strategically make things crystallise in their favour. To describe these politics of crystallisation, we focus on what Charles Tilly (2003) calls the “political entrepreneur”: a “transversal figure” (Wilson 2009: 16) who tries to carve out the contours of an initially fluid political field to gain political capital. As Tilly writes with regard to the conflict in former Yugoslavia:
In this regard and many others, the crystallisation approach decidedly runs counter to Robert Kaplan's The Coming Anarchy (1994). Though Kaplan uses a similar sort of “crystallisation” imagery (he likens West African young men to “loose molecules in a very unstable social fluid, a fluid that was clearly on the verge of igniting”), his perspective discards human agency as much as the open-endedness of history. Mainly intent on creating an apocalyptic image of the global future, Kaplan abuses “the experience of a number of African countries [as] merely a dress rehearsal for an ecological fate towards which humanity is ineluctably moving” (Mkandawire 2002: 183).
[Political entrepreneurs] specialize in activating (and sometimes deactivating) boundaries, stories, and relations, as when Bosnian Serb leaders sharpened boundaries between Serbs and their Muslim or Croatian neighbors with whom Bosnians of Serbian lineage had long mingled, married, traded, and collaborated. They specialize in connecting (and sometimes disconnecting) distinct groups and networks, as when those same leaders integrated armed Serbian gangs into larger nationalist coalitions. They specialize in coordination, as when those leaders organized joint action on the part of those coalitions. Political entrepreneurs specialize, finally, in representation, as when Bosnian Serb leaders claimed to speak for all Bosnians of Serbian lineage […]. In these ways, political entrepreneurs wield significant influence over the presence, absence, form, loci, and intensity of collective violence. (Tilly 2003: 34)
Sam, who owned a shop at Kisekka Market, was a paradigmatic political entrepreneur from the opposition who tried to turn fluid dynamics into tangible events and political camps. 18 An ambitious young politician who stood for a parliamentary seat, he had built diverse networks within three different opposition parties of which he was a member. Much of his political clout came from commanding a so-called youth brigade with a branch in Kisekka Market that could spread rumors, execute protests, blackmail political rivals, or confront police in favour of any politician who “gets a problem” and is willing to pay for its solution. He recounted stories of how he organised 30 of his Kisekka Market members into “command,” “information,” and “implementation” sections during the Kayunga riots, and mobilised urban Baganda youth by emphasising that “the government doesn't like your king!” Another time, they staged a violent attack on an opposition politician, to then publicly allege that the perpetrators were government thugs. Once, they destroyed a public toilet to tell bystanders and residents that the government was going to tear down the informal settlements and build new houses in the area, sparking protests against the Kampala Capital City Authority (KCCA). Once the information was out, it crystallised into a phenomenon, and had already gathered momentum by the time the KCCA dismissed the allegations and rumours as false and fabricated.
Interview with Sam, 15 October 2012, Makerere, Kampala.
Media coverage was central in these schemes; the goal was to “throw information at the public eye,” 19 possibly even vis-à-vis international audiences and donors that both the regime and civil society heavily depended on (see Tangri and Mwenda 2006, 2010; Tripp 2010; Kagoro 2015a). 20 Both police and activists therefore informed and paid journalists to report about the spectacle from a specific angle. 21 A former opposition activist elaborated,
Interview with Sam, 15 October 2012, Makerere, Kampala.
Interviews with various informants, e.g. Sam, 15 October 2012, Makerere, Kampala; Dixon and Ron, 17 October 2012, Makerere, Kampala; Gerald, 4 October 2012, downtown Kampala.
Conversation with Gerald, 4 October 2012, Kampala.
And, the good thing, how all this [protest mobilisation] is achieved, is media. Cause we buy media. Guys buy the media, […] and the media reach the scene, very early enough and keep waiting. Then they start filming. Police will start beating the [journalists], hé? Not to cover the event. They are calling the radio station. “Ahhh, police are running, roughing up, they are taking my camera,” directly reporting [live], you get? So those on radio also tell others […]. By evening, the whole city is on fire. 22
Interview with Dixon and Ron, 17 October 2012, Makerere, Kampala.
