individualistic and collectivistic orientations: examining

Individualistic and Collectivistic Orientations:Examining the Relationship between

Ethnicity and National Attachment in Ghana

Kofi Takyi Asante*University of Ghana

AbstractEthnicity is regarded as the greatest threat to national attachment by both politi-cians and scholars. However, ethnicity is only one of the many forms of identifi-cation which could potentially clash with national attachment. This studytherefore examines the relationship between ethnicity and national attachmentby asking a general question: what is the impact of alternative group loyaltieson national attachment? To answer this question, I develop a measure of nationalattachment drawing on several sentiments oriented towards the state. Using a sur-vey of 996 university students, I find varying degrees of intensity of theselected identities. Specifically, while descriptive analysis supports recent reportsof the declining salience of ethnicity in Ghana, inferential analysis contradictstheoretical expectations that the increasing salience of ethnicity would negativelyaffect national attachment. Conceptually, it is possible to map the various identitiesonto a collectivistic-individualistic scale. Individualistic orientations underminenational attachment, while collectivistic orientations boost it. I argue that ratherthan being contradictory impulses, ethnicity and national attachment are bothunderlain by the same collectivistic orientation, pointing to the importance of socialrootedness. I deploy qualitative and historical data to give substance and texture tothese findings.

Introduction

The question of national unity has exercised the minds of researchers since nationalindependence appeared on the political horizon across Africa. Modernization the-orists in the 1950s and 1960s proposed an evolutionary model of development andnational integration in which newly independent countries would follow the pathsbeaten by European countries. As the dreams of political and economic develop-ment turned to nightmare in the 1970s, another set of scholars turned to accounts

* Kofi Takyi Asante is a Research Fellow at the Institute of Statistical, Social and Eco-nomic Research (ISSER) at the University of Ghana, Legon, Accra. His research interestsinclude state formation, the sociology of colonialism, citizenship, and economic sociology.

2

Kofi Takyi Asante: Individualistic and Collectivistic Orientations

of historical exploitation to explain the continuing salience of ethnicity. Since theso-called third wave of democratization in the late 1980s, ethnicity has again comeunder the spotlight as electoral competition created divisive politics across the de-mocratizing world.

Ethnicity, or its less polite synonym ‘tribalism’, is an elusive concept. It was ini-tially used among early anthropologists ‘as a heuristic category […] with only in-tuitive meanings attached to it’, and later attempts at a more concrete definitionproved to be ‘not enlightening’ (Ekeh 1990:662). As used in the literature, ethnic-ity refers to feelings of attachment to an ‘imagined community’ at the sub-nationallevel, a community often defined in terms of common language, history, culture,or political systems (for a review of this large body of work, see Lentz 1995).Strong ethnic attachments may or may not come with discriminatory attitudes to-wards outsiders, and a critical unpacking of instances of ethnic animosity wouldreveal underlying factors that point away from questions of belonging and towardsstruggles over power or resources. Analysing instances of ethnic-based violenceinflicted on Eastern Nigerians by Northern Nigerians, for example, led RichardSklar (1967:7) to conclude that the ‘causes of tribal violence in Africa, like thoseof mob violence elsewhere, might be traced to the prevailing conditions of pov-erty, insecurity, and the lack of opportunity for satisfying employment’. Indeed,most identity-based conflicts occur among co-ethnics and have to do with strug-gles over what it means to be a good member of a particular ethnic community(Lonsdale 1994).

This paper is a first stab at a much broader project aimed at theorizing citizen-ship in Ghana. For such an endeavour, perhaps the most apposite starting point isethnicity, because a strong sense of ethnic attachment or tribalism has been‘widely supposed to be the most formidable barrier to national unity in Africa’(Sklar 1967). My purpose in this paper is simple; I seek to measure the strengthof national attachment in Ghana, and the effect of ethnicity on this attachment.To do this, I frame the problem more generally, asking: what is the impact of al-ternative group loyalties on national attachment? To answer this question, I selectfive forms of identification whose salience can be theorized as directly in compe-tition with the state for the loyalties of citizens.

I deploy qualitative and historical data to give substance and texture to the sta-tistical findings. In particular, I emphasize the importance of a historical perspec-tive on citizenship and identification. This is important because a key weakness ofscholarship on national belonging in Africa has been the ‘tend[ency] to fore-shorten historical time and to privilege the present at the expense of the past’(Hunter 2016b:1). Examining the ways in which ‘basic assumptions about moralrights and obligations’ (Owusu 1989:373; Lonsdale 1994) are embedded in localcultures helps us to understand the ways in which national belonging is imaginedand performed.

