Tensions,Confrontations,and Consensus: WhatsApp Use in Kenyan Electoral Politics

Abstract

Social media has enhanced culturally grounded debates and ethnicity-based exclusionism during Kenya’s WhatsApp group deliberations. These practices are often more pronounced during elections when WhatsApp becomes a social media platform of choice for various groups. Exclusionism takes at least two vectors. On the one hand, tensions and confrontations emerge as electoral deliberations proceed. On the other hand, consensus on WhatsApp spaces also emerges during the electoral deliberations as in-groups forge common ground. To analyze the splits in elections-based deliberations on WhatsApp, the article starts with an interrogation of how history and culture inform the Kenyan people’s use of WhatsApp space for electoral deliberations. Among others, the people use the platform to resolve perceived and real political injustices perpetrated in the society by previous governments, resulting in tensions and confrontations. The article argues that the origin of the tensions is the fact that most African people have dual “citizenship,” whereby they simultaneously belong to their ethnic communities while being part of the broader Kenyan nation. It theoretically engages Ekeh and Langmia to argue that the people concurrently fall into the primordial publics and civic publics, leading to dialectical tensions, confrontation, and consensus between the two categorizations. These tensions, confrontations, and consensuses spill over into the WhatsApp groups, with electoral deliberations registering the peak in these splits. The case of the Bungoma County and its linkage to the broader Kenyan nation serve as the site for empirical data. The method for data collection is qualitative. The article uses participant observation and face-to-face interviews to investigate how the people’s historical and cultural values inform their use of WhatsApp space during electoral deliberations. The findings reveal how the people’s historical and cultural values shape their use of WhatsApp space. The implication of the research is to explore the African people’s cultural adaptation of technology and to invite more Afrocentric theorization on technology adaptation in African societies.

Introduction

Ethnic divisions have marked the Kenyan political landscape since independence in 1963. However, the reintroduction of multiparty democracy in 1991 saw a meteoric rise in ethnicity-motivated tensions to a level hitherto unprecedented (Shilaho, 2008). These ethnicity-inspired political tensions, confrontations, and consensus are like what Ekeh (1975) labeled as dialectical tensions. They originate from perceived and real unresolved political injustices perpetrated on the people along ethnic lines by colonial and post-independence Kenyan governments. This article investigates how historical and ethno-cultural values shape Bungoma County people’s use of WhatsApp space for electoral deliberations.

This article conceptualizes ethno-cultural values as the indigenous ethnic logic that shapes people’s electoral deliberations. The objective is to establish how historical and ethno-cultural values inform people’s use of WhatsApp spaces during electoral deliberations. The main research question is: How do Bungoma people’s historical ethno-cultural values shape WhatsApp use during electoral deliberations? The subsidiary questions are: How does the majority or minority status inform people’s WhatsApp use during electoral deliberations? How does the reward system influence WhatsApp use during electoral deliberations? How does affect influence WhatsApp use during electoral deliberations?

The article is divided into four sections. The first section introduces some of the terms and concepts that will be leveraged to the study. This section also sets the context under which WhatsApp deliberations have become significant during electioneering. The second section analyses relevant literature, traces affective social media use to colonial-era media practices, and engages with theoretical debates on the Afrocentric public. The third section introduces our qualitative approach to analyze how affective historical and ethnicity-inspired cultural values shape WhatsApp use during electoral deliberations. The fourth section empirically shows how the affective values shape WhatsApp use in Bungoma County. We conclude by summarizing the findings and the implication of the research about Afrocentric theorization on technology adaptation in African societies.

As we shall expound in the theoretical framework section, the governments, consisting of individuals simultaneously belonging to both the primordial and civic publics, committed the injustice due to inherent dialectical tensions within them (Ekeh, 1975). We define primordial publics as those that belong to specific ethnic communities while civic publics are taken as those that bend toward national or nationalistic worldviews. The tensions and confrontations between them lead to exclusionism, resource deprivation, and loss of life. The WhatsApp space was first adopted in Kenyan political deliberations during the 2013 general elections, leading to the transposition of tensions, confrontations, and consensus onto this virtual space. This is loosely comparable to what Langmia (2016, p. 268) discusses as a “false online identity.” These dynamics are also like what Goffman (1959, p. 1) seminally idealized as a life performance involving “the presentation of self in everyday life.” In this article, the term majoritarian means having a majority in numbers, while minoritarian means minority in numbers. The terms shall also be used to analyze and discuss the majority or minority in WhatsApp group composition.

According to Kenya’s Digital Statistics (KDS, 2023), WhatsApp is the most used space in Kenya at 89.9% of the population. Its popularity is partly explained by the fact that it is easier to use and due to comparatively low-cost internet packages commonly known as data bundles (KDS, 2023) provided by mobile telecommunication service providers. Its use varies enormously. For instance, some use WhatsApp spaces to evaluate leadership performance in public offices (Omanga, 2019). Others perceive the space as safe for use in electoral deliberations compared to other social media platforms (Ooko, 2023). However, according to Kenya National Commission on Human Rights (KNCHR, 2017), WhatsApp also enables people to affectively exclude others based on ethnicity, among other social stratifications in society. Jane Arnold (2011, citing Arnold & Brown, 1999) defines affect as the “aspects of emotion, feeling, mood or attitude which condition behaviour.” The use of WhatsApp spaces in political discourse is primed with affect. In Kenya, most of the 47 administrative units, referred to as counties, are populated by ethnic majorities. The people in Bungoma County affectively exclude other ethnicities perceived as beneficiaries of the past governments’ political injustice. Ekeh (1975) discusses this as the rewarding primordial publics in exchange for political support. In this article, the affective politics concept means recycling emotions into electoral deliberations as new voices render old ones inaudible (Mbembe, 2017). The affective politics is shaped by Bungoma County people’s cultural and political history. Ekeh (1975) traces the origin of affect on African people to the co-existence of primordial and civic publics, which results in conflicts and confrontations. For example, the majority ethnic group in Bungoma County, the Bukusu, had numerous confrontations with neighboring ethnic groups and the colonial administration, over economic resources, political freedom (Wefwafwa, 2014), and identity.