External observation and media coverage was crucial in particular because of the observers' partial knowledge of what was taking place – partial in terms of both its incompleteness and the political biases that would guide the observers' interpretations of what they saw. Police and protesters actively appealed to that partiality: protesters would place women activists close to cameras to visualise their vulnerability; 23 in front of other cameras, police commanders would explain the need to protect law-abiding citizens from irresponsible rabble-rousers. Journalists gravitated towards these different versions as avenues for advancement. “Everything is politicised,” said Daily Monitor journalist John Njoroge:
Conversation with Gerald, 4 October 2012, Kampala.
You cannot get by, in Uganda, just by simply being the average person. If you want to get ahead, you must affiliate yourself with some form of political persuasion. 24
Interview with John Njoroge, 22 October 2012, Kampala.
These affiliations were neither long-term nor clear-cut, however, as many journalists took “sides” as those sides emerged, or, more precisely, as the journalists helped those sides emerge. After all, opposition and police emerged as two “sides” to a considerable degree through journalistic observation, through being seen and talked about as bounded groups from the outside. Just like during the 2009 Kayunga riots, where Ganda ethnicity was an attribute ascribed to the rioters rather than one the rioters experienced, there are good reasons to believe that the groupness of government and opposition forces is at least as much a matter of outside attribution as of internal experience.
Political entrepreneurs, who actively constructed what emerged as political realities, would accord rather little credibility to these constructions themselves. Even when asked about his own youth brigade, Sam, the above-mentioned political entrepreneur, said dryly: “We use these youths. You know, we want power, we want to get legislative seats.” Many political entrepreneurs had worked for both the opposition and the NRM; some had switched sides multiple times during their career and some were working for both simultaneously. Although senior politicians shifted political sides as well – ex-PM Amama Mbabazi, who ran for president against Museveni in 2016, is a noteworthy case in point – such dynamics were magnified amongst a large population of young urban graduates who aspired to a successful career. As most opportunities for professional and economic advancement depended on connections, and since the important nodal points were situated within the NRM government, which is close to synonymous with the state, many youth were almost inevitably drawn to these political networks and the enormous sums of money that circulated through them. 25 Adam, a former youth mobiliser and protester for the Black Monday Movement in Kampala, for instance, had first registered with the NRM in 2006. “I realised that I cannot get anything if I am not with [the] NRM,” he said, and eventually worked his way up to becoming the NRM Youth League chairman in Kampala's Makindye District. He left the NRM in 2011, disappointed in its system of ubiquitous nepotism in which he, as a fatherless young man with no family ties, was unable to advance. He recalled,
According to Andrew Mwenda (2011), the NRM spent an equivalent of USD 350 million on the 2011 election campaign, which would amount to a per capita spending of approximately USD 10, dwarfing Barack Obama's former record spending of USD 2.50 per American citizen in 2008.
People were wondering: Why are you running from where money is? And where are you going to go? And I was disturbed, really, [because] I never trusted in any opposition party.
He then coordinated the Uganda Youth Platform, an opposition movement by General Sejusa, the former coordinator of the intelligence services under Museveni.
Cross-cutting political boundaries was so entrenched that political entrepreneurs could coordinate and connive across the lines they would draw. Dixon, for instance, a police spy at Kisekka Market working for the inspector general of police, usually sought to prevent upheavals by channelling money to opposition groups who would otherwise organise riots. 26 But when “things go sour at the end of the month,” he also collaborated with them:
Interview with Dixon and Ron, 17 October 2012, Makerere, Kampala.
I'm broke, so I organize my gang […]. They tell him [the inspector general of police] all sorts of lies. That maybe the opposition today approached us; they're saying tomorrow we should go for this demonstration. So, chief will call me and say, “Dixon, you have to go to Kisekka and, you know, handle this situation.” And eventually he will give me some small money, I will also get my commission, I go and (laughing) meet my gangstas and give them something and we survive. Yeah, this is (pauses), basically, this is how our government has survived. For all these years! 27
For a similar point on urban patronage politics in Kampala, see Goodfellow and Titeca (2012).