In what follows, I review the literature on ethnicity, focusing specifically on itstheorized relationship with national identification. I then turn to the main empiri-cal data: a survey of university students conducted between March and April 2017in Ghana. Results from descriptive analysis support recent findings about the de-clining salience of ethnicity in Ghana. However, the effect of ethnicity on national

Studies in Ethnicity and Nationalism: Vol. 20, No. 1, 2020

3

attachment is counterintuitive. Conceptually, I find that alternative group identifi-cations can be mapped onto either individualistic or collectivistic orientations.Statistical tests show that individualistic orientations undermine national attach-ment, while collectivistic orientations boost it. I use qualitative and archival datato expatiate on the statistical results and to evaluate prevailing theories of ethnicand national identification.

Ethnicity and the Nation-State in West Africa

At independence, Sub-Saharan African leaders felt the heavy weight of the re-sponsibilities they were inheriting from the departing colonial administrators.Radical economic transformation was necessary to ‘catch up’ with the West.But they feared that their state-building agenda risked getting derailed by the per-sistence of strong sub-national attachments. They therefore needed to transformtheir people from parochial ‘tribesmen’ into cosmopolitan citizens, and to enlargetheir vistas to accommodate the diversity of peoples who now composed thesenew nations. At the All-African People’s Conference held from 5 to 13 December1958 in Accra, it was declared that ‘tribalism’ was an ‘obstacle’ lying in the wayof ‘the unity […] the political evolution [… and] the rapid liberation of Africa’(quoted in Sklar 1960:493).

This anxiety was also shared by scholars. The literature can be broadly classi-fied into modernization and colonial legacy explanations (Eifert et al. 2010; Rob-inson 2014). While modernization theorists argued that with economic and socialdevelopment, narrow ethnic sentiments would eventually wither away, theirdependency-theory counterparts attributed the intensity of sub-national alle-giances to the legacy of colonialism.

Modernization theory builds on theoretical insights from structural functional-ism. Talcott Parsons, the foremost structural functionalist, details the conceptualdistinctions between traditional and modern societies in his famous pattern vari-ables, a set of five dichotomous variables ‘focused on the relational aspect ofthe role structure of the social system’ (Parsons 1991 [1951]:43). The binary op-positions of the pattern variables projected an image of traditional societies as pa-rochial, in contrast to the cosmopolitanism of modern societies. Modernizationwas going to deliver traditional peoples from the shackles of their own irrationalaffections and prejudices.

Modernization scholars found newly independent states to be crucial arenas totest their theories and to offer prescriptions for welding together into unified statestheir motley collections of ethnic groups. What was needed for national unity anddevelopment was for these countries to lose their traditional orientations. WaltRostow (1959:7), for instance, believed that the development of a national spiritwas an important precondition for take-off, because such ‘a definitive politicaltransformation’ can ‘[harness these] national energies, talents, and resourcesaround the concrete tasks of economic growth’. Like Rostow, modernizationtheorists generally agreed that economic development would only be possible ifpre-existing social norms and practices were overhauled to bring them closer tothe Western standard. This rested on a circular argument in which ‘the values,


4

institutions, and patterns of action of traditional society were both an expressionand a cause of underdevelopment and constitute the main obstacles in the wayof modernization’ (Valenzuela and Valenzuela 1978:538).

Modernization theorists believed that with economic development, ethnic sen-timents would fizzle out. Urbanization, industrialization, and the widening ofwage employment were going to be crucial elements in this process. Wage em-ployment would reduce reliance on ethnic networks for access to agricultural land.Together with industrialization, these would pull villagers into towns and cities,forcing them to interact with people of all ethnicities, and this in turn would leadto greater tolerance and the development of a national orientation (Robin-son 2014). Moreover, investment in massive developmental projects and the ex-pansion of social welfare by the state would call forth from within the hearts ofthe citizenry affection and loyalty. For their part, nationalist politicians alsoembarked on deliberate campaigns and projects to generate feelings of attachmentto the state. Just a few decades after independence, however, students of West Af-rican states observed mixed results. For instance, Abner Cohen (1969) found thatin some instances, urbanization weakened ethnic attachment, while in others, eth-nic sentiments were intensified; processes he referred to as detribalization and re-tribalization, respectively. Sometimes, there was even the phenomenon ofsupertribalization, where related ethnic groups combined in urban settings toform mass umbrella associations (Rouch 1956).