According to Kenya Population and Housing Census (KPHC, 2019), the ethnic composition in Kenya in the 2019 census indicates that the top 10 ethnicities include Kikuyu at 17.2%, Luhya at 13.8%, Kalenjin at 12.9%, Luo at 10.5%, Kamba at 10.1%, Somali at 6.2%, Kisii at 5.7%, Mijikenda at 5.1%, Meru at 4.3%, Turkana at 2.6%, and Masai at 2.2%. The country has about 43 ethnic groups. As aforesaid, in Bungoma County, the Bukusu dialect of the larger Luhya ethnic community forms the majority of people. The second largest ethnic majority in the county is the Tachoni dialect from the same larger Luhya ethnicity. The third ethnic majority is the Sabaot dialect from the larger Kalenjin ethnicity. The fourth ethnic majority in the county is the Ateso-speaking people, who form less than 0.5% of the total population (KPHC, 2019).

Bungoma County is in the Western Kenya region neighboring eastern Uganda. It is one of the 47 counties in the country. The county is the fifth most populated in Kenya after Nairobi, Kiambu, Nakuru, and Kakamega (Government of Kenya, 2020). The present-day residents of Bungoma County take pride in embracing their indigenous ethnic and cultural values (Wefwafwa, 2014) while remaining a part of the larger nation of Kenya. This is loosely comparable to Ekeh’s (1975) discussion of the African people as simultaneously primordial and civic public. For instance, male circumcision, an ethno-cultural practice, is entrenched in the county and is often intertwined with the politics of ethnicity (Lamont, 2018). Arguably, this unique form of long-running traditional practices spills over into Bungoma people’s electoral deliberations, along ethnic lines. How historical and ethno-cultural values influence Bungoma peoples’ WhatsApp use during electoral deliberations forms a significant motivation for this research. Overall, there is a need to understand how people’s values shape their use of digital technologies in the context of competitive political dynamics.

The Kenyan political system is presidential, with elections taking place every 5 years. The most important period of the electoral cycle is usually 3 months before the polling day and immediately after polls, before the winner is sworn in. The period involves political party nomination of the candidates to contest in the countrywide elections at the levels of county assembly (ward), national assembly or parliament (constituency), women county representative, senatorial (senate), gubernatorial (county executive), and presidency (The [Kenyan] Election Act, 2011). Article 196 of Kenya’s 2010 Constitution allows those who win the party tickets to campaign for 3 months. The campaigns enable them to sell their agendas to the public before the countrywide elections. The people perceive the WhatsApp space as important for their participation in electoral processes and for electoral contestants to sell their agendas (Ooko, 2023). The people believe they have relative ownership of their WhatsApp spaces compared to radio and TV in Kenya. The perception is that radio and TV are prone to government censorship and commercialization.

WhatsApp Space in Bungoma County

The WhatsApp groups in Bungoma County operate at many levels, including international, national, regional, county, constituency, ward, clan, and family. A WhatsApp user can belong to all the group levels simultaneously. For instance, a family member from a particular clan can simultaneously belong to various WhatsApp groups: ward-level, constituency-level, county-level, national, and international. Access to WhatsApp spaces is through smartphones for the majority of the people, while others use internet cafés. Entrepreneurs run small-scale media services including internet cafés that offer internet services (County Government of Bungoma, 2020). The cafés enable people who can only access smartphones occasionally to use WhatsApp. They initially sync their smartphones to the café desktops to enable access to WhatsApp space for up to 14 days, before they are required to log in again/sync. The cafés also avail facilities such as electricity sockets for recharging cell phone batteries at a fee, to enable the people to stay connected. Some Bungoma County people living outside the county, who stay in other urban centers within and outside Kenya, from where they engage in gainful employment and other economic activities, also join the groups. Due to the majority people’s cultural/ethnic/primordial inclinations during electoral deliberations, the group members exclude the minority ethnic groups from WhatsApp groups, and they become emotive/affective when they suspect minority presence (in the groups) because they are perceived as political traitors. This prompts the minorities to infiltrate the groups through misrepresentation (Langmia, 2016), as a form of resistance to the exclusion. The excluded minorities surreptitiously access information on the political calculations of the majority group. This implies that the WhatsApp space is considered among the most important sources of information during electoral deliberations. It is made unavailable to minority groups to deny them political participation.

Based on the previous description, this study is critical because it investigates how history and culture affectively inform the majority and minority people’s use of WhatsApp space for exclusion and/or inclusion during electoral deliberations. The research was guided by the following questions: How does the simultaneous majority/minority status emerge during WhatsApp use? How does Afrocentric reward/punishment inform the people’s WhatsApp use? And how does insecurity/affective politics shape the people’s WhatsApp use?

Reviewed Literature

This sub-section traces Bungoma County’s affective politics in WhatsApp and social media to historical and ethno-cultural values. It takes a historical and ethno-cultural approach for contextualized evaluation of affective WhatsApp electoral deliberations. The literature review on WhatsApp is linked to social media in general. This approach is informed by the fact that the people of Bungoma County do not use the WhatsApp platform in isolation of other social media platforms during electoral deliberations. They use the platform in combination with other social media. They share messages from other social networking sites in their WhatsApp groups, extending and causing different affective reactions. Thus, the approach enables a holistic review of the affective WhatsApp electoral deliberations.

Kenyan Media Landscape and the Emergence of Ethnic/Civic Politics

During colonial rule and immediately after Kenya’s independence, the media landscape in Bungoma District, now renamed Bungoma County after the promulgation of the 2010 constitution, was dominated by the government-owned media—Kenya Broadcasting Corporation (KBC). The BBC ran regional news programs for a limited number of hours daily. According to Article 23 of the Kenya Information and Communication Act (KIC Act, 2012), it is the function of the government to provide and regulate the mass media infrastructure. However, private media have taken root since the era of media liberalization that started in the early 1990s, coinciding with the return to multiparty politics.