In sum, uncertainty and ambiguity constituted a “productive resource” for political entrepreneurs (Cooper and Pratten 2015: 3), even vis-à-vis their bosses, as it allowed them to invent schemes and create realities, which they hoped would gather momentum. Such momentum depended on how different forces would react to a specific “seed” – for example, an event of public uproar – and political entrepreneurs prided themselves on knowing that very well. “I know youth politics, hé?” Ron said, having worked for both the opposition and the government:
I know violence, how it comes up, hé? I know how opposition operates, and I know how government operates, properly! […] In this nation I know.
Such knowledge was relative to a surrounding that political entrepreneurs tended to picture as comparatively naïve. Gerald, less self-confident than Ron, said,
There are some things I understand. But I want to tell you that the youth who follow me, know absolutely little about so many things. About so many things. So, this person, if I decide to put something in him [an idea or a political affiliation], me, myself, I can.
In a context of manipulable surroundings, then, the scope of schemes depended on how well connected a political entrepreneur was. 28 Those able to draw from a broad range of connections could induce the most wide-ranging crystallisation processes, while others had to pitch their ideas to better-placed peers to benefit from their connections. For instance, during the interview with Dixon, the police spy, his phones rang and vibrated incessantly. “You see my phones here?” he asked at some point. “I receive sooo many calls every day, these gangsters, whoever calling me […] [I receive] soooo many ideas every day!” Asked to specify the kind of ideas, Dixon answered, “Blackmail” – threatening someone's power by exposing subversive information. While the content of the information itself was comparatively negligible and oftentimes fabricated, the threat consisted of the expected effect – what would happen if the information gathered momentum.
On “connectionwork” in Kampala's music industry, see Schneidermann (2014).
The NRM state was in a strikingly similar position: blackmailed on charges of corruption, which is common knowledge across the country, the subversive element was not the information that corruption existed but that frustrations about it gathered political momentum in large-scale protest. The question preoccupying Dixon was how long the government could pay the ransom:
The moment the government runs out of money, we get [it gets] scared, trust me: (whispering suspiciously) “Now, these people, criticising on the street, what shall we give them?” If there is no money?
Indeed, it is in this context that the state has increasingly counted on police to handle the issue.
Policing the Metastable City
Uganda's police had long remained in the shadows of Museveni's National Resistance Army (NRA). While the army embodied the Ugandan liberation after the Bush War from 1981 to 1986, the police, adopted from the former Obote regime, were poorly equipped, notoriously underfunded, and unheeded in public. The fact that it is now a highly visible, well-equipped, and equally prominent and controversial institution in Ugandan politics that no longer requires military assistance in anti-riot operations has a lot to do with the current inspector general of police (IGP), Kale Kayihura. Kayihura had joined Museveni in the Bush War in 1982 after completing his LLM at the London School of Economics, and was then appointed to lead sensitive operations in Congo and to conduct internal investigations. Kayihura became IGP in 2005, and the police's manpower has increased enormously since then, from 8,000 in 2001 to an estimated 60,000 today (Kagoro 2014: 120). From 2004 to 2014, the police budget grew from UGX 75 billion to 303 billion (i.e. from USD 22 million to 90 million), 29 the number of vehicles has mushroomed, 30 and for the 2016 elections period alone, police secured UGX 212 billion (USD 63 million) for mobilising, recruiting, training, and equipment purchases. 31 Recently, the police established a new anti-riot division, the so-called Field Force Unit (see Baker 2015: 381–382), and in 2013 President Museveni signed the Public Order Management Bill, which gives police discretionary authority to break up gatherings of as few as three people in a public arena who are deliberating political issues (Kagoro 2014: 114–115).
See Ministry of Internal Affairs, Ministerial Policy Statement FY 2013/2014. Presented to parliament for debate of estimates of revenue and expenditure by James Baba, minister of State for Internal Affairs, on 30 June 2013; 83–134.