Not all scholars of ethnicity in West Africa were worried. Richard Sklar’s pro-vocatively titled paper, ‘The Contribution of Tribalism to Nationalism in WesternNigeria’, is an example. He argues that without the pan-tribalism which gave ‘im-petus to the growth of mass political parties […] The British Government wouldnot, in principle, have transferred power’ to the leaders of the nationalist struggle(Sklar 1960:493, 502). The young Immanuel Wallerstein (1960) argues similarlythat ethnic mobilization was a necessary step towards the ultimate destination ofexclusive national attachment. Invoking the pattern variables – central to the mod-ernization thesis which he would later come to oppose – he argues thattribalization was a positive development because it reduced the salience of famil-ial ties, which he considered more damaging to national attachment. Echoing Ed-ward Banfield’s (1958) then prevailing notion of ‘amoral familism’, he argues thatfamily attachments compel one to focus on a very narrow sphere of life, whereasethnic allegiances orient the affections towards a much larger, more diffuse, moreabstract social sphere, ultimately embracing the entire nation.1 From this stand-point, ethnic politics was never a threat, because it mobilized individuals to partic-ipate in national life.

In contrast to modernization theory, colonial legacy explanations lay the blameon colonialism. Arbitrary colonial boundaries are said to have laid the foundationfor intractable inter-ethnic tensions by cobbling together diverse and sometimesantagonistic ethnic groups, resulting in suspicion and tensions in the new states(Young 1979). Furthermore, the cynical manipulation of ethnic identities by colo-nial administrators, in their quest to entrench colonial power, exacerbated ethnictensions. This argument draws on the famous ‘invention of tradition’ thesis(Hobsbawm and Ranger 1983). In their attempts to bolster the authority of


5

traditional leaders with whom they collaborated, colonial regimes sought toclearly define customs and identities, in the process freezing identities whichhad been marked by fluidity in the past; a process further reinforced by nationalistpoliticians in the ensuing power struggles at the dawn of independence(Ekeh 1975). Unlike modernization theory, which locates the problem in culture,this historical institutionalist view provides a path-dependent account whichlocates the problem in the specific developmental trajectory on which colonialismhad set these countries (Mamdani 1996).

Anthropologists and historians have since moved beyond top-down notions of‘invention’ to argue that transformation in ethnic identities was an endeavourwhich could not have been dominated by colonial officials. Colonial initiativeswere challenged or abetted by different actors in colonial societies, as they strug-gled to secure their perceived interests (Berry 1992; Lentz 1995; Spear 2003). Ter-ence Ranger, author of the Africa chapter in The Invention of Tradition (1983), hassince recanted his position, and has suggested that scholars should shift from afocus on ‘invention’ to one of ‘imagination’ in acknowledgement of the realiza-tion that official colonial attempts were simply one out of many endeavours toremake ethnic identities; and that, in contrast to ‘invention’, imagination acknowl-edges the contingencies that characterized the outcomes of official attempts(Ranger 1993).

Scholars have also moved beyond a static view of identities. Instead, they seethem as constantly shifting principles by which people relate to one another,and which are perpetually adjusted in response to given situations. In the GoldCoast, although plans by colonial officials to imprint fixity and stability on mercu-rial social conditions ‘could nowhere be translated into practice’, local chiefs triedto ‘appropriate’ these official colonial idioms in a bid to entrench their own power(Lentz 2000:116–7). The shifting meanings of ethnicity have continued into thepostcolonial period. In her study on politics and ethnicity in Ghana, NaomiChazan (1982) observes that in practice, ethnicity could mean anything from lan-guage groups, to communities, to administrative regions, depending on the matterat hand. It is thus necessary to guard against the assumption of stability or evenobjective existence of these identities, and to instead investigate how actors thinkabout themselves in relation to these identities.

Data and Methods

Ghana’s system of electoral democracy seems to exhibit clear evidence of politi-cized ethnicity. Its two dominant political parties have strong ethnic support bases.The National Democratic Congress (NDC) draws substantial support from theEwe-dominated Volta Region, while Asante and other Akan groups in the Ashantiand Eastern Regions overwhelmingly support the New Patriotic Party (NPP).However, it is impossible for either party to win power by relying solely on theirethnic bases (Arthur 2009). At any rate, Ghanaian voter decision is influenced bya host of non-identity-based factors, such as manifesto promises and governmentperformance (Lindberg and Morrison 2008). Since 1993, power has alternated be-tween these two political parties twice. Thus, observers of Ghanaian politics


6

applaud the country for its relatively stable electoral system, regarding it as an is-land of peace in a turbulent region. Nonetheless, sceptics warn that such evalua-tions suffer from the twin flaws of excessive haste and naïve optimism(Adjei 2012; Jockers et al. 2010).