The media liberalization in the 1990s, initiated in Kenya by private and government media stakeholders, drastically changed the media landscape. For the most part, the media and communications sector liberalization freed up the airwaves following market demand rather than government decisions (KIC Act, 2012). It marked a turning point in the media and communication ecology, enabling policy frameworks to safeguard the media space to facilitate private media operations. It also enabled the acquisition of information and communication technology (ICT) equipment to improve and establish new media forms (Oriare et al., 2010). The media infrastructure was laid out across Kenya during the 1990s and into the 2000s, encouraging media entrepreneurs to invest in private media (KIC Act, 2012). Such infrastructural investment included internet fiber cables and telecommunication boosters that facilitated nationwide internet coverage.

Today, there is 2G, 3G, 4G, and 5G internet coverage across Kenya to allow for fast internet connection, enabling people to engage in online economic activities. This has opened up the communicative and economic potential of rural Kenya (Mandela, 2016). Before WhatsApp electoral deliberations came into vogue in 2013, the most popular form of public engagement included the use of FM radio station talk shows that consisted of phone-in sessions. The stations mushroomed after the airwaves were liberalized or freed. However, the radio was prone to government interference, ownership control/interests, and (over)-commercialization, prompting people to adopt social media, especially WhatsApp. The internet availability has inspired both the majoritarian Bukusu people and the minoritarian Kikuyu, Sabaot, tachoni, and Ateso, among others, residing in Bungoma’s numerous fast-urbanizing areas such as Chwele, Lugulu, Webuye, Kimilili, and Kibabii to use ethnically inclined WhatsApp groups for electoral deliberations. Besides the ethnic inclinations, people use social media for nationwide civic engagements.

Digital media in Kenyan civic and electoral politics is attributable to the 1990s media liberalization. Like Mutsvairo and Rønning’s (2020) discussion on social media’s role and struggles against authoritarianism, its success in addressing electoral injustice in Kenya remained contested until the 2017 elections. It was deployed for election monitoring with landmark success (Lynch et al., 2019). Before the 2017 elections, the citizenry fruitlessly engaged in election-related violence every election year to protest electoral malpractices, since the reintroduction of multiparty democracy in 1992 (Lynch et al., 2019; Wanyande, 1995). The violence climaxed in 2007 with over 1,000 people killed, and property worth millions of dollars destroyed (Lynch et al., 2019). In 2013, there were massive allegations by opposition parties and civil society that the incumbent government used technology to commit electoral malpractice during the presidential elections (KNCHR, 2017). In the electoral petition filed by opposition parties and the civil society, the opposition parties’ plea for transparency by the electoral body, under the hashtag #OpenTheServers, did not yield much (KNCHR, 2017). Citing the Kenya Law (2012, p. 6), the Supreme Court upheld the 2013 presidential elections citing “lack of sufficient evidence of electoral malpractice.” The opposition parties and civil society used social media in 2017 to complement their election monitoring. The consequence was the collection of evidence on electoral malpractices, which in part resulted in the successful 2017 presidential electoral results petition. This led to the nullification of the presidential election by the Supreme Court for the first time in Kenyan history (KNCHR, 2017, p. 28). In 2022, the Supreme Court turned the tables against the opposition which had appealed for the presidential election. The Supreme Court cited the use of digital media by the opposition and sections of civil society to manipulate evidence of electoral malpractice (Obulutsa & Houreld, 2022).

Insecurity/Affective Politics and the Trajectory of Kenyan Media

The historical and ethno-cultural values that have shaped the affective WhatsApp/social media use are traceable to colonial and post-colonial environments in Kenya. Busia (1967) discusses how colonialists affectively disrupted African indigenous media and the social structures that existed to establish non-indigenous media that would advance colonial interests, including social inequality. The inequality affectively prompted the Kenyan people who had acquired Western education to set up and own media to counter it (Mamdani, 2016). Wanyande (1995) explores the constant conflicts between African independent governments and the post-independent media in Kenya, arguing that immediately after Kenya attained independence, the new African governments did not decolonize the communicative outlets. Instead, they embarked on amending legislation to curtail people’s freedom of expression, like colonialists. The urban population became increasingly skeptical of the government-owned KBC. By the 1990s, people were consuming the private media that publicly criticized the government (Wanyande, 1995). The major weakness of the pioneer private media was the lack of infrastructure for nationwide reach, making them a preserve of a few urban elites. By the late 1990s and early 2000s, private vernacular radio stations had mushroomed across the county due to liberalization, leading to even more affective programming. Mitullah et al. (2015) discuss how Kenyan ethnic people’s oral tradition influenced the adoption of the mobile telephone into vernacular radio broadcasting’s affective public debates. The oral traditions were ultimately transposed onto social media platforms. The adoption shaped the people’s affective use of the radio stations during phone-in sessions, not only in the capital, Nairobi, but also in counties such as Bungoma (Oriare et al., 2010). According to the Media Council of Kenya (MCK, 2020), at some point, almost every Kenyan ethnic group had an ethnic language radio station. However, the government intensified censorship on the radio, arguing it provided public space for political incitement, especially in political oppositional zones such as Bungoma County (Oriare et al., 2010). People escaped government censorship by using mobile phone Short Message Service (SMS) for political mobilization (Ogola, 2015). However, the government censured SMSs too, arguing they invaded privacy and fuelled insecurity.

Ogola (2015) discusses how people’s ethinicized interpretation of government censorship as a curtailment of their freedom along ethnic lines led to the adaption of social media, notably, blogs for electoral deliberations. This marked the onset of social media use in Kenyan electoral deliberations. Although Wildermuth (2021) contests the increasing number of online users in Kenya, there was an emerging brand of citizen journalism involving individuals with powerful online presence such as Robert Alai, Denis Itumbi, Boniface Mwangi, and Cyprian Nyakundi with massive followings, larger than some traditional media houses. Notably, the debates on the individuals’ social media pages were affective and ethnically inclined (Ogola, 2015). Analogically, whereas elitist colonial and post-colonial media produced affective content for dissemination to the masses, the WhatsApp/social media users, who constituted the masses, were the dominant (co-)producers of the affective content, including in Bungoma County. Borrowing from Ekeh (1975), it is observable that during the colonial media era, affect was produced by the civic publics, while today, it is (co-)produced by the primordial publics. Agbozo et al. (2024) discuss how affective electoral politics in specific sub-Saharan contexts inform people’s choices on information technology. This is also the case for Bungoma County.

