Why do different cultures form and persist? Learning from the case of Makerere University

Abstract

Culture is a central concept in the social sciences. It is also difficult to examine rigorously. I study the oldest university in East Africa and a cradle of political elites, Makerere University, where halls of residence developed distinct cultures in the 1970s such that some hall cultures are activist (e.g. Lumumba Hall) while others are respectful to authorities (e.g. Livingstone Hall) even though assignment to halls has been random since 1970. I leverage this unique setting to understand how culture forms and affects the values and behaviours of young adults. Participant observation, interviews and archives suggest that cultural differences arose from critical junctures that biased group (hall) composition and from intergroup (inter-hall) competition. Hall governments promote cultural and institutional persistence through the intergenerational transmission of norms and practices, thereby highlighting the role of political hierarchy in reproducing culture.

INTRODUCTION

Makerere University, in Uganda, is the oldest university in East Africa. It was founded in 1922 as Makerere College and has been a cradle of political, social and economic elites in East Africa ever since. Makerere alumni include former presidents Milton Obote of Uganda, Julius Nyerere of Tanzania, Mwai Kibaki of Kenya and Joseph Kabila of the Democratic Republic of the Congo. The University is thus well-known in East Africa and beyond.

One less well-known feature of Makerere University is that its halls of residence, where students reside, developed distinct cultures in the 1960s and 1970s. Livingstone Hall, which opened in 1959 and was named after missionary and explorer David Livingstone, has long been known as a ‘hall of Gentlemen’ because its culture emphasises a respectful and quiet demeanour. Lumumba Hall, named after Congo's independence leader Patrice Lumumba, has been a socially and politically ‘activist hall’ since its opening in 1971. Lumumba Hall played an important role in the limited student resistance to Idi Amin's dictatorship (1971–1979). Its leaders still organise university-wide demonstrations against grievances such as increased tuition fees. Northcote Hall, inaugurated in 1952, was a socially cohesive ‘hall of statesmen’ that developed a well-defined political and military hierarchy (the State Supreme Revolutionary Command Council).

Even less well-known is that the Administration has allocated students to halls of residence randomly since 1970.¹ Therefore, the halls of Makerere constitute a randomised natural experiment that allows me to examine the influence of culture on attitudes and behaviours – in a setting where the broader social environment of Makerere and Uganda is shared. This is exactly what I study in a companion article, where I survey recent students and alumni to understand the extent to which exposure to a new culture influences young adults. I find that random assignment to a new culture influences students’ identity, levels of interpersonal trust, generosity and activism but not their academic performance, political preferences and ideology, among other outcomes (Ricart-Huguet & Paluck 2023). Overall, the influence of hall culture on interpersonal outcomes (e.g. trust) is higher than on individual outcomes, many of which are not affected by culture (e.g. career choice).

The combination of these two facts – cultural differentiation and random allocation – raises a fascinating puzzle: Why did different cultures emerge at all given that the Administration assigned students to halls randomly since 1970 and given that all students share the same campus-wide and even hall-level formal institutions? While the companion article just summarised examines the effects of hall cultures, this article examines their origins and evolution.

To account for the origins and evolution of these cultures, I conducted 43 in-depth interviews with senior alumni, 50 interviews with more recent alumni, six focus groups, and collected primary and archival materials in addition to six non-consecutive months of participant observation. Methodologically, the study is conceived as a set of comparative ethnographic case studies (the groups or halls) within a larger case study (Makerere University).

I argue that intergroup (interhall) competition for status facilitates cultural formation, consistent with optimal distinctiveness theory (ODT). Brewer (1991: 477), who pioneered ODT, argues that ‘social identity can be viewed as a compromise between assimilation and differentiation from others’. We value belonging to an ethnic group but we also value comparing our group to another. Further, Hornsey & Hogg (1999: 543) experimentally confirmed that ‘subgroup differentiation [can be] a response to an overly inclusive group’. Makerere was the only university in Uganda until 1989, an extreme case of an overly inclusive group. Hence, interhall rather than interuniversity competitions were central to university life (de Bunsen 1967).

However, ODT does not explain the set of values and behaviours a hall converges upon (the particular ‘cultural equilibrium’). Why is the social norm to be an activist in Lumumba Hall but a gentleman in Livingstone Hall? I show that contingent events (critical junctures) in the 1960s, when assignment to halls was not yet random, biased the initial population of some halls and led these to converge on certain values and behaviours. Further, three other halls that did not experience a contingent event that biased their populations did not develop a culture. Thus, intergroup competition and a shock to a group's composition may be necessary to form a culture.²

Cultures persist even under randomisation, I argue, because of two reasons. First, the yearly replacement rate at the halls is high but not too high (about 25% considering most degrees are four years). Second, each group (hall residents) elects a government or cabinet, elected by and composed of students, that is key to the intergenerational transmission of culture from seniors to freshmen. I also document how randomisation in 1970 led to cultural change in Northcote Hall and a mass expulsion of its residents led to the hall's cultural death.

I define culture as shared values, social norms and symbols (a system of meaning) linked to a set of behavioural practices (i.e. customary or socially approved behaviours) (Sewell Jr. 1999). This definition is compatible with others that specifically concern student subcultures, ‘a segment of the student body at a given institution holding a value orientation distinctive of that of the college community’ (Gottlieb & Hodgkins 1963: 272). Culture and institutions are sometimes confused because they overlap, but I clarify that culture cannot be reduced to informal institutions, socially sanctioned but unwritten rules, or to formal institutions, legally sanctioned and typically written rules (North 1990). For example, activism in Lumumba Hall and gentlemanliness in Livingstone Hall are informal institutions that are embedded in their respective hall cultures. By contrast, some symbols such as flags and myths, whether of halls of residence at Makerere or of nation-states, do not necessarily prescribe rules and yet they provide meaning – i.e. they are not institutions but are part of the culture.

In sum, the first contribution in this article is to improve our understanding of how cultures originate and evolve through the combination of rich comparative case studies and existing social science theory (next section). We know that culture affects social capital (Putnam 1993), democratic attitudes (Almond & Verba 1963), identity (Laitin 1986), conflict (Huntington 1993) and economic growth (Algan & Cahuc 2010), but we know much less about how cultural differences originate.³ While evolutionary pressures and environmental differences can explain long-run differences in culture (Henrich 2017; Giuliano & Nunn 2021), this article shows that short-run inter-group competition combined with contingent events can also generate cultural differences, even when the local environment (Makerere and Uganda) is shared.