The number of armored personnel carriers (APCs) has grown from 0 to 30; patrol pickups (trucks) from 125 to 591; motorcycles from 0 to 3,556; and saloon cars from 41 to 232. See Uganda Police Force Fleet Statistics FY 2003/2004–2013/14.
Kayihura also negotiated that police may retain all non-tax revenues such as traffic fines rather than remitting them to the Uganda Revenue Authority (URA) (see Kagoro 2014: 120).
Despite these legal-institutional instruments, the police tend to operate not as a bounded institution but an open-ended network, defying distinctions as to who is in the police and who is not. Street kids indicate traffic offenders to policemen to later pocket a fraction of the bribe (“chai”). Paramilitary youth brigades join police in ad-hoc anti-riot operations on the basis of personal and financial arrangements whose specifics remain opaque to most of those involved. 32 At police stations, informal middlemen, so-called kayungerizi, liaise constantly between police officers, suspects, and complainants in nebulous shuttle mediations about allegations, bribes, and brokerages. Akin to Kisekka Market, such jockeying involves a multiplicity of connections, phone calls to people of potential influence, some of them fake, some of them real, most of them transgressing institutional boundaries. Police actively engender such entanglements, and have amassed a plethora of informants, to the point where Kampala's urbanites half-jokingly say that “if four people meet, you can be sure that one is a spy.” 33 Just like during the Ugandan Bush War, when civilians were military-trained under “the democratisation of the gun” discourse to later participate in the removal of the Obote regime (Mudoola 1991: 239; Museveni 2000: 80), and akin to the post-war nationwide mchaka-mchaka politico-military trainings of civilians that interlaced Ugandan society with a marked military ethos (Kagoro 2015b: 183), current “crime preventer” trainings are turning millions of Ugandan citizens into police partners, including for political mobilisation purposes on behalf of the ruling party. 34
Interview with journalist Angelo Izama, 18 October 2012, Kisementi, Kampala.
E.g. informal conversation with Adam at Kisekka Market, Kampala, 5 October 2012.
For more on crime preventers, see Moses Khiza, “Here is the absurdity of crime preventers.” In: The Observer, Kampala, 28 August 2015, online: www.observer.ug/viewpoint/39546-here-is-the-absurdity-of-crime-preventers (3 June 2016). See also “Crime preventers demand payment for Museveni Kololo rally, anti-Mbabazi demo.” In: The Observer, Kampala, 6 November 2015, online: www.observer.ug/news-headlines/40873-crime-preventers-demand-payment-for-museveni-kololo-rally-anti-mbabazi-demo (3 June 2016).
One of the police's greatest assets in the spy apparatus are boda-boda motorcycle taxi drivers. Estimated at 200,000 and continuously growing in number, 35 they are critical for Kampala's traffic infrastructure and were long thought of as “raw material for the opposition.” 36 They now constitute a main source of intelligence information for police, especially, in the IGP's words, “to fight the opposition rioters who only want to destabilise the city.” 37 The boda-boda drivers' political affiliation shifted notably after the Kayunga riots, in which they featured prominently as protesters and mobile linkages between riot settings. Soon after the riots, IGP Kale Kayihura installed and co-opted the Boda-Boda 2010 Association, organised the motorcycle trade with a company in India that supplies motorcycles at a cheap price, and became the middleman between the boda-boda drivers and President Museveni. In 2010 Kampala's boda-boda leadership, headed by Abdallah Kitaka, was flown on the presidential jet to Arua, in northwestern Uganda, to meet Museveni, who later handed out more than 500 motorcycles for free and thousands more under favourable repayment plans. Today, innumerable boda-boda riders work closely with police, sometimes in arrests of suspected traffic offenders, but mostly in the field of intelligence. Hundreds have undergone crime prevention trainings, including stripping and shooting AK47 guns, martial arts, and ideology classes. Boda-Boda 2010 Association offices are often painted yellow, the color of the ruling party, and feature pictures of President Museveni, IGP Kale Kayihura, and the police's former director of operations, Andrew Felix Kaweesi. 38 Police are always invited to oversee elections of boda-boda leaders.