Papacharissi (2016, p. 17) contends that social media “evoke affective reactions” as the people stand up for justice. Affective politics on social media including the WhatsApp spaces, although universal, can be particularized to Bungoma County and Kenya during the 2007, 2013, 2017, and 2022 elections. The people’s ethnicized logic perceives election malpractices along the ethnic prismatic lenses, which results in ethnically inclined virtual and physical public protests and violence (KNCHR, 2017). Facebook was especially instrumental during the 2007 elections while WhatsApp was the most used application in 2013, 2017, and 2022 (Ooko, 2023). For instance, in 2013, 2017, and 2022, Alai and Itumbi (two prominent social media influencers) used their Facebook pages to share screenshots of WhatsApp group chats, alleging they were political elites’ secret electoral strategies. Such posts elicited ethnicity-inclined public electoral discussions, as the people revisited the real and perceived historical electoral injustices the past governments committed to people of their ethnicities. Ushahidi (2018) links social media/WhatsApp ethnicity-inclined deliberations to physical violence within counties in Kenya, including Bungoma. There is no documentation about the emotional violence caused by virtual attacks during deliberations in Bungoma County and Kenya due to the challenges associated with quantifying emotions (KNCHR, 2017). Although the highest-recorded election-related violence in Bungoma County and Kenya was in 2007, the highest emotional violence may have been during the 2013, 2017, and 2022 elections (BBC, 2022; KNCHR, 2017). We can surmise that emotions went viral online unlike in 2007 when social media use was still in its formative stages. The incumbent and opposition parties ratcheted the highest level of social media and ICT technologies in their electoral campaigns. Ironically, the Kenyan elections in 2013, 2017, and 2022 were cited as relatively peaceful compared to previous ones (Obulutsa & Houreld, 2022). Again, the “violence” could have deceptively shifted online meaning that emotional violence rather than physical violence was on the ascent.

Bungoma County people prefer the WhatsApp platform for electoral deliberations because they perceive it to provide safety due to the end-to-end encryption. This is despite Ooko’s (2023) assertion that encryption is false affordance, which does not necessarily ensure safety. Mbembe (2017) argues that social media, including WhatsApp, makes people operate on psychic bonds rather than material structures, which makes it difficult for their voices to be ignored in society. Castells (2008), although not explicitly focused on the political sphere, argues that the advantage of the internet is that it allows for the forging of weak egalitarian ties with strangers, in a classless pattern of interaction. This allows for the coming together of large numbers of people to demand accountability and justice, among other tenets of a democratic society. Ooko (2021, p. 149) explores how ethnic logics enable and limits people’s use of the WhatsApp space in Kabula ward, Bumula Constituency of Bungoma County, in which ethnically inclined members resent the “powers that previously marginalized them.” Although not explicit about underlying ethnic logic, Omanga (2019) discusses how the people of Nakuru County use the WhatsApp space to evaluate county governance. They come together to deliberate in WhatsApp groups on how to keep the County government in check.

Mbembe (2017) draws from the work of Comaroff (2012) to argue that the digital public sphere has facilitated a new form of politics he called “politics of the affect,” whereby people use social media including WhatsApp to actualise affective politics at the expense of logic. The affect preoccupies the people with more emotions and less sobriety during electoral deliberations. In his analysis of Pentecostalism and populism in Africa, Comaroff (2012, p. 13) argues that politics of affect is compounded by African people’s belief that “politics, governance, and public debates are rooted in a world of invisible powers—of witches, spirits, and demons.” We can localize the African people’s belief to Kenya, whereby the current leader, President William Ruto of the Kalenjin ethnicity, attends numerous church services to present to his political supporters the image of a spiritually upright leader. In his political rallies, he often openly refers to the opposition leader Odinga of the Luo ethnicity as a “witch doctor” (Nation, 2021). In the Kenyan context, this insinuates spiritual impurity along ethnic lines. Ruto aims to win more supporters from other ethnicities based on his self-proclaimed spiritual purity and discredit Odinga as a spiritually impure political opponent. The exchanges between Ruto and Odinga often jump from physical political platforms onto social media platforms such as WhatsApp. Social media, including WhatsApp, provides space for people’s implicit and explicit emotive deliberations and expressions.

After the introduction of county governments in 2010, the politics of ethnic majority versus ethnic minority representation in the Bungoma County have played out in the 2013, 2017, and 2022 gubernatorial elections (BBC, 2022; KNCHR, 2017). In the three elections, the winning Bukusu gubernatorial candidates picked their deputies from the non-Bukusu ethnicities. This enabled the winning candidates to get the support of the ethnic minorities needed to tip the scale. The Bungoma County Government Board (BCGB, 2020) prescribes ethnic/dialect/clan representation during job hiring at the county, sub-counties, constituencies, and wards. These ethnic prescriptions largely inform the WhatsApp group electoral deliberations. The ethnic majority perceives ethnic Bungoma County minorities including the Sabaot/Kalenin community in Mt. Elgon Constituency as sympathetic to the government of the late Daniel Arap Moi. (Moi was from the Kalenjin community.) The Sabaot/Kalenjin are alleged to have benefited economically (KNCHR, 2017) during the Moi presidency. This is arguably among the root causes of suspicion within neighboring ethnicities, in real life and during WhatsApp deliberations (KNCHR, 2017). The ethnic groups opposed to governments suffer forms of marginalization including not getting a fair share of national resources, which induces perceptions of economic lack among the people (KNCHR, 2017; KPHC, 2019). The sense of marginalization facilitates the perpetration of electoral malpractice including vote buying, which enables unpopular ethnic minority candidates with more economic resources to win elections (KNCHR, 2017; Lynch et al., 2019). To date, Bungoma County is prone to election-related violence because of perceived and real electoral injustice (KNCHR, 2017). Youth unemployment has pushed a large number of young people to take up informal jobs in the public transport sector, where they use bicycles, motorized tricycles, and motorcycle popularly known as boda boda and tuktuk to offer short-distance transport services to people (KPHC, 2019). Notably, the bodaboda and Tuktuk youths are hired by political elites during public rallies to generate content used on social media/WhatsApp to imply that the politicians have a massive following (Lynch et al., 2019). Against this background, people use WhatsApp spaces for affective ethnically inclined electoral engagements.