The second contribution is to conduct a deep examination into the young elites educated at Makerere since independence from colonial rule to further understand how they impacted Ugandan politics and society. In doing so, I join the work of scholars (Byaruhanga 2006; Mills 2006) and politicians (Mahama 2012; Rukikaire 2019) who have highlighted the importance of higher education, and of the elites educated therein, to understand contemporary African politics and society. The remainder of the article describes my fieldwork and then threads an analytical narrative of cultural formation, persistence and change at Makerere University.

CULTURAL FORMATION AND DIFFERENTIATION BETWEEN GROUPS

Cultures are commonly viewed as ‘slow-changing’ (Guiso et al. 2006; Ross 2008: 156) but there are also examples of short-term change (Mead 1956; Alesina & Fuchs-Schündeln 2007).⁴ I present four explanations that could lead to short-run cultural formation: names as symbolic focal points; differences fostered by a central authority; intergroup competition; and contingent events. The case studies provide limited support for the first explanation and no support for the second, but ample support for the third and fourth.

First, a name can be a salient feature or symbol around which early members of a group could converge. Names can be focal points in the game-theoretic sense of relevant coordination points (Schelling 1960) and in the common sense of being important to an organisation. Sir Geoffry Northcote was a colonial officer involved with military operations in Kenya (consistent with Northcote Hall's paramilitary and hierarchical culture); David Livingstone was regarded as a British gentleman (Livingstone Hall is a hall of gentlemen); and Patrice Lumumba was a Congolese pan-Africanist activist and politician (Lumumba Hall is an activist hall). The empirical sections show that the name was irrelevant to hall culture (in Northcote Hall) or the result of hall culture (Lumumba Hall). Only in the case of Livingstone Hall might the name have provided a focal point to coordinate on norms that were antithetical to those of the other halls.⁵

Second, a central authority may instil or impose particular values and behaviours (Young 1996: 106). From lab experiments (Tajfel & Turner 2004) to divide-and-rule colonial policy and state-building, those in power have engineered differences between smaller and larger groups (e.g. ethnic minority vs. ethnic majority) for strategic purposes, such as controlling their subjects or increasing intergroup competition (Horowitz 1985). Similarly, university authorities (e.g. a vice-chancellor or a dean) could have fostered hall cultures. Not one interviewee suggested such top-down effort by Makerere authorities. If anything, the Dean of Students explained that the randomisation method he implemented in 1970 was ‘meant to prevent differences’ and homogenise populations across halls (interview 02, see Table A.1 in supplementary online material for the list of interviews).

A third explanation is that intergroup competition for status results in cultural differences that allow individuals to balance two opposing needs: social inclusion and social differentiation (Brewer 2007: 731). Optimal Distinctiveness Theory (ODT) argues that ‘social identity can be viewed as a compromise between assimilation and differentiation from others, where the need for deindividuation is satisfied within in-groups, while the need for distinctiveness is met through intergroup comparisons’ (Brewer 1991: 477).⁶ The in-group can be hall members, co-ethnics or co-nationals and out-group members are the rest (this is consistent with Anderson's (1983) definition of the nation as a ‘limited’ imagined community). Hornsey & Hogg (1999: 544), among others, experimentally validate ODT by showing that ‘membership of a highly inclusive superordinate group [University of Queensland] will provoke a search for subgroup differentiation [their major]’. Makerere was the only university in Uganda until 1989, an extreme case of an overly inclusive group. Interhall rather than interuniversity competitions facilitated differentiation while fostering hall-level social identity.

Scholars that view social norms as the equilibria of repeated coordination games present a mathematical but fundamentally similar intuition to ODT. ‘Starting from arbitrary initial conditions, different [subgroups] may well end up with different norms. That is, there will be near-uniformity of behavior within each village and substantially different behaviors across villages. This is the local conformity/global diversity effect’ (Burke & Young 2011: 314). Makerere students socialise in halls, so we may observe ‘local conformity’ in halls and ‘global diversity’ at the university.

I argue that, while useful, ODT is not sufficient to understand cultural formation because it does not explain (i) convergence on particular norms and behaviours and (ii) why not all halls developed a culture. A fourth hypothesis can account for these two puzzles: contingent events that bias the composition of a population, either by selection ab initio or by chance after a shock. Such events can make some values and behaviours more prevalent than others and hence facilitate coordination on those. The empirical sections argue that two contingencies in the 1960s, when assignment to halls was not yet random, favoured convergence on certain norms and behaviours at Northcote, Livingstone and Lumumba halls. Halls that did not experience these contingencies – the partly unintended consequences of decisions by two Deans of Students – did not develop a culture (Mitchell Hall, University Hall and Nkrumah Hall).⁷ Thus, intergroup competition and an early shock to the group's composition may be necessary to form a culture.

Mechanisms of cultural persistence and change

Cultural formation and cultural evolution are partly distinct processes. The period in which members of a group converge on an initial set of social norms is different from subsequent periods during which initial norms may persist or change. In the case of Makerere University, the period of cultural formation begins in the early 1960s and ends in the early 1970s: all available evidence indicates that hall cultures did not exist before the mid-1960s and no new culture formed after the early 1970s. After the mid-1970s, we observe much cultural persistence and some change.

An efficiency view of culture and institutions, as in the early North & Thomas (1973), would predict that all groups should emulate the best-suited culture and rules for the goals at hand (e.g. academic performance and campus political leadership), such that the most efficient set of norms prevails. This view is commonly rejected even by social scientists that view culture and institutions as rational (Bednar & Page 2007). In reality, from the later North (1990) to anthropologists and evolutionary game theorists (Axelrod 1986), many emphasise that ‘every society has its own set of social norms. This is because, in many situations, coordination on a behaviour is more important than which behaviour is taken. In such a situation, multiple equilibria arise’ (Matsui 1996: 265). The halls at Makerere are no different: multiple cultural equilibria remain in spite of the high intergroup contact and proximity between halls (five-minute walks). As social psychology and evolutionary game theory highlight, a distinct social identity is valuable in itself.