These figures were revealed during a meeting between the RPC Kawempe and all Kampala city division boda-boda chairmen held at Makerere, Kikoni on 26 March 2015.
Interview with a senior police officer at the rank of assistant inspector general of police (AIGP) and with a regional police commander, both in Kampala, on 18 March 2015 and on 26 March 2015, respectively.
Conversation with the IGP Kale Kayihura, 18 April 2015, Kololo, Kampala.
Kaweesi is now the director of Human Resource Development.
Given these overall developments, the crystallisation of police may easily be misunderstood as an institutionalisation towards greater stability, when in fact police, too, are inherently metastable – representing a fluid terrain whose inside and outside are constantly negotiated and blurred, where positions and alliances can shift overnight and turn into rivalries, contingent on myriad informal relations whose specifics remain opaque, and where money and spies, rather than securing control, amplify the agitation. For instance, senior officers recalled an incident where the director of the Kampala Metropolitan Police went to meet with the National Forum of Crime Preventers, expecting to be received with due honour, when the crime preventers told him squarely,
It is a good thing you have come, you did a good thing, because all those that did not come here to see us did not succeed, they were stripped of their jobs. Without us you can't work in this city and we can cause your sacking. 39
Interview with a senior police officer on 20 November 2015, Kampala.
As another senior officer intimated with respect to boda-boda riders,
In 2011 after the presidential elections […] I mobilised over 20,000 boda-boda riders, gave them fuel [worth] 20,000 [shillings] each so that they can ride ahead of Mzee Museveni's convoy to Kololo for the swearing-in ceremony. This of course would make it seem that the youth have been overwhelmingly and with enthusiasm behind the president. It would also make the function more colourful. But the boda-boda cannot be trusted 100 per cent; they are slippery characters. After escorting Mzee, they immediately turned their bikes to go and also pick [up] Besigye [the arch-rival of Museveni] from Entebbe Airport [Besigye was returning from Kenya where he had gone for treatment after the police had pepper-sprayed his face during a political rally], saying, “Oyo tumutusiza kati katu kime omulwadde.” [“Since we have safely escorted this one let us go and collect the one who is sick.”] 40
Interview with a senior police officer at the rank of AIGP, 12 March 2015 at Bukoto, Kampala.
Added to the “slippery character” of the police's supposed allies, subversion may also come from within the police forces. A substantial minority of police personnel self-reported feeling “hostile to the sitting government,” and disunity about the politicisation and militarisation of police exists amongst the highest echelons. 41 Competition over better-paid posts, commonly referred as “wet” deployments, leads to diverse conflicts and rivalries. Influential officials are frequently blackmailed by their peers, and disgruntled Field Force Unit officers have even undermined anti-riot interventions by shooting tear gas canisters when the order had been not to, sometimes at schools or hospitals, to make their superiors lose their position. 42 In that regard, protests not only emerge as an external threat, but also trigger uncertainty within the police as to what will crystallise out of their own ranks.
Interview with anonymous police officer, 22 March 2014 in Kampala.
Interview with the AIGP held on 15 March 2015 in Wakiso.
At the same time, while the police's own ranks can pose risks, the opposition can be won over. The IGP himself is known to frequently take rioting youth to his personal home in Muyenga to dissuade them from organising and participating in riots – offering tea and food, handing out money, promising development projects to individual ringleaders, integrating others into the spy apparatus, keeping them tied up in talks when they are supposed to be mobilising, and appealing to their hopes that they can make it within the current metastable order, and that they need to continue trying like everyone else. Ron, a former opposition mobiliser, when asked why he would stay with the IGP instead of mobilising fellow protesters, exclaimed,
This is the general who has called you, everything is there. He's telling you how he used to do what you're doing. […] Personal experience. [inaudible] That these [opposition] people are not going to help you, they are doing selfish businesses, you're young, you lose your life.