Theoretical Framework—Afrocentric Reward/Punishment

This sub-section theoretically anchors the article on Ekeh’s (1975) Afrocentric conceptualization of the public sphere and Langmia’s (2016) Teleco-presence theory of identity. Here Ekeh’s conceptualization is originally from the sociological background but has since been borrowed into media and politics. In the concept, Ekeh identifies two public realms of the African public sphere in post-colonial Africa: primordial and civic publics. Primordial publics largely consist of ethnic groupings and kinships, while civil publics consist of those in the civil service structure such as civil servants, military, and police, among others. He argues that the two publics have different types of moral linkages to the private realm: in a way different from the Western conception. The primordial public emphasizes social identity, while the civic public emphasizes material gain. He attributes the emergence of the two publics to Africa’s politico-historical and sociological factors including the European-originated bourgeois colonial administrators and the African bourgeois class born out of the colonial experience, respectively. The African people simultaneously belong to both the civic and primordial public. The two publics have unwritten logics of dialectics, that is, it is legitimate to rob the amoral civic public for the benefit of the primordial public. In turn, while in government positions, the civic publics rob the state to benefit the primordial publics. This is loosely comparable to Langmia’s (2016) discussion of “false” online identity, arguing it lacks the “genuineness embedded in the bi-directional interchange” (p. 268). We can use Omulo’s (2023) discussion of former Kenyan President Uhuru’s legacy to argue that both online and real-life people only become genuine when they expect to be rewarded for their genuineness.

Osaghae (2006, p. 237) contrasts Ekeh’s (1975) concept of the African public sphere, with the Western scholars’ conceptualization. He contends that the Western public sphere only focuses on “formal” organizations with the capacity to change the state, excluding the “primordial” public from their analysis. It is the “equivalent of civil society” and/or opposition parties in Africa. We can localize Ekeh’s (1975) concepts to WhatsApp use in Bungoma County, where the primordial publics consist of the ethnic relations between the majoritarian Bukusu people and the minoritarian people. On the other hand, the civic publics consist of the relations among the societal elites including County government officials and businesspeople. Even as the concept of societal elites is used here, we take caution about the problematic conceptualization of African middle class, within Ekeh’s (1975) sense.

The concept of publics is also visible in Naudé’s (2023) analysis of how South African people use online engagements to construct the meaning of self and society. Social media enables young people to “connect with each other and the world, refine their philosophy on life, and reflect on their views of the future” (p. 608). Ekeh (1975) observes that the two publics inform the African people’s conception of citizenship as citizenship is dependent on whether or not it is conceived in terms of the primordial or civic public. The majority Africans are primordial publics because they originate from ethnic groups. They are moral and obliged to “voluntarily” contribute to ethnic associations in exchange for intangible rewards such as social identity or psychological belonging/security. For instance, the minority Kikuyu people in Bungoma County seek ethnic relations with other counties consisting of majority Kikuyus such as Nyeri, Murang’a, and Kiambu for social identity and psychological belonging, even as they form civic publics in Bungoma. When in their civic public mode in Bungoma County, the people—both Kikuyus and other ethnicities—become amoral and over-emphasize economic/material value. For example, they seek to secure tenders, jobs, and contracts among others with the county government.

According to Ekeh (1975), the African people’s simultaneous belonging to the primordial and civic space brings dialectical tensions and confrontations within individuals. He argues, “all” African (primordial) people seek material gain from the civic publics, who only give grudgingly and under compulsion in exchange for social identity (p. 108). This constitutes the tensions and confrontations within the WhatsApp groups during electoral deliberations. The political elites make tribal and clan political alliances (primordial) to attain political power (civic). This causes further tensions during WhatsApp electoral deliberations within different ethnicities and political persuasions. We can loosely compare Ekeh’s (1975) dual publics with Langmia’s false online identity, in which both shift positions depending on the expected gains. The study’s theoretical contribution is to use Afrocentric theorization to test contemporary phenomena of social media use with a particular focus on WhatsApp.

Methodology

The article used both traditional and digital ethnography (Hine, 2020). The combination cushioned against each other’s weaknesses. We selected the five most popular politics-related WhatsApp groups ranked by a survey conducted by a leading vernacular FM radio station in the county. The survey was done through local people’s phone-in sessions in which they listed the most popular WhatsApp groups they belong to or know about. The survey indicated countless politically related WhatsApp groups in the County, but the most notable ones were about 10. The WhatsApp groups’ membership was spread across the nine constituencies in the County. One of the authors joined the groups with members’ consent, as a covert participant observer for 3 months—September to December 2020. The consent was sought through the group administrators, who notified members about the researchers’ presence and objective. The members did not object. This is likely because two of the three researchers were born in Bungoma County, and their (user)names did not raise any suspicions from the members. Administrators reminded the members about the researchers’ presence and objective in the groups every morning for 3 months via a post. Although the researchers expected that their presence in the groups would somehow shape the deliberations, there were no observed changes in the deliberations.

While in the groups, the researcher identified and selected interview respondents who regularly and articulately deliberated issues in the groups. The researcher assumed that articulate respondents would be more forthcoming with information and express themselves clearly. However, some active members selected from the groups did not live in Bungoma County and so did not have firsthand information sought in the research. They were thus left out. There were pseudo accounts in the groups, which the researchers interpreted as the minorities’ resistance to majoritarian exclusionism during WhatsApp electoral deliberations.