To examine the impact of ethnicity on national identification, I ask a more gen-eral question: what is the impact of potentially competing group identifications onnational attachment? In addition to ethnicity, I include religion. RaufuMustapha (1986) points out that religion can, under certain circumstances, bean equally salient mode of attachment. Under colonial rule, religion became an in-strument of protest. In West Africa, churches were Africanized and the doctrinesof sects like the Jehovah’s Witnesses were transformed into potent weapons ofanti-colonial resistance (Hodgkin 1956). I also include professional identities; ifmodernization theory is correct, we should expect greater salience of professionalidentities to result in greater national attachment. Pan-African identity, while not asub-national attachment, could potentially detract from the loyalty which citizenswould otherwise offer the state. There is some overlap between the forms of iden-tification selected here and other measures presented in the literature, such as byEifert et al. (2010:497).2

Since the relationship between these forms of allegiance and national identifica-tion are theorized as a zero-sum game, I propose the following hypothesis: greaterlevels of attachment to non-state identities would result in lower levels of attach-ment to the state. Secondly, we can expect from the Durkheimian assumptions un-derlying modernization theory that greater attachment to a professional identitywould result in greater national attachment.

This paper is mainly based on data from a survey administered to 996 studentsat the University of Ghana and the Ghana Institute of Management and Profes-sional Administration between March and April 2017. Questionnaires were ad-ministered to students at both undergraduate and graduate levels. This isadmittedly unrepresentative. Nevertheless, the views of students are important,because they are the section of the population whose daily round of activitiesgives them the greatest exposure to the legitimating ideologies of the state. In ad-dition, they constitute ‘the political and economic elite’ in its embryonic form,which ‘makes their ideas on citizenship especially relevant’ (Godefroidtet al. 2016:7–8). Since this survey is not nationally representative, I use data fromRound 6 of the nationally representative Afrobarometer survey (2015), a datasetof 2,400 respondents, to get a sense of how closely the attitudes of my sample re-spondents approach national averages. I further contextualize the quantitativefindings using qualitative data from ongoing interviews, and I historicize themby using archival materials and the rich historical and anthropological stock ofknowledge on mutations of ethnic identity.

Dependent variable

The dependent variable is national attachment. I define this as the sum of citizens’emotional responses towards the state. National attachment is an amalgam of dif-ferent sentiments. To measure this concept, the questionnaire included a battery of


7

questions regarding (1) sense of belonging, (2) national pride, (3) sense of havinga personal stake in the country, (4) willingness to die for the country, and (5) aself-rated sense of patriotism. Respondents were presented with a set of state-ments (see Figure 1) and asked to indicate the extent of their agreement on aLikert scale with values ranging from 1 (strongly agree) to 5 (strongly disagree),with 3 as a neutral midpoint. To make for intuitive interpretation I reverse-codedthe Likert items. Feelings of national pride and national belonging were highamong the sampled students (4.38 and 4.30, respectively, out of a maximum of5). A self-professed sense of patriotism and a sense of having a personal stakein the country were only slightly lower (4.1 and 3.9, respectively). However, inspite of this clear sense of national loyalty, respondents balked at the idea ofsacrificing their lives for the country. Figure 1 shows that ‘dying for the country’falls below the mid-point line, indicating a small amount of disagreement with thestatement.

Most studies which attempt to measure national attachment use theLinz-Moreno question. This question asks respondents to place themselves onan identification scale that ranges from ‘solely ethnic’ to ‘solely national’, witha midpoint indicating equal ‘ethnic and national identification’. Exclusive use ofthe Linz-Moreno question to measure the relative salience of ethnic or nationalattachment makes it impossible to measure independently the salience of eachof these identifications (Guinjoan and Rodon 2015). I address this weakness by

Figure 1. Dimensions of national attachment. [Colour figure can be viewed atwileyonlinelibrary.com]


8

http://wileyonlinelibrary.com

generating a newmeasure of national attachment. To do this, I combine the individ-ual national sentiment questions into a composite variable as a measure of nationalattachment (Cronbach’s alpha, 0.79). I give twice more weight to willingness tosacrifice oneself and a sense of having a stake in the country, because these twoitems indicate a more intense and possibly action-oriented component of nationalsentiment compared to the other three more notional components.

Independent variables

Due to the design of the survey, education was naturally controlled for. The age ofrespondents ranged from 17 to 64 years, with a mean of 21.6 years. The other de-mographic variables are summarized in Table 1. Females made up 51.2% of thesample. Because of the youthfulness of the sample, most were single. In termsof partisan distribution, the governing New Patriotic Party (NPP) enjoys 43.4%support among students, with the main opposition National Democratic Congress(NDC) commanding only 11.6%. The level of support for the smaller parties com-bined is negligible, a reflection of the de facto two-party character of Ghanaian

Table 1. Demographic variables.

Variable Number %

Age Mean 21.6Range 17–64

Sex Female 498 51.2Male 474 48.8

972 100Marital status Single 920 95.8

Married 35 3.6Divorced 3 0.3Widowed 2 0.2

960 100Political affiliation* NPP 360 43.4

NDC 96 11.6Neutral 332 40.0PPP 28 3.4CPP 6 0.6GFP 2 0.2GCPP 1 0.1Other 1 0.1Multiple parties 3 0.4

829 100

*NPP (New Patriotic Party), NDC (National Democratic Congress), PPP (ProgressivePeople’s Party), CPP (Convention People’s Party), GFP (Ghana Freedom Party), GCPP(Great Consolidated Popular Party).