What mechanisms explain the persistence of different cultures? This is interesting because random allocation since 1970 should weaken cultures over time as new members replace old ones. First, culture may persist absent self-selection if the population replacement rate in every period is low enough. The yearly student replacement rate in college is about 25% because most bachelor's degrees last four years. At the limit, a 100% population replacement rate should prevent the inter-generational transmission of culture. That is the goal of genocidal dictators or, less dramatically, the result of expelling all members of a group or residents of a hall, as we will see.

Second, the political hierarchy of a group may facilitate cultural persistence. Standard evolutionary models consider societies with ‘no central authority’ where individuals are atomised units evolving in a social system (Axelrod 1986: 1095; Burke & Young 2011: 318). In reality, political hierarchies exist even in moderately complex societies. Leaders influence the group because of their power and high status in the social hierarchy (Paluck et al. 2016: 566). Universities also have political hierarchies. Each hall at Makerere includes a hall government that is elected by and composed of students. Each hall government consists of a Chairman and a cabinet, including a Minister of Culture, all of whom are elected individually. Two reasons arguably make hall governments a ‘hard case’ to show how political hierarchies contribute to reproducing culture. First, students are elected yearly and a student can be in the cabinet for at most three years (sophomore through senior year). Longer tenures would facilitate cultural reproduction. Second, students are elected democratically after campaigns that combine material rewards in exchange for support (clientelism) with symbolic displays of adherence to hall culture. Democracy in small societies tends to foster egalitarian norms and thus diminish the role of hierarchical leaders. Nonetheless, I argue that hall hierarchies matter and that these leaders have had an outsized influence on campus affairs and, in fact, on Ugandan politics (many became Members of Parliament, see Table A.1).

HISTORICAL CONTEXT

Founded in 1922 by the British Government as Makerere College, Makerere is the oldest university in East Africa.⁸ In 1919, Governor of Uganda Robert Coryndon set up a committee to build the first training college of East Africa due to the increasing need to train a skilled workforce in the region (MacPherson 1964: 2).⁹ Makerere became a university in 1949. Most faculty and senior administrators were British before Ugandan independence in 1962, when the process of Africanisation at Makerere began to accelerate (Sicherman 2006).

Makerere has long been a knowledge hub in East Africa (MacPherson 1964), a cradle of African elites (Goldthorpe 1965), and the centre of social and political activism in Uganda (Byaruhanga 2006). It follows, then, that the setting and individuals studied here are not representative of Ugandan society or of any society at large. Makerere students were and continue to be disproportionately represented in the civil service, politics and every high-status sector of society (Goldthorpe 1965; Byaruhanga 2006). Alumni include heads of state and prime ministers such as Joseph Kabila (DRC), Julius Nyerere and Benjamin Mkapa (Tanzania), Mwai Kibaki (Kenya), and Milton Obote and Yusuf Lule (Uganda) (Figure A.7).

Halls of residence are integral to Makerere's history. Mills (2006) provides a rich social history of Makerere that embeds student life on campus and at the halls into the broader educational aspirations and political events of colonial and independent Uganda. This article, by contrast, tries to understand why different cultures originated and developed in the second half of the 20th century given that, different halls notwithstanding, all students shared the same campus.

Data and cases

To understand cultural formation and persistence, I completed six months of fieldwork between 2015 and 2018 and conducted nearly 100 semi-structured interviews between 2014 and 2018, 43 of them with alumni who graduated before 2000 (listed in Figure A.1) and the rest with post-2000 students and Makerere staff. I always resided on campus during these non-consecutive six months because it greatly facilitates participant observation. I joined campus events during that time while keeping my research interest very generic to avoid revealing its purpose. The interviews were all conducted in English, in person, and in Kampala.¹⁰ While some interviewees are retired, others enjoy prominent positions in the country, from Prime Minister Rugunda to multiple MPs and well-known faculty. The average interview lasted about 90 minutes. Prior to the interviews, I had conducted six two-hour focus groups with recent and senior alumni.

Alumni were selected based on their past roles on campus and availability (some were deceased and others abroad). The goal, in the spirit of many process-tracing studies, was not a representative sample but to ‘obtain information about specific events and processes’ (Tansey 2007: 768–9). I complement the interviews with primary materials that include university documents and reports and with a large-N student and alumni survey that are the core of a quantitative article on the effects of hall cultures (Ricart-Huguet & Paluck 2023), rather than on their origins as is the case of this article.

Participant observation and interviews present a very similar picture of hall cultures. The one caveat is that some interviewees from halls without any cultural distinctiveness claimed their hall had a culture but were unable to describe it beyond using ‘culture’ as a buzzword, as I had suspected from my early fieldwork and other details discussed below.

I leverage the Makerere halls of residence as multiple case studies to present a narrative, both thick and analytical, that improves our understanding of culture when groups share the broader environment. Table I provides a description of halls with a distinct culture to guide the reader through the remainder of the article.

Table I. Hall descriptions

Note: The table lists hall attributes and the adjectives most commonly mentioned by students in surveys and interviews. Northcote Hall was closed in 1996 and reopened as Nsibirwa Hall in 1998. The hall changed dramatically, as I explain in the text. This may be why interviewees used both ‘noisy’ and ‘quiet’ when describing the hall (those mentioning ‘noisy’ likely have Northcote Hall in mind while those mentioning ‘quiet’ likely have Nsibirwa Hall in mind). The three halls without a distinct culture (Mitchell, Nkrumah and University Halls) are not listed because they lack common descriptors.

THE ORIGINS OF CULTURAL DIFFERENCES (1960S)

In the 1950s, ‘the Makerere Undergraduates’ Guild’ already organised ‘many student activities’ including a Graduation Ball, fund raising, and a newspaper that preceded The Makererean (MacPherson 1964: 84). However, alumni in the 1950s explicitly stated that there were no cultural differences between halls during their time. ‘There was no culture during our time’ (interview 04 with lawyer Mugenyi ’57). ‘We never had […] culture’ (interview 01 with Minister Nkangi ’50). This is reasonable because competition between halls was impossible; before Northcote Hall opened in 1952 there was only one hall because annual intake was below 200.