Ron later worked for the IGP himself: “He's a nice person, this guy. He listens to you. If you have personal challenges, you seek something, he gives you money.” 43 The IGP's personal budget per quarter for such ad-hoc purposes is now at UGX 3.6 billion (approximately USD 1 million), making him a key player in the urban politics of crystallisation. 44
Interview with Ron, Kampala, 16 March 2014.
Personal conversations with senior officers.
Money, however, was only one way of silencing dissent. While police pay journalists who report favourably about their operations, 45 they muzzle others whose reports may threaten the regime (Human Rights Watch 2010b). In the wake of the 2009 Kayunga riots and “in addition to repressing the demonstrations in the street, the state shut down four radio stations” and banned the so-called ebimeeza – open radio debates that had been key sites of political deliberation (see Brisset-Foucault 2013a: 241). Numerous newspaper offices and radio stations have been raided and shut down over the years, and various journalists have been arrested, leading to considerable insecurity and precariousness amongst journalists without political protection (see Tripp 2010: 96-101). At the same time, many journalists seek such protection from police. Police headquarters are crowded with journalists who hope to establish good relations with influential officers, to have a number to call in difficult situations, but also to get the latest news, a fuel card, or UGX 100,000 for positive coverage. This goes for journalists from both private and state-owned media, as Uganda's media outlets tend to defy the idea of stable political camps. Journalists from the state-owned New Vision newspaper, for instance, might be critical of police, while the previously police-raided Independent news magazine might be full of praise, due to a diverse range of favours, threats, personal rivalries, and other contingencies.
Ethnographic observations by Jude Kagoro.
An important upshot from this discussion on police is that any crystallising entity in a metastable context must maintain responsiveness to change and uncertainty. Just like the Boda-Boda 2010 Association or different media outlets, police were forced to become neither too stable, bounded, or paralysed in an ever-shifting environment, nor too unstable, fluid, or indistinguishable from that environment. While ensuring visibility as a distinct institution, police also connived with boda-boda drivers, some of whom had been involved in anti-government protests, some of whom still were, and some of whom sympathised with the protests, but would argue that “the difference between the government and the opposition is that the opposition doesn't have power. That's all.” 46
Interview with Lawrence, boda-boda driver, 22 February 2014, Kololo, Kampala.
Conclusion
Thinking through these dynamics of protest and policing in Kampala in terms of stable political camps, where a bounded opposition stands against a bounded government or state apparatus, ignores both the relations that transgress these boundaries and the shifting political positions between these camps. Though political camps do obviously exist in Kampala, this paper has suggested treating them not as structures, not as a priori, permanent, and definitive parameters of agency, but as emerging fields of gravity once a protest, riot, or any other “seed” transforms the metastable milieu into an antipodal political spectrum. This process is akin to what Brubaker (2002) has argued about ethnicity and the formation of ethnic groups. Brubaker, who notably defines ethnicity as a “crystallization of group feeling” (2002: 167), rejects the concept of “ethnic groups” because it presupposes the “groupness,” whose genesis the analyst is actually supposed to account for (2002: 176). Since ethnicity is contextually fluctuating, waxing and waning over time, and oftentimes a corollary of conflict rather than the underlying cause of it, it should be thought of as a project or an event rather than a collective entity: schemed by political entrepreneurs, solidified through external representation by media, and self-perpetuating as it crystallises (Brubaker 2002: 168).