9

electoral politics. The share of support for the two main parties in the sample alsoreflects recent national trends. In the 2016 general elections, the NPP swept theboard for votes in all university campuses in the country.

The main independent variables measure the relative salience of the selectedidentities: religion, ethnicity, profession/education, family, and pan-African iden-tity. The preamble to this set of questions closely follows the wording of theLinz-Moreno question in the Afrobarometer survey. The questionnaire presentedrespondents with a hypothetical choice scenario and asked: would you considerany of the options below more important than being Ghanaian? Responses rangedfrom ‘strongly agree’ to ‘strongly disagree’, with a ‘neutral’ middle category. Ialso asked questions measuring participation in civic life (voting, voluntary groupmembership, and protesting) as control variables.

Figure 2 graphically represents the mean scores of the variables (identificationvariables are the first five bars on the left). Family attachment is the most salient(4.34), whereas ethnic group has the least salience relative to national identifica-tion. In fact, the ethnic score of 2.97 is evidence of a tiny mean disinclinationamong the students to choose ethnic over national identity. All other forms ofidentification – religion, studies/profession, pan-African – are above the neutralmidpoint line.

The low salience of ethnic scores is consistent with findings from the nationallyrepresentative Afrobarometer survey (see Figure 3). Eleven per cent of all respon-dents in that survey chose a predominantly ethnic attachment in response to thesurvey’s Linz-Moreno question. The comparable figure for students at tertiarylevel in that sample was almost 8%. In another study of Ghanaian university stu-dents, Godefroidt et al. (2016) found a similar pattern of responses to theirLinz-Moreno question. They also found something which to them was a puzzlingobservation; namely, that Ghanaian students’ conception of citizenship was acombination of both ethnic and civic elements.

Figure 2. Independent variables. [Colour figure can be viewed atwileyonlinelibrary.com]


10


The other independent variables in my survey data are measures of participationin national life. Voting is the most popular form of civic engagement, but nationallevel elections (3.9) far outweighed local level elections (2.96) in importance (Fig-ure 2). Participation in local level elections actually shows a slight disinclination.The trend of disinclination was much more obvious in all the other measures ofparticipation. Demonstrations and protests are the most unpopular civic actsamong my respondents, even though universities are generally hotbeds of unrestand protest (Balsvik 1998).

The Afrobarometer survey shows a similar trend of participation across differ-ent spheres of civic action. As shown in Figure 4, electoral participation remains

Figure 3. Ethnic versus national identification. [Colour figure can be viewed atwileyonlinelibrary.com]

Figure 4. Participation in civic acts. [Colour figure can be viewed atwileyonlinelibrary.com]


11



by far the most popular civic act. Other forms of civic engagement such as asso-ciational life and contact with leaders remain low. Participation in direct forms ofcivic protest among Afrobarometer respondents was as low as it was with my stu-dent sample.

Findings

For the analysis, I created dummy variables for the five identification questions.‘Agree’ and ‘Strongly agree’ were coded 1 (‘Yes’), and the remaining responsecategories were coded 0 (‘No’). I also created dummies for sex (with ‘Male’ asthe reference category) and party affiliation (NDC, NPP, and Neutral). Figure 5graphically represents the results of the OLS regression prediction of the strengthof national attachment as a function of other forms of identification. Model 1 in-cludes only the estimated effects of the main independent variables. Religion,Africa, and ethnic identity have positive effects, while the effects ofstudies/profession and family are negative. All of these effects are significant atthe 0.01 alpha level, with the exception of Africa and religion, which are signifi-cant at the 0.05 level. This finding rejects the modernization theory-inspiredhypothesis that leads us to predict an inverse relationship between ethnicity andnational attachment. It also rejects another hypothesis of modernization theorythat a positive relationship should exist between professional identification andnational attachment.

Although the Afrobarometer survey does not contain a directly comparablequestion, we can infer from other questions (if not confirm) with at least someconfidence that the causal direction of ethnicity here is not due to random error.To gauge tolerance, Afrobarometer respondents were asked to answer a batteryof question about attitudes towards having certain social groups as neighbours.Only 4.1% of national respondents expressed discomfort at having non-ethnicneighbours. As can be seen from Figure 6, more university students (11.3%)actually expressed aversion for members of other ethnic groups. Moreover, theseattitudes appear to be independent of the strength of ethnic attachment. The sec-ond chart in Figure 6 compares respondents whose identities were mainly ethnic(see Figure 3 above) with the full Afrobarometer sample. In both groups, tolerancefor non-ethnics was overwhelming, at more than 90%.