Until 1969, alumni ranked the halls of residence as part of their application to Makerere during their senior year in secondary school (interview 02 with Dean Kihuguru ’54, interview 07 with Vice-Chancellor Kirya ’62). Their first or second choice was always respected. Some would rank halls haphazardly (most were first-generation students with little knowledge of Makerere) and others in accordance with proximity to their department or information from older siblings.

Dean Dinwiddy and early differentiation at Northcote Hall

The first contingent event takes place at the dawn of Uganda's independence in 1962. Hugh Dinwiddy, Warden of Northcote Hall, also becomes the University Dean of Students. His concurrent appointments until 1969 led him to play a central role in the formation of Northcote Hall's culture.

Dean and Warden Dinwiddy: a principal-agent problem

Hugh Dinwiddy (1912–2009) was a Cambridge-educated scholar of English literature but also a famous cricketer and avid sportsman (Figure 2). Every year, Dean Dinwiddy would contact the headmasters of Uganda's secondary schools and even visit the schools to inquire about the best students and sportsmen in their graduating classes (interviews 23 with Ekudu, 02 with Kihuguru, 15 with Seemanobe). Warden Dinwiddy would encourage those students to rank Northcote as their first choice. Dean Dinwiddy would then ensure their choice was respected.

Figure 1. Extract from annual reports by halls submitted to the University Council.

Note: The top of the left page (Northcote Hall Annual Report 1963–4) lists the two roles of Hugh Dinwiddy as Warden and as Dean. The right page (Northcote Annual Report 1961–2) shows an impressive list of visitors and so-called ‘older statesmen’ (i.e. former Northcote Hall residents) in that year alone: Prime Minister Milton Obote, Governor of Uganda Walter Coutts, and professors from British and American universities.

Figure 2. Picture of former Dean of Students Hugh Dinwiddy.

Note: Hugh Dinwiddy, 19, at Radley College, Oxfordshire. Source: Radley College's blog.

Dinwiddy tried to convince sportsmen to switch halls even after they had arrived on campus: ‘I was residing in Livingstone. Then Dinwiddy came to convince me to go to Northcote, he offered very many offers but I rejected all of them and stayed in Livingstone’ (interview 15 with John Baptist Ssemanobe, former head of the Uganda Football Association).

The wardens (principals) might have preferred a ‘fair’ allocation rule but the Dean (agent) had his own priorities. His charisma and popularity among students might have facilitated his ‘machinations’. Dean Dinwiddy and his successor, Dean Kihuguru, were the only two Makerere administrators that were unanimously praised by interviewees (cf. Mills 2006; Sicherman 2006).

Northcote Hall: A rose by any other name would smell as sweet

Dean Dinwiddy's selection process affected halls differently. Its consequences for cultural formation soon became apparent. So-called ‘Northcoters’ concentrated the best students and sportsmen and devised the motto ‘We either win or they lose’. Their prominence led Northcote Hall to be the first hall to have a ‘cultural ideology and symbols’ (Appendix B provides a description of Northcote's culture). This led many Northcoters to join in their hall's pride even if they were not sportsmen, a phenomenon known in social psychology as ‘basking in reflected glory’ (Cialdini et al. 1976).

Northcoters began to differentiate themselves by developing a distinguishable group identity grounded on objectively better performance at interhall events. The Northcote section of the 1966 Makerere Annual Report explains: ‘We had claimed … the Inter-Hall Championship Trophy for all Inter-Hall games and sports for the fifth time in seven years, the Debating Competition, held for the first time between halls, and the Chess Competition for the fifth year running’ (de Bunsen 1967: 188). ‘[We were] very outgoing, known for sports, dancing and dominating all the time’ (interview 21 with MP Jack Sabiiti). In-group identity was already developed by the late 1960s: ‘I was emotionally attached to Northcote, it was the only hall for sports, culture, dance, social activities and literacy’ (interview 13 with Minister of Tourism Kamuntu ’68).^{11, 12}

CULTURAL FORMATION (1970S)

Several contingent events take place in Uganda in 1970–71: Makerere appoints a new Dean of Students (our second contingent event); Lumumba Hall and Africa Hall (a female hall) open their doors; and Commander Idi Amin leads a successful coup d’état. Collectively, they explain cultural formation and differentiation between halls, sometimes in ways that are far from obvious.

Dean Kihuguru randomises to eliminate intergroup differences

After Dinwiddy retired in 1969, the University Council appointed Makerere alumnus and educator George Kihuguru as the next Dean of Students but not as Warden of Northcote (Figure 3).

The Dean then in 1960s to 1969 was the Warden of Northcote which brought a lot of politics in the allocation. He would go to the different high schools and he would encourage only good sportsmen to his hall. This made Northcote vibrant and it was winning trophies and that's why they developed a saying that ‘we either win or they lose’. This was unfair for the other halls and [brought] a lot of dissatisfaction. When I took up the office in 1970, I changed the system from making choices to random. (Interview 02 with Dean Kihuguru)

Kihuguru unilaterally decided to allocate students via alphabetical randomisation to eliminate differences between halls. Kihuguru would assign the first student in the alphabetical list to hall 1, student 2 to hall 2, student n to hall n, student n + 1 to hall 1, etc. until the end of the admissions list. Kihuguru conducted the process by himself in pen and paper or, as student intake increased over time, directly supervised it.

Figure 3. Interview with former Dean of Students George Kihuguru (1970–1995)

Note: From left to right, Edwin Mayoki (research team leader), Joan Ricart-Huguet and Dean George Kihuguru.

Kihuguru's alphabetically random process recognised that surnames from ethnolinguistic groups tend to cluster around certain letters of the alphabet. Thus, assigning surnames A–H to hall 1, I–P to hall 2, etc. would have led to clustering while Kihuguru wished to balance halls for ethnicity, athletic prowess and any other dimension (interview 37 with Kayiggya). A few alumni resided in halls other than the one they were randomly assigned for various reasons, from accommodation for disabilities to successful pressure of some elite parents on the University (interview 29 with Deputy Dean Kateega). However, survey data suggests that only 3% of the 1970–2000 alumni changed hall, including those with legitimate reasons (Ricart-Huguet & Paluck 2023). (An oft-told story is that the son of dictator Idi Amin was the only student with a single room and a TV.) In sum, the system was random for almost everyone (interview 33 with General Kulaigye).