After 3 months of observation, in-depth face-to-face interviews were conducted during another 3 months of traditional ethnography in the County between January and March 2021. The interviews took place with two categories of participants: first, the WhatsApp users from Bungoma County, and second, the WhatsApp group administrators from the five selected WhatsApp groups. During interviews, researchers discovered a network of information, whereby people living in the County shared political information with people living outside the county. Those living outside the county actively shared the information on various WhatsApp groups as if it were their own. The researchers used the snowballing method to sample the inactive grassroots members who shared the information with those outside the county. Face-to-face interviews were conducted with them too. The interviews were “semi-structured/semi-standardized” in nature to allow for the adaptation of the research instrument to the level of comprehension and articulacy of the respondents (Fielding, 2012, p. 124). In some cases where respondents could not communicate in English, the researchers asked questions in Swahili. This compromised the findings because some information was lost during translation. Some colloquial and slang words and pictograms such as emojis could not be readily translatable into the Swahili language. The researchers relied on descriptions and demonstrations of symbols including emojis to familiarize respondents with the questions asked.

There are nine constituencies in Bungoma County, namely, Kanduyi, Bumula, Kabuchai, Sirisia, Kimilili, Mt Elgon, Tongaren, Webuye East, and Webuye West. The researchers sampled between five and seven respondents from each constituency depending on the population size. The most populated constituency was Kanduyi where seven respondents were sampled, while Mt Elgon was the least populated constituency where five respondents were sampled. A total of 59 respondents were sampled. The rationale for 59 respondents was based on the qualitative information they possessed and how articulate they were during online engagements. Thus, the sampling method was convenient and not scientific. We considered respondents as a sample representation of the total population in Bungoma County because online communities tend to be homogeneous (Hine, 2020). After interviewing 42 respondents, we started observing saturated responses. We stopped interviewing after 59 responses. Increasing the number of respondents may have increased costs with no significant difference in the findings. The interview questions mainly sought to understand the people’s circumstances and experiences by first obtaining their background and then probing for detailed information on how they participate in WhatsApp group deliberations; the causes of the tensions and confrontations in the groups; how tensions and confrontations affect their participation; and the support systems participants rely on from the administrators. Among the challenges experienced is people’s reluctance to participate, saying they had been interviewed “way too many times.” The researchers informed them about the importance of the research, and they agreed to take part voluntarily.

For data collection, the researchers retrieved viral WhatsApp voice messages relating to electoral deliberations and observations of the WhatsApp group electoral deliberations about Bungoma County. Viral messages are those largely shared within the five groups. The researchers also acquired information about interpretation-peculiar behaviors among the people being researched. These involved subtle but decisive gestures/actions informed by the respondents’ cultural practices; expressed through emojis, GIFs, and textual symbols such as exclamation marks and dots, among others.

We used a thematic analysis approach, where they used starting themes as “WhatsApp use,” “tensions,” “confrontations,” and “consensus.” The themes were obtained from the interview responses, the content of the WhatsApp viral messages, and observations. The coding was manually done to retain as much of the qualitative values as possible. We use an inductive approach to coding with the units of analysis as “simultaneous majority/minority status,” “Afrocentric Rewards,” and “Affect.” Immediately we embarked on the coding process, and many additional themes emerged from the data, including infiltrating groups, punishment, and insecurity, among others, discussed in the following findings section. The research adopted a hierarchical coding method to allow for relationships between codes to be established and worked with (Kelle, 2000, pp. 287–289).

Findings and Discussions

The Simultaneous Majority/Minority Status and WhatsApp Use

In this sub-section, we discuss the findings on WhatsApp use in Bungoma County, in which historically and ethnically originated tensions, confrontations, and consensus are featured. The WhatsApp group membership is largely based on ethnicity, clan, and/or other ethnicity-related considerations. The people historically contextualize the ethnic considerations within colonial and post-independent governments. The ethnic majoritarian Bukusu people form their own WhatsApp groups for electoral deliberations and deny admission to minorities into the groups. For example, they deny the minoritarian Kikuyu people admission into their groups, ostensibly because they undermine Bukusus by obtaining political information from groups to sell to their elite Kikuyu political opponents, who benefit from colonial structures and ties. The minorities are not willing to be left out of these majoritarian groups because they want to stay informed about the majoritarian political plans, hence the dialectical tensions and confrontations from within. The information is important in case majoritarian people plan attacks on them as is often the case. The attacks may be both online and physical, emanating from deliberations that trigger the majoritarian people’s perception that minorities are beneficiaries of the previous governments’ political injustices. An interlocutor argued that ethnic minority elites were allocated resources by colonialists and subsequent governments in exchange for political support, at the time when the majority Bukusus were strongly in opposition politics.

It was observed that the minorities infiltrate the WhatsApp groups with fake usernames that suggest they belong to the majoritarian Bukusu people’s ethnicity. The members in the WhatsApp groups largely deliberate in the English language, which enables the infiltrating minorities to take part in the deliberations without suspicion. If the majority people suspect the presence of the minorities, they start deliberating in the Bukusu language dialect. The minority Kikuyu people would remain silent during deliberations since they do not understand the dialect. The group members would appeal to the administrators to eject all the silent members on suspicion that they are infiltrators. In an attempt not to be seen as silent members, some infiltrators try using emojis and GIFs, among other symbols during Bukusu dialect deliberations. However, some infiltrators use them counterproductively. For example, an interlocutor narrated how she used a laughter emoji to react to a chat about a member saying that the majoritarian people do not go to minority people’s ancestral counties to influence their politics. The interlocutor said she was asked using the Bukusu dialect why she was laughing. She could not respond, and so she was ejected. The cycle of ejections and infiltrations into the groups takes place, with the majority people seeking to guard the groups’ political information and the minority people seeking to obtain the information. Minority people are unwilling to start their own WhatsApp groups because groups without majoritarian inclusion would not enable them to acquire information about the majority of people’s political plans.