After adding demographic controls to my regression model, the effect sizes donot change much. There is a small increase in the effect of religion and slight de-creases in the effects for familial, pan-African, and ethnic identities. However, theeffect sizes and direction of these demographic variables are interesting in theirown right. Being single leads to a large decrease in national attachment, while par-tisan neutrality, rather than partisanship, negatively affects national attachment.However, none of these demographic variables were significant. The full modelincludes the variables for civic participation and they all have positive effects ex-cept participation in a public protest, which is also significant. Including participa-tion controls only slightly reduces the impact of the identification variables.

Further examination renders these findings less surprising. For example,Levstik and Groth (2005) found in a study of high-school students in Ghana that


12

Figure 5. OLS predictions of strength of national attachment. [Colour figurecan be viewed at wileyonlinelibrary.com]


13


ethnic salience was not incompatible with intense expressions of nationalism.Among the students in their study, ethnic histories fed into, rather than detractedfrom national histories. They argue that this attitude was important in fosteringtolerance for ethnic diversity among these students.

Similar sentiments were expressed by respondents in the ongoing qualitative in-terviews I am conducting as part of this project.3 My respondents repeatedly madereference to the colonial origins of the country, a fact which made them keenlyalive to the responsibility that ‘we’ve got to find a way to live together’. Thus,in their definitions of Ghanaianness, their first impulse was to refer to the legalstipulations: possession of a Ghanaian passport, birth certificate, and the like. Amusician, Odo Nkoaa,4 said that the essence of Ghanaianness ‘is just a piece ofpaper left by the British’. When respondents referred to cultural ingredients ofthe Ghanaian identity, this reference was always vague. It was only upon furtherprobing that they mentioned specific cultural norms and practices, such as pecu-liarities of speech or greetings, themselves not exclusively Ghanaian traits.

Like the respondents in Levstik and Groth’s study, my respondents tended tosee no contradiction between a strong ethnic and national identification. Theywarned about the dangers of sliding into bigotry and assuming an air of superior-ity, but in itself, they found nothing wrong with celebrating one’s ethnicity. In-deed, journalist Naa Oyo confessed that a sense of ethnic belonging came morenaturally to her than national identification. As she said, ‘I don’t struggle withwhat it means to be Asante. There’s a certain connection there that I don’t feelwith being Ghanaian’. This largely arose from her sense of the failures of the

Figure 6. Attitudes to having non-ethnic or foreign neighbours. [Colour figurecan be viewed at wileyonlinelibrary.com]


14


Ghanaian state to, as she so poetically put it, make her ‘feel like a beloved’. Thissentiment mirrors some scholarship on the Ghanaian state. The Ghanaian state hasbeen described as ‘a pain disseminating rather than a welfare state’; as a reluctantor inefficient distributor of welfare benefits, but a generous dispenser of pain

Figure 7. Predicted probabilities for the dependent variable. [Colour figurecan be viewed at wileyonlinelibrary.com]


15


(Anyimadu 2006:5). Jonathan Frimpong-Ansah (1991) has given Ghana thebloodcurdling soubriquet, ‘vampire state’.

Individualistic and Collectivistic Orientations

The regression results seem to map onto an underlying pattern. The variableswhich presented negative effects (career, family, singlehood, and partisan neutral-ity) are variables that point to an underlying individualistic orientation. On theother hand, religion, pan-African identity, ethnicity, and partisanship all point to-wards a collectivistic orientation. Although family is a collective unit, in relationto other collective categories like village or ethnic group it indexes a private realm.By extension, the variables which focus on the interests of the family, if at oddswith particular communal interests, would lead to an individual taking decisionswhich would be considered at odds with the group interest. This is what EdwardBanfield (1958) famously referred to as ‘amoral familism’, which he identified asa key reason why the southern Italian village he studied could not embark on col-lective action.

Recognizing the underlying pattern of individualistic and collectivistic orienta-tions removes the surprise that might have been occasioned by the positive effectof ethnicity.5 The most telling evidence comes from the predicted values for mar-ital status and political affiliation (Figure 7). As we should expect from the direc-tion of the regression results, marriage and widowhood yield positive predictions.Divorce generates an even greater negative effect than singlehood. Again, thisconfirms the inference that it is individualistic orientation that is driving theeffects; divorce is evidence of a strong self-assertive spirit, especially in a socialcontext where marriage is so highly valued and where adults of both sexes are ‘ex-pected to marry and have children’ (Takyi 2001:79). We find a similar pattern forpartisan affiliation. The predicted probabilities for the ruling and all the importantopposition parties are also positive, even though there is reason to assume thatmembership in an opposition party could weaken national attachment. It is actu-ally partisan neutrality which yields a negative prediction. This evidence is impor-tant because non-partisanship is often touted as evidence of closer identificationwith the nation-state, and is accorded greater normative value than the supposeddivisiveness of partisanship.6