In this paper, we applied a similar idea to opposition and government forces in Kampala, outlining how political entrepreneurs construe situations through which opposition and government forces emerge as bounded entities that people gravitate towards. Since such gravitation processes are essential for party politics to work, the incentives to incite them are by no means restricted to the Ugandan opposition. The NRM, too, capitalises profoundly on events that polarise and make the opposition visible as a threat to the regime – especially when these events are violent. Indeed, the NRM has long benefitted from security threats to expand its military apparatus and present itself as the only alternative to violence and mayhem (Baker 2015; Branch and Mampilly 2015). As Goodfellow (2014) has shown in a variety of cases, NRM politicians have in recent years strategically stirred up legislative debates whose main effects were political unrest and state crackdown. Rather than trying to implement the debated laws – which appeared to be a secondary concern at best – the NRM used the violence they provoked as a motive to criminalise the opposition, demonstrate the regime's dominance, and justify further “legal maneuvers” that restricted political space and “did little to ease underlying tensions” (Goodfellow 2014: 2). 47
Designating the Ugandan state as “politically stable” or “politically unstable” becomes problematic in this context, not only because both are simultaneously true (cf. Tripp 2010). Rather, it is problematic because such designations tend to absorb the very processes that they refer to, the whole gamut of politicking that goes into crafting the “stable” or the “unstable” state, the actual making of politics as they appear. To account for these processes, Simondon's framework provides a rich set of instruments that can be applied to matters far beyond the question of urban politics in Kampala. One is Simondon's idea of the pre-individual, metastable realm. Put simply, one could think of the metastable realm as a truly empirical world that is inconceivable because it is not yet conceptualised, not yet divided into distinct phenomena that one could describe, assess, and evaluate. Additionally, this idea implies that the emergence of any conceivable reality from the metastable realm is contingent on the different interests of those who partake in defining its contours, and thus inevitably controversial, as we elaborated in the “politics of crystallisation” section. Related to that, Simondon's framework hints at the need for a self-reflexive understanding of how our analytical concepts also work as agents of crystallisation (Brubaker and Cooper 2000; Combes 2013; Lakoff and Johnson 2003). A concept like “police,” as we have shown in the final section, bundles a diversity of oftentimes incongruous elements into something homogeneous and perceptible, for better or for worse, and scholars face the complex challenge of finding methods and languages that account for what gets lost in that process. AbdouMaliq Simone's work on urbanity remains noteworthy in this regard, as he refrains from fixing and determining what he describes, and instead highlights and performs interdependences and elusiveness. As Simone said in a recent interview, “I guess what has always interested me about urban life is that aspect which it refuses to be folded into” (Philipps 2015: 58).
In a context of allegedly ubiquitous security threats, such a political strategy is clearly of global concern.
The crystallisation approach, finally, aside from lending itself to investigating the metastable city as experienced by its residents (Simone 2014: 4), or to scrutinising protest movements (Philipps 2016), encourages analysts to reconstruct the crystallisation of realities about African politics that emerge in a rather top-down fashion. Indices used in global governance, such as the Fragile States Index, 48 are a crucial case in point (see Merry, Davis, and Kingsbury 2015). Widely accepted and used as objective measurements, indices construe “easily digestible” realities (The Fund for Peace 2016) from exceedingly more complex national contexts, which these indices hierarchise according to standards mostly set in the Global North – with marked effects on governance (Cooley and Snyder 2015; Merry 2011; see also Jerven 2013, 2015). A Simondonian “politics of crystallisation” approach would problematise, but also retrace how such indices work. As regards the Fragile States Index, such an approach would, for instance, question the idea of political (in)stability as the aggregate sum of standardised indicators and, with that, the index's ability to perform an “early warning” function on that basis (The Fund for Peace 2016). The so-called “Arab Spring” has amply manifested how contentious politics can crystallise unpredictably, in unlikely settings and in unforeseeable ways, and how they can destabilise states across and beyond the MENA region, which the Fragile States Index and other “expert institutions” had ranked as comparatively stable at the time (The Fund for Peace 2011; see also Philipps 2016). Nevertheless, as indicators provide “a transition from ambiguity to certainty” (Merry 2011: S88), the crystallisation framework is well equipped to also retrace and retrieve the diverse processes and relations that indicators absorb, and to explore the functions of indices at the nexus of global politics and knowledge production. As such, Simondon's theoretical apparatus provides valuable tools for both classic empirical analyses and critical reflections on African politics as an object of knowledge.
The Fragile States Index, which until 2014 used to be called the Failed States Index, has been published annually since 2005 by the American think tank The Fund for Peace and the journal Foreign Policy. It is based on 12 social, economic, and political indicators, each with an average of 14 sub-indicators. Criticisms of the index abound. See, e.g., Beehner and Young (2012); Evers (2014); Leigh (2012).