We cannot ascertain whether random allocation has contributed to the reduction in ethnic and regional conflict in Uganda since the mid-1980s. However, we do know that exposure to out-groups in key formative years reduces inter-ethnic discrimination (Scacco & Warren 2018) and that the stakes were high because almost all senior civil servants, politicians and high-ranking military officers are Makerere alumni. We also know that, in spite of randomisation, differences between halls actually increased with the emergence of two new cultures in the 1970s at Lumumba and Livingstone halls. Why?

Lumumba Hall: ‘a hall of rejects’ since the time of Idi Amin

Lumumba residents were activists ‘from the beginning’, said Stephen Dagada, one of the first leaders of Lumumba Hall (interview 22). This is difficult to square with random assignment. While Dean Kihuguru assigned freshmen randomly, he asked wardens from other halls to provide him with a list of sophomores and juniors to populate Lumumba. According to his successor, Dean Ekudu (interview 23), the result was a concentration of activists and ‘undesirables’:

The students that were moved from other halls to Lumumba were basically naughty students that were not liked by the wardens of other halls. When the rejects went to Lumumba, they became rivals of those who were in Northcote. They would have this rivalry during sports and wherever they would meet.

This initial concentration of activists provided the basis for residents to develop a group/hall identity and combine it with their broader Makerere identity, consistent with ODT. Dagada (interview 22) argued that he and others chose the name out of a need for recognition:

Lumumba [male hall] and Africa [female hall] started the same year, 1971, as hall X and hall Y, respectively. [We debated] in the Senior Common Room and got to a consensus for Lumumba Hall which was then forwarded to the University Council by the hall leadership and was therefore adopted. I was the Minister of Information [of Lumumba Hall] and was entrusted to invite the Congolese Ambassador to officiate the end of year party and we officially announced the name. … We wanted to recognise a hero who was killed in cold blood trying to fight for Pan-Africanism. … We needed to be recognised as the hall had just been opened with us as the pioneers.

The 1970s were not conducive to activism in Uganda, however. Commander Idi Amin gained power in 1971 via a coup d’état and proceeded to brutally stifle dissent during his tenure (1971–1979): ‘The university became a haven for military intelligence’ and the ‘Vice-Chancellor was brutally murdered in 1972, allegedly by government operatives, accused of being an agent for former President Milton Obote. … Clandestine student politics remained, but based mainly in halls of residence’ (Byaruhanga 2006: 63). ‘The students’ surreptitious plans to overthrow Idi Amin's military government were designed in Lumumba Hall’ (Byaruhanga 2006: 69). A freshman activist at Lumumba Hall in 1976 recalls Operation Entebbe, when a hijacked plane by Palestinians with 106 hostages was rescued by Israeli commandos:

The day of the crisis in Entebbe Airport, Amin had intelligence on campus. Some students agitated to remove Amin that same night. Some military officers came to encourage us but they were in reality pro-Amin, it was a trick to expose activist students. … Soldiers came to Lumumba as they knew havoc came from Lumumba to wipe us. … We ran away and they arrested some people. (Interview 26 with Barnabas Nawangwe)

Professor Nawangwe went into exile in 1976 and only returned in 1980 after Idi Amin had been deposed. He became the Vice-Chancellor of Makerere in 2018.

Livingstone Hall's culture: a quest for distinctiveness

Livingstone Hall opened in 1959 and ‘was named by the University Council because the students had little say in making decisions during that time [under British rule]’ (interview 06 with Minister of Finance Rukikaire). It ‘was named after the explorer and they called themselves the gentlemen’ (interview 23 with Dean Ekudu). ‘[We were] not radicals, not noisy but gentlemen’ (interview 15 with Ssemanobe). General Kulaigye (interview 33), who later fought against the Lord's Resistance Army, was not keen on Livingstone at first: ‘I wanted to reside in Lumumba because of its name, because of Pan-Africanism’ but Livingstone Hall ‘taught me to handle things calmly’. Minister Rukikaire even stated: ‘I am still a focused person and gentle because of what my hall made me.’

Livingstone alumni all agree on cultural descriptors such as ‘gentle, calm, and modest’ (interview 14 with Dr Abbas) but ‘can't really trace why Livingstone [residents] were called gentlemen’ (Rukikaire). The early oral history of Livingstone Hall is more ambiguous than that of Northcote or Lumumba halls. Dr Abbas and MP Sabiiti (interview 21) suggest that cultural formation at Livingstone was also the result of a quest for distinctiveness and contingent events that biased their early composition – although this time likely aided by the eponym, ‘Dr Livingstone’, which provided values that stood in contrast to those of Northcote and of Lumumba.

Livingstone's biased composition in the 1960s was likely due to medical students selecting it because of its proximity to the University Hospital in Mulago Hill. Intergroup rivalry and perhaps a sense of inferiority toward Northcote Hall may have spearheaded the development of an almost opposite identity: reserved, quiet and gentle instead of outgoing, noisy and aggressive. The incipient culture of Livingstone therefore satisfied their quest for distinctiveness. As Minister Kamuntu (interview 13) put it, the culture of Livingstone developed ‘as a reaction to Northcote's dominance’. For all their losses in sports, Livingstone fared better in coopting the new female hall, Africa Hall, located between ‘gentle’ Livingstone and ‘rough’ Northcote and courted by both – a key reason for the rivalry between those two male halls. In the 1970s, Africa and Livingstone Halls formalised the ‘Afrostone solidarity’, leaving Northcote without a female partner hall.

Other male halls

While understanding cultural formation and persistence is central to this article, it is also interesting to consider why three other male halls (Mitchell, Nkrumah and University Halls) did not develop a culture. The characterisations of these halls by former residents are discordant. University Hall alumni characterise their hall alternatively as ‘carefree’, ‘serious’, ‘hard-working’, ‘not sure’ and ‘gentlemen’ (interviews 08 with MP Nabwiso and 016 with Professor Ntozi). At times, respondents stated that their hall ‘didn't have a culture’ (interview 20 with Ambassador Baba) and that ‘there was not much that distinguished [my] University Hall from other halls’ (interview 07 with Vice-Chancellor Kirya). Analogously, Mitchell alumni see their hall as ‘respectful’ (interview 05 with Rubahaiyo), ‘[a hall for] free people, we did what we wanted’ (03 with Kyemba), and ‘gentle’ (41 with Matovu). The Makerere Journal states that ‘such ‘Culture’ [at Mitchell Hall] is difficult to define’ (de Bunsen 1967: 188). Most would state not to know why these halls failed to develop a culture (e.g. interview 28 with MP Kassiano). A few answered with tautologies: ‘some halls were not so active and hence had no culture’ (interview 21 with MP Sabiiti).