After the electoral period, WhatsApp groups become less active. The tensions and confrontations subside. Visibly, consensus is embraced as members co-exist with less suspicion. Drawing from Ekeh’s (1975) unwritten logic of dialectics, we can explain the factors for consensus. Arguably, the minority Kikuyus (civic public) reward some of their economic resources to the majority Bukusus (primordial public) in exchange for co-existence/belonging. In any case, the tensions and confrontations result from people’s desire to elect candidates who would favor them during economic resource allocation.

Based on the aforementioned analysis, we conceptualized that a section of Bungoma people voluntarily and simultaneously belong to the minority and majority grouping on WhatsApp and in real life. We termed the belonging as “simultaneous majority/minority status.” In relation to this, our respondents Wairimu and Lagat (not their real names) narrated that although they reside in Bungoma, they also associate with other counties. Thus, they simultaneously form a minority and a majority, physically and on WhatsApp groups in more than one county. Most Kikuyu respondents said they are unwilling to leave Bungoma County and the WhatsApp groups, where they are minorities, to relocate to the majority Kikuyu counties such as Murang’a, where they would be part of majoritarian WhatsApp groups. They argue they have economic wealth in Bungoma County which ties them there. As a result, they simultaneously belong to Bungoma and Murang’a, easily moving between the primordial public and civic public. They obtain social belonging by associating with the latter, which is their ancestorial origin, from where they share ethnicity with the majoritarian people. Therefore, the Kikuyu minorities in Bungoma County also form a majority in Murang’a County; from where they also deny WhatsApp group admissions to ethnicities that form minorities there. Out of Kenya’s 47 counties, most of the 43 Kenyan ethnicities form an ethnic majority in the county of their origin, and an ethnic minority in another. The exception includes counties such as Nakuru, whose indigenous people were displaced to settle Kikuyu ethnic people, whose land was first forcefully acquired by colonialists and later by Kikuyu elites, including the first Kenyan president Kenyatta. The aforementioned analysis shows that Ekeh’s (1975) conceptualization of African publics does not entirely fit the Kenyan current and historical political conditions. Despite the misfit, the majority/minority compositions directly inform the WhatsApp platform membership and deliberations with African characteristics.

Wairimu says:

. . . I use these WhatsApp usernames because I need to masquerade as a (majoritarian) Bukusu person without suspicion. This way I will stay informed about the politics of the County. But most important, nijue kama kuna msako inakuja nijipange (to keep me aware of impending politically motivated physical attacks so that I may escape in time).

We can borrow from Wiredu’s (1997) “super consensus” democracy to expound on tribal relations involving simultaneous majority/minority status. In this case, we shall use the Sabaot/Bukusu people’s ethnic relations within Bungoma County. Like Kikuyus, the Sabaot people form a minority in Bungoma County, although they simultaneously form a majority in Mt Elgon Constituency within Bungoma County. In the constituency, the Bukusu people form the minority, even though they simultaneously form a majority in the larger Bungoma County. Thus, the Sabaot and Bukusu ethnicities have a simultaneous majority and minority status within Bungoma County. The minority Bukusu ethnicity in Mt Elgon Constituency does not contemplate relocating to join the majority Bukusus in neighboring constituencies such as Kimilili, Kabuchai, and Siririsa; even though Mt Elgon Constituency has the highest level of insecurity, gendered intimidation, cultural and ethnic discriminations, among others. Factors such as economic resource (re)allocation, historical/cultural heritage, and clannism, among others, inform the people’s choice to stay as a minority, as they easily move between the two publics—primordial and civic. The minority-majority relations are forged in confrontational tensions motivated by the sharing of scarce resources, ethno-cultural identities, and mutual suspicions. The WhatsApp spaces are used by ethnic majority to deliberate on how to deprive the ethnic minority of the resources. This is over and above depriving the minority of the WhatsApp communicative space during electoral deliberations. Ethnic minority people falsify their online identity along the lines of Langmia’s (2016) theorization, to survive in majoritarian WhatsApp groups. Lagat narration that:

. . . Before they removed me from the groups, I started receiving threats on my phone about my WhatsApp group posts on corruption. They [those threatening] had just killed a friend who openly criticised the incumbent in the County. Earlier they [incumbents—County administration?] had refused to allocate me a tender to make furniture for the nearby school even though I am a qualified engineer born and brought up in the place . . .

From the narrations by Wairimu and Lagat mentioned earlier, during WhatsApp electoral deliberations, the ethnic majority people emphasize resource ownership/allocation and deprivation of ethnic minorities, who are perceived as political “enemies.” If we draw on Appadurai’s (2004, p. 63) conception of WhatsApp as a “valuable communicative resource,” we can affirm that the neo-colonial structures and government ideologies enable ethnic majoritarian people to keep minorities uninformed to allocate themselves more economic resources, as rewards (Ekeh, 1975).

In addition, the neo-colonial structures and governments enable ethnic majority Sabaots within the Mt. Elgon Constituency to use WhatsApp space to employ brutal tactics (Matolino, 2016, p. 41) during political engagements with the ethnic minority Bukusu people. The constituency ethnic minorities are tracked from WhatsApp groups and physically identified for punishment because of their deliberations that are critical of the incumbents. In other cases, the ethnic majority Sabaots hold WhatsApp electoral deliberations in their vernacular language to prevent the ethnic minority Bukusu people who may have infiltrated the groups from understanding the messages.

Lagat narrates that:

. . . They [killers] had just killed a friend who openly criticised the incumbent before they started threatening me. “Tutakujia kichwa yako na 100k . . .” (The killers brag that they are paid 100k/USD 1000 after the killing a critic to the incumbent). The police don’t help.