My in-depth interviews provide further qualitative evidence for the inferred un-derlying pattern. Respondents repeatedly mentioned ethnic belonging as an im-portant component of their sense of national identity. Naa Oyo, who reports ahigh salience of her ethnic identification, claims nevertheless to hold both nationaland ethnic identifications harmoniously: ‘I can be both Asante and Ghanaian’.Much of her notion of belonging came from her father, whom she describes asbeing ‘pro-Ghana, pro-CPP, pro-Asante’.7 It is telling that all three identifications– country, political party, ethnic group – fall on the collectivistic end of the orien-tation scale. Oyo’s father inculcated in her the importance of all these forms ofbelonging because, she said, ‘society mattered to him’. It is clear here, as I havealso shown in the analysis of the statistical results, that the collectivistic orienta-tion which underpins ethnic identification is what also underpins attachment to


16

the state. The expectation of modernization theorists that nationalistic identitieswould supplant ethnic loyalties is contradicted by these data. Even more impor-tantly, my study shows that both forms of identification are underlain by the samecollectivistic orientation.

Social Rootedness:8 Towards a Theory of Ethnicity and National Attachment

In spite of their other differences, there is agreement between modernizationand colonial legacy theories on the perverse effects of ethnicity on nationalattachment. Modernization theory conceives of the relationship between ethnicand national identification as a zero-sum game. Even Sklar (1960) andWallerstein (1960) view ethnicity as simply a help-along-the-way, a catalyst whichwill self-obliterate after performing its good work. They assume that with the tran-sition from tradition to modernity, ethnic identification would atrophy. It was thatmisunderstanding that led Godefroidt et al. (2016), on finding that Ghanaian stu-dents had both ethnic and civic conceptions of citizenship, to describe these con-ceptions as ‘postmodern’. I will argue that there is nothing postmodern aboutsimultaneously holding both civic and ethnic conceptions of citizenship; and,further, that this phenomenon is not even new.

Durkheim’s theory of the relationship between mechanical and organic solidar-ity has been hugely influential with modernization scholars, in spite of the fact thatin The Division of Labour (Durkheim 1933) there appears to be a lack of clarityabout the actual nature of these solidarities. On the one hand, Durkheim arguesthat solidarity based on interdependence and self-interest would totally supplantthe prevailing primordial moralities which underlie different forms of solidarityin traditional societies. On the other hand, he recognized that in addition toassociational/occupational ethics and the complex division of labour, there wasstill a need for a more generalized morality – drawing from the traditional order– to hold modern societies together (for an extended discussion of this criticism,see Pope and Johnson 1983). Thus, the agonistic conception of the relationshipbetween organic and mechanical solidarity is problematic.

Colonial legacy arguments, on the other hand, place the blame for ethnicity’sdamaging effect on the hubristic imposition of colonial boundaries, and on thecynical manipulation of apparent ethnic differences by colonial authorities in theirquest to maintain control. Moreover, the ‘importedness’ of the state itself in post-colonial societies, so the argument goes, has resulted in estranged state-societyrelations. Badie and Birnbaum claim that:

to this day the ‘state’ is no more than an imported artifact in both Africa andAsia, a pale copy of utterly alien European social and political systems, aforeign body that is not only inefficient and a burden on society but also afomenter of violence. (Badie and Birnbaum 1983:99)

Such arguments can, however, be taken to extremes by scholars of ethnicity inpostcolonial countries, and may end up prematurely foreclosing close scrutiny.The underlying assumption is that colonialism motivated people to look inwardsinto their ethnic groups, rather than outwards to the emerging state. Nonetheless,


17

there is limited evidence for the proposition that an ‘organic’ state would be moreacceptable to its people than an ‘imposed’ or ‘imported’ one. Historically, theemergence of states has been marked by struggle. Theories of state formation inEurope highlight the formative role of (figurative) banditry (Olson 1993), coer-cion, and violence; Charles Tilly (2002 [1985]) likens the process to the operationof organized crime.

Similarly, although colonialism created deep social fractures, those crackssometimes became strategic spaces for negotiation. In the history of the GoldCoast, there were moments of protest and agitation, and there were moments ofcooperation. With the consolidation of colonial rule, the educated elite and tradi-tional rulers jostled for influence in the emerging juridical order. Commoners didnot remain passive in this process either. An emerging body of scholarship on thedynamics of colonialism has started to show that the widely accepted dichotomybetween citizenship and subjecthood is simplistic and precludes serious analysisof the negotiations over rights and belonging (Cooper 2005; Hussin 2007;Hunter 2016a).