Why did these three halls not develop a culture? All halls share the same institutions at the campus and even hall level (i.e. all halls have a per capita budget) and all six male halls are located within 0.6 square miles (Figure A.13). ‘Every hall wants to have an identity’, said Vice-Chancellor Nawangwe (interview 26), but only some succeed. These negative cases are useful to show that the need for distinctiveness posited by ODT (Brewer 1991) is probably necessary but not sufficient.

The three halls that developed a culture experienced some contingent event that biased their composition at a critical juncture (Dinwiddy's bias for Northcote in the early 1960s; Kihuguru's ‘mistake’ in 1970 that allowed student ‘rejects’ to concentrate in Lumumba in 1970; and, arguably, the proximity of Livingstone Hall to both the University Hospital and to rival Northcote Hall). None of the other three halls experienced a contingent event.

‘Other halls tried to copy Northcote's culture’ (interview 34 with Dr Lubaale). Northcote's immediate neighbour hall was named Nkrumah Hall in 1968 at the insistence of some residents. Chiefly among them was Uganda's current Prime Minister Ruhakana Rugunda (interview 18). They erected a statue of Ghanaian President Kwame Nkrumah and even referred to it as ‘Osagyefo’ (‘redeemer’ in the Akan language), a powerful statement by hall leaders who espoused the values of solidarity and Pan-Africanism. One might expect the hall to become a paragon of pan-Africanism, since it was the first hall to be named after an African person. Far from it, Nkrumah Hall alumni, including PM Rugunda, were unable to distil values or behaviours particular to their hall.¹³ The need for distinctiveness may have been just as high at Nkrumah Hall but, without an event that biased the hall's initial composition, it was unable to form a culture.

Female halls

The discussion has centred on male halls because the history of Makerere is heavily gendered. Two female halls, Africa and Mary Stuart, developed cultures that were analogous to those of their neighbouring male halls (Table I), thereby following rather than leading from a cultural standpoint.

Africa Hall formed a so-called ‘solidarity’ (a close relationship) with neighbouring Livingstone Hall in the 1970s called Afrostone. Africa Hall residents became known as ‘Ladies’. Their culture became the gendered transposition of Livingstone's. Lumumba Hall and neighbouring Mary Stuart Hall, known as Box because of its shape (Mills 2006: 255), established the Lumbox solidarity also in the 1970s. Mary Stuart Hall's motto ‘Train a woman, a nation trained’ (interview 031 with Kabenge) was fitting because it matched the activism of Lumumba Hall and because Mary Stuart, an educator and Officer of the British Empire, played ‘a tremendous role in the transformation of social life in the country’ and in the ‘promotion of women's education’ (New Vision 2002).

Several reasons probably explain why male halls led the way in cultural differentiation and female halls followed. One is that Makerere was a male-only university until 1945 (Mills 2006: 254; Sicherman 2006). Another is that women suffered discrimination both from Ugandan society and from the colonial state, as evidenced by the exclusive focus on male higher education early on. Sexism has long been a feature of college campuses in Uganda and elsewhere, sometimes blatantly and sometimes more insidiously (Moffatt 1989). A third and more specific reason is that women had to worry about their physical security before concerning themselves with hall culture. Mills (2006) recounts that, while living on campus, he ‘often heard male students’ shouting ‘Malayaa’ (prostitute in Swahili) at women. Specioza Kazibwe, former Vice President of Uganda (1994–2003), organised courses in martial arts in the 1970s as chairperson of Mary Stuart Hall ‘so that women could ward off assaults’ (Sicherman 2006: 121).

CULTURAL PERSISTENCE (AND CHANGE) SINCE THE 1970S

Primary documents, quantitative survey data, interviews with recent alumni and months of participant observation show significant cultural persistence (Appendices B and C provide further materials). Figures 4 and 5 use data from the alumni survey to show that some differences in hall culture are reflected in behaviours, such as activism and quietness in class. An email conversation with a 2016 graduate conveys the same idea qualitatively:

Shadrack: Anyways what was your study about? Never really got it!

Joan: […] if you Shadrack had been assigned to Lumumba instead of Livingstone, do you think anything would be different? If so, what and why?

Shadrack: ‘haha that's true. I think a lot would be different. I would be more confident being rowdy. A bit more dramatic as a person and more involved with university demonstrations. I think with a bit more pride or I would experience some culture shock finding the behaviour of the Lumumba boys a bit too extreme. Being in Livingstone enabled me [to] blossom a bit as a rowdy person because we are so few who are that energetic but also instilled this pride in me of being calm, collected and rational. I like the peace and order of Livingstone now and I can't really imagine myself in any other hall. I guess therefore I wouldn't mind being in Lumumba because of my previous background and would have probably grown in to a Lumumba stereotype but we can never be too sure though I must add that in my first year only did I identify with the ‘way’ of the Lumumbists but right now i believe the chaos can be avoided.’ (Email conversation with Shadrack Manano, former Deputy Disciplinary Minister at Livingstone Hall, June 2015)

Two contemporary primary documents also show that this stark contrast remains. Lumumbists are admonished by the Dean of Students for engaging in inappropriate behaviour (Figure 6), while Livingstone Gentlemen distribute a Code of Conduct to incoming freshmen (Figures 7 and A.3).

Figure 4. Activism over time: demonstrations joined by alumni respondents while on campus.

Note: The graph shows levels of activism by hall over time using a five-year moving average. ‘(M)’ stands for male halls and ‘(F)’ for female hall. Lumumbists and Northcoters demonstrate more than Livingstone Gentlemen throughout the period. Differences between either and Livingstone respondents are statistically significant (p < 0.01). Mary Stuart Boxers demonstrate more than Africa Ladies but the difference is not statistically significant.