Afrocentric Reward/Punishment and WhatsApp Use

In this article, the application of Ekeh (1975), the concept of Afrocentric reward means the peculiar African politics in which individuals are rewarded based on their political support. Ekeh (1975) discusses that civil servants find it legitimate to rob the civic public for the benefit of the primordial public, to nurture it to ensure it does not come to harm. However, this section of our findings reveals an ill-fit in the application of the concept. From the aforementioned Lagat’s narration, we realize that the incumbents punish the WhatsApp platform whistleblowers for exposing corruption and other societal evils during their regimes. This brings tensions and confrontations seminally discussed in Ekeh’s (1975) concept of the African public sphere. In his argument, the two African publics are based on the reward system, where the civic publics such as civil servants in the county are rewarded with jobs, tenders, and contracts, among other economic benefits. Our findings reveal a contrary stance whereby those critical of the incumbent are punished. The incumbents punish critics because they interpret criticism as a non-African concept borrowed from Western democracy. For instance, the information network relations we discovered during fieldwork existed because the inactive members feared attacks if they posted information critical of the incumbents. Thus, they felt okay when the people outside the county posted the information even if it was not attributed to them. Those outside the County felt secure enough to post the information because they were physically distant from the incumbents.

In indigenous Africa, the ethnic minority either waited for their turn to lead or walked away to establish their own “kingdoms,” villages, among other political units (Busia, 1967). From Lagat’s narration, the incumbents assume that the formal structures such as the civil society and the official opposition supposed to check them should be from other ethnicities and not their own. Thus, Lagat is punished by incumbents from his place of birth because they do not recognize him as a formal watchdog, due to their shared ethnicity. This falls outside Ekeh’s (1975) scope of conceptualization of the African publics.

The incumbents believe they can offer monetary rewards to their ethnic people in exchange for intangible rewards such as being praised as the best leaders. They expect those monetarily rewarded to praise them in the WhatsApp groups as the best-performing leaders (Ekeh, 1975). This leads to surveillance and counter-surveillance among people during WhatsApp electoral deliberations. The incumbents expect those rewarded to campaign for them, online and offline, and ultimately vote for them during elections despite their poor performance in public offices. These lead to tensions and confrontations, which only usher in consensus after elections once the winners are announced.

The cycle of surveillance, suspicion, and exclusionism/punishment recurs every election period, in real life and in WhatsApp groups. According to interview responses, even if the excluded/punished people became incumbents through the next elections, they would need to materially reward their primordial publics in exchange for the needed intangible rewards to enable them to stay in power (Ekeh, 1975). Lagat observed that even if the people removed from WhatsApp groups were made the new WhatsApp administrators in the various groups, they would seek to exclude/punish others perceived as opponents for political gain. Thus, the relative stability that holds the people together before the next electoral cycle resides in Wiredu’s (1997) aforesaid “super consensus” democracy, in which they believe that they will not be permanently placed in a minority position.

Insecurity/Affective Politics and WhatsApp Space

Insecurity/affect shapes how people use WhatsApp space in electoral deliberations. We can conflate the previous narrations with Matolino’s (2016) discussion on the agreement by Afrocentric scholars that violence was a highly observable feature in indigenous Africa. We can also philosophically locate the insecurity/affect in Bungoma County WhatsApp use in Mbembe’s (2001, pp. 101–102) description of indigenous Africa as “chaotically pluralistic.” He draws from Comaroff’s (2012) description of Pentecostalism as a catalyst for new politics of affect in Africa, to argue that the WhatsApp space re-mediates the politics of affect (Mbembe, 2017). Thus, the WhatsApp space actualizes the affective politics that is both externally/colonially and internally/ethnically originated. Colonial history and ethnicity inform the un/fair resource allocation within the county fueling the affect. WhatsApp has enabled people to transpose the affect onto virtual platforms. Goffman (1959, p. 1) calls it the relaying of people’s real-life performative actions.

Namisi narrates that:

. . . Therefore, mababuu zetu hawawatambui hapa (our ancestors do not recognise them [minority people] here in Bungoma County) . . .

From the narration, we can see that WhatsApp space enables people to compound the politics of affect. It allows the people to express the belief that “politics, governance, and public debates are rooted in a world of invisible powers—of witches, spirits, and demons Comaroff (2012, p. 13).” As a result, the WhatsApp electoral deliberations are robbed of logic, making ethnic minority people vulnerable to marginalization. Although it is difficult to quantify the spirituality of the people of Bungoma County, it tremendously informs how and why they use WhatsApp space in electoral deliberations. Therefore, WhatsApp space provides a platform for the intersectionality between the people’s spirituality and politics of affect; constituting the fuel that keeps the people’s emotions raging and logic waning.

However, if we read Mbembe’s (2001, pp. 101–102) description of indigenous Africa as “chaotically pluralistic” in a contemporary way, we can interpret the people’s affect as the presence of space for public expression. Comparatively, many nations across the globe with the most “progressive” constitutions allow for the public space displayed in Mbembe’s (2001) description. Like Bungoma County, citizenry globally, including in countries such as the United States and United Kingdom, is discernibly intolerant to others, with renewed energy for explicit rejection of conventional notions of reason, in what some scholars associate with the rise of nationalism. Therefore, although particular to Bungoma, some tensions and confrontations witnessed during the WhatsApp electoral deliberations are transnational.

Conclusion

The people of Bungoma County adapted the WhatsApp platform to their electoral deliberations to resolve their historical and cultural issues. The tensions, confrontations, and consensus that characterize the deliberations are attributable to African ethnic culture and history. The African people’s dual “citizenship” enables simultaneous belonging to both primordial and civic publics. They also simultaneously belong to the majority and minority ethnic groupings. This belonging is largely responsible for the tensions and confrontations, which spill into WhatsApp groups during electoral deliberations. The ethnic majoritarian people in power reward/punish the ethnic minorities by excluding them from electoral deliberations and resource allocation. This prompts the minorities to assume false online identities, as a form of resistance to exclusionism. Although unjust, insecure, affective, and devoid of logic, the WhatsApp electoral deliberations are a sign of the presence of space for Bungoma people’s freedom of expression, comparable to the present-day Western world. After the elections, the people reach a consensus and co-exist in the county until the next electoral cycle. The implication of the research is to invite more Afrocentric theorization on technology adaptation in African societies, to further explore how the African people are culturally adaptive to technology.