Moreover, there is a tendency to focus selectively on instances of inter-ethnictension, but this does not constitute the full picture. Recent scholarship has startedrevisiting interpretations of the pre-colonial past which have privileged warfarebetween states. In a recent paper, Karl Haas takes issue with interpretations ofAsante-Dagbon relations which emphasize conquest and Asante control, showingthat these ‘have been greatly exaggerated’ (2017:224). Indeed, there is sufficienthistorical evidence of inter-ethnic cooperation or co-existence before and duringcolonial rule. For instance, in his autobiography, Milestones in the History ofthe Gold Coast, Nii Kwabena Bonne (1953) describes himself as Mantse ofOsu Alata (in Accra) and Oyokohene of Asante. That Kwabena Bonne could holdtraditional leadership titles in both states points to the fluidity of notions of be-longing both before and during colonialism.9 Moreover, many communitiesthroughout the country have historically hosted ‘stranger quarters’ or ‘settle-ments’. Things were by no means always paradisal, and this arrangement camewith the inevitable tensions and open conflicts, but such periods of conflict punc-tuated long periods of harmonious co-existence.10 It was not uncommon forstranger quarters ultimately to be assimilated formally into the social and politicalstructures of their hosts.

Patterns of economic relations, migration, and settlement often defied attemptsby both colonial officials and traditional rulers to define or impose rigid ethnicidentities. An interesting case is the attempt by Paramount King of AkyemAbuakwa, Nana Sir Ofori Atta, to define a distinctive Akyem identity. This hege-monic attempt was countered by commoners. In a rich account of the ensuingdrama, Richard Rathbone describes the role played by the asafo, a traditional mil-itary band:11

[…] in all respects a profoundly Akan institution, [it] was composed not onlyof Akyemfo commoners but also of Asante, Akuapems and, more strikingly,Ga and Northern residents of the kingdom. Accordingly it would seem fair toargue that Ofori Atta’s agenda, which increasingly defined a[n Akyem]


18

citizenship based upon something close to ethnicity, did not go uncontested.It seems fair to conclude that Ofori Atta’s definition of citizenship presentedhis peoplewith a heavier burden, the increasing obligations of being subjects,rather than expanding their access to rights. This propensity allowed someunlikely bedfellows to emerge in the numerous struggles against the palace’sattempts to centralize power in the kingdom. (Rathbone 1996:518)

Ethnicity is a fact of social life, a reflection of the human desire for social rooted-ness. This rootedness, whether found in ethnic, religious, or national communi-ties, does not automatically express itself in automatic antagonism to othercommunities. In fact, as John Lonsdale (1994:132) argues: ‘Ethnicity is alwayswith us; it makes us moral – and thus social – beings’. Moreover, as I have pointedout above, there was a pattern of inter-ethnic relations before and during colonialrule. Thus, the notion that ethnic co-existence was suddenly going to be a problemin the newly independent states misrepresents the problem of postcolonial nationbuilding, as does the corresponding view that the people needed to be tutored bytheir ‘bettered brethren’ who had acquired the precious gift of cosmopolitanism.These problematic assumptions, however, continue to inform much of the schol-arship on ethnicity.

Conclusion

The statistical analyses reported in this paper show that, contrary to prevailingtheoretical expectations, strong ethnic identification actually predicts strongernational attachment. I draw on qualitative and archival data to give flesh to theskeleton of the statistical findings. I argue that rather than being a zero-sum game,ethnic and national identifications are actually manifestations of the same under-lying orientation. All the forms of identification which are collectivistic in orien-tation – ethnic, pan-African, and religious – have positive effects on nationalattachment. On the other hand, individualistic orientations – family, career, andpolitical neutrality – reduce the strength of national attachment (see Table 2).

What does this mean, in practical terms, for contemporary problems of nationalunity? As noted in the introduction, anxieties about the negative effects of ethnic-ity pervade public discourse in Ghana and, indeed, in many other African states. Ina public lecture in 2008, political scientist Kwame Ninsin expressed a concernabout how politics ‘has tended towards tribalism’, warning that the increasingsalience of ethnic identities ‘threatens stability and peace’ (Ninsin 2008).

Inter-ethnic conflicts pose serious threats to national cohesion. At the sametime, most local conflicts in Ghana have occurred not between but within ethnic

Table 2. Orientations and national attachment.

Individualistic orientations Collectivistic orientations

Family ReligionCareer (profession/studies) Ethnic groupPolitical neutrality Pan-African identity


19

groups. These conflicts have flared up over such matters as access to land or chief-taincy disputes (Tsikata and Seini 2004). And where there have been ethnicallymotivated political tensions or conflicts, the culprits have mostly been membersof the elite class. Indeed, Ninsin has pointed out elsewhere that ‘primordial iden-tities like ethnicity and religion have become part of the ideology of dominationby which the political class manipulates the electorate to enhance their electoralfortunes’ (