Figure 5. Quietness over time: class behaviour by alumni respondents while on campus.

Note: The graph shows class behaviour by hall over time using a five-year moving average. ‘(M)’ stands for male halls and ‘(F)’ for female hall. Livingstone Hall is quieter than Lumumba and Northcote Halls (p < 0.01), while Africa Hall is quieter than Mary Stuart Hall (p < 0.01). The differences are substantively small in a 0–3 scale where higher means quieter.

Figure 6. Notice of ‘Naked Mile’ during the Culture Week (2015) by members of Lumumba Hall.

Note: The Dean of Students notifies Lumbox students of inappropriate behavior. Interestingly, the ‘brashness’ is consistent with earlier decades but the cultural practice of the ‘Naked Mile’ is a cultural innovation showing some change occurring alongside broader cultural persistence (reproduced with permission from the 2014-15 Culture Minister of Mary Stuart Hall).

Figure 7. Code of Conduct of Livingstone Hall (2014–2015, pages 3–4).

Note: Pages from the Livingstone Hall Code of Conduct (2014–2015), publicly distributed to its residents.

Mechanisms of cultural persistence

The two factors theorised earlier, a low enough annual population replacement rate and a hall's political hierarchy, repeatedly arise as central to cultural reproduction via in-group socialisation. The share of new freshmen is around 25% because most degrees are four years (or three or three plus one in the past, following the British system). A 25–30% yearly replacement rate is high and a homogeneous minority could a priori challenge the cultural status quo. Just consider a 25–30% annual immigration rate from country A into country B. However, randomisation prevents this homogeneity of newcomers and thus coordination on an alternative cultural equilibrium by making this 25–30% as diverse as Makerere itself.

The second mechanism of cultural reproduction is the hall political hierarchy, composed of elected student governments. Each year since the 1960s, students elect a hall cabinet composed solely of students. Students can run for a dozen positions that include Chairman/Chairlady, Speaker, Minister of Interior and Minister of Culture. Elections are often contested for these top positions and campaigns can be heated.¹⁴

Hall cabinet ministers hold multiple responsibilities, notably socialising freshmen into the hall's culture. Some activities have changed over time in content but their goal – socialisation and adaption into the hall's culture – remains. In the 1960s, cultural activities included choral societies, a travelling theatre group, regular dances, Divali celebrations (until Idi Amin expelled the South Asian minority in 1972) and poetry competitions (de Bunsen 1967: 215–16). ‘Most students [at Northcote] were involved but mostly the fresher students would be trained’ (interview 01 with Nkanji). Other events even involved hall alumni: ‘Every year, every hall had a reunion day. During this reunion, former students who are referred to as ‘ancestors’ would come together with the current students and participate in bull-roasting’ (interview 33 Kulaigye, also 34 Lubaale).

Hall ministers today continue to promote freshmen identification with their hall's culture. Orientation Week activities are aimed at strengthening freshmen's social ties with the hall. Morning jogs, organised by each hall leadership, have long been important for socialisation (Figure A.9). Other hall social events take place in the evening, such as porridge nights. Culture Week, when students celebrate the culture of their hall, closes the academic year (Figure A.4).

Finally, it is important to note that cultural practices may not be accepted by all. Leaders can sanction social deviants through a mixture of social pressure and social exclusion, as in many other societies (Henrich 2021). Hall ministers can sanction deviants to foster norm adoption (Figure 8). I witnessed the hall ministers knocking insistently (and loudly, in the case of Lumumba) on the doors of recalcitrant first-year students who wanted to sleep in rather than join the 6 am morning jogs that I joined on occasion during Orientation Week.

Figure 8. Socialisation during Orientation Week.

Note: Article in The Observer, 11 August 2010.

Cultural change and death: the case of Northcote Hall

Thus far I emphasised cultural persistence, yet cultures may change even in the span of 25 years. Contact with Americans during the Second World War ‘culturally transformed’ New Guineans from the island of Manus (Mead 1956). At Makerere, Northcote Hall presents an interesting case of cultural change (1970s–1980s) and eventually of cultural death (1990s). Just as low enough replacement rates are a mechanism that facilitates cultural reproduction, a high enough replacement rate, illustrated by the case of Northcote Hall, led to the cultural death of the hall – i.e. to the disappearance of its system of meaning and associated behavioural practices.

Randomisation ended Northcote Hall's dominance on campus, upon which they had built a culture that emphasised pride and social cohesion. It forced Northcoters to rethink their identity. They progressively developed a militaristic culture absent in the 1960s. ‘The culture [became] militant. Students behaved like soldiers nicknaming themselves as general, field marshal, etc.’ (interview 25 with Nkanji) and ‘students would be taken through military drills’ (interview 23 with Ekudu). Northcote chairmen Kassiano (1977–78) and Lubaale (1987–88) explain:

As soon as a student would come to Northcote, that student would be initiated into the Northcote culture. … The drum was a unifying factor as the students danced to the drum-beats and the songs in the courtyard which was referred to as the parliament of Northcote. During football matches, the hall had a state car, the enemies of the hall call it tractor [it actually was a tractor], in which the hall chairman would be driven to the match as the commander in chief of the State Supreme Revolutionary Command Council (SSRCC). (Interview 28 with MP Wadri Kassiano)

The students could be inaugurated into the culture and would be woken up to do military training for about 5 days. … Northcote students were known to be the military men and Northcote was called State Supreme Revolutionary Council … In Northcote, if a student had no rank and did anything on Gongom [a statue and cultural symbol at Lumumba], that student would be promoted. … The Defense Council of Northcote would meet at midnight for disciplinary meetings. (Interview 34 with Dr Lubaale)

Joan: Where did you get the state car or tractor from?

Lubaale: We would hire it from the Estates Department. Northcoters would have their drums, the tractor, as if people were going to war. … The State Car [tractor] would have the officers seated in front. And then everybody [would be] in camouflage [uniforms].

Joan: But were these uniforms real?

Lubaale: Yes, like mine was real [others were not].

Joan: From the UPDF [Ugandan Army]?

Lubaale: Actually, it was from the NRA [National Resistance Army], there was no UPDF at that time [1988–91].

Joan: How did you get it?

Lubaale: From my colleague who was in the State House. He was a bodyguard for Museveni [President of Uganda]. (Interview 34 with Lubaale)

These lived experiences are very consistent with Mills (2006: 247), who provides his first-hand account of being woken up in 1995 by Northcote's ‘young men in camouflage trousers’ who were ‘being inducted into its quasi-militaristic culture, complete with uniforms, marching songs and passing-out parades’.

It is hard to overstate the role of cultural symbols and behaviours in providing meaning, above and beyond institutions, after reading these passages. Northcoters’ self-conception as ‘statesmen’ may have been key for the adoption of a militaristic culture. When asked, Minister of Tourism Kamuntu (interview 13) replied: ‘What state doesn't have a military?’ Change seems to have been gradual, akin to cultural formation in Livingstone Hall, with military titles and behaviours being adopted in the late 1960s and early 1970s. ‘Northcote used to be rivalled by Lumumba Hall’ (interview 28 with Kassiano) since the 1970s, so a military identity afforded an image of strength towards other halls and of cultural distinctiveness, consistent with ODT.

By the 1990s, Kassiano admitted that ‘the culture had degenerated into militarism, which was evident in the military attires worn by the hall students’. Northcote's rivalry with neighbouring Livingstone Hall, lingering since their lost contest to build ties with the residents of Africa Hall in the 1970s, led them to ‘add ground pepper and glass [shards] to the food at [an Afrostone solidarity] dinner’ in 1995 (interview 23 with Ekudu). ‘The horror of pouring grounded glass in the food, which the country came to know about’, led to the shutdown of Northcote Hall (interviews 29 with Kateega, also 43 with Lukabala). Dr Lubaale's version is somewhat different:

Joan: One more thing. [The University] closed down Northcote 10 years later [1995]. Since you were on campus [as a Lecturer], maybe you can explain to us what happened?

Lubaale: The closure of Northcote … eventually culture evolved and other people say it was going to extremes. … There was another time that same year when it was Afrostone day, when you come together and you celebrate. There was a dance in Africa in the dining. Somebody entered with pepper and put it on the floor. People dance and they start [mimics sneezing]. That one was referred to as ‘chemical warfare’. The one of the faeces was referred to as ‘biological warfare’. Then everyone gets out, they clean the place, they enter, and they make sure no Northcoter can enter. … After that, there was an allegation that Northcoters went and put glass in the food in Livingstone. It is an allegation because my cousin, who teaches in Internal Medicine, was at Livingstone. … He was able to eat the food, he didn't die or go the hospital, but they claim Northcoters had put glass in the food. So: nobody produced the sample of the food, no proof, but somebody just made an allegation, people worked on rumour, and that was all. So that was how Northcote was closed.

Edwin: But was the pepper thing true?

Lubaale: The one for dancing? Ah, that one is true.

Joan: And the biological warfare is true as well?

Lubaale: Yes, that one is true.

Joan: You're sure?

Lubaale: That one I'm sure, I was around. (Interview 34 with Dr Lubaale)

Dr Lubaale, a committed Northcoter, denies that Northcoters put glass in the food but the story was confirmed by the other interviewees above and by Mills (2006: 248). As a result of the incident, Dean Ekudu and Vice-Chancellor Kirya decided to close the hall and expel all its students, a decision that put Ekudu and Kirya in physical danger (interview 07 with Kirya).

Northcote reopened in 1997 with new students and a new name. The University Council named it after ‘Nsibirwa, the former Prime Minister of the Buganda Kingdom in Uganda, because he offered his land to Makerere university’ (interview 23 with Ekudu). Some new residents, aware of the hall's history, tried to revive the culture, even ‘renaming’ Nsibirwa as Northcote State Is Brilliant in Revolutionary War Affairs (N.S.I.B.I.R.W.A) (Nassar 2015). However, the student replacement between 1995 and 1997 proved too sharp of a cultural discontinuity, the wishes of prominent Northcote alumni notwithstanding. ‘The hall culture has changed greatly because it was so militaristic until when the hall was closed for misbehaviour’ (interview 34 with Lubaale). ‘From the time they expelled students and closed the hall, it has never been the same again. As we speak [2016], they do not have a distinct culture’ (interview 01 with Nkanji).¹⁵

CONCLUSION

This article examines the case of Makerere University's resident halls to show that cultural differences between groups can form and persist even when all groups share the same campus environment, live in close proximity, and new members do not differ between groups ex ante because of random assignment. Compared with existing ethnographies in other university settings (Moffatt 1989; Becker et al. 1992), the multiple case studies (halls) within Makerere present a unique research design to understand cultural formation and intergroup competition because they allow for a more controlled comparison than most natural experiments. The patterns uncovered here are probably more general, however: they are likely analogous to relations between groups in other universities, between ministries in a government, or between departments in any organisation, all of which may have somewhat different cultures.

I argue that biased composition of groups ab initio, resulting from contingent events (critical junctures), and a quest for optimal distinctiveness (ODT) between groups led to the formation of cultures at Makerere's halls. I also attempt to explain why, even if ‘every hall wants to have an identity’, as Vice-Chancellor Nawangwe put it, alumni in three of the six male halls were unable to distil any consistent hall values or behaviours. I argue that these negative cases did not experience any contingent event that biased their composition before randomisation, suggesting that the need for distinctiveness is insufficient to explain cultural formation.

How cultures emerge and persist is a question of enduring importance. This article provides one answer, by no means unique, that complements existing accounts (Swidler 1986; Young 1996; DiMaggio 1997). My emphasis on social identity does not reject instrumental rationality. A stark dichotomy between Durkheim's homo sociologicus, whose behaviour is dictated by social norms, and Adam Smith's homo economicus, whose behaviour is dictated by rationality, is a false dichotomy (Elster 1989). Evolutionary approaches (Axelrod 1986; Matsui 1996) support the view that both instrumental and social motivations explain social norm adoption and cultural adaptation.

Finally, the findings speak to ‘micro’ social psychological theories of identity formation as well as to ‘macro’ political economy theories of ethnic and cultural diversity by examining the conditions under which identity and cultural differences emerge to begin with. Further, my findings suggest that when there is no group identity that provides distinctiveness, individuals may not only select it among the existing menu of options (Posner 2005) but may even be able to form it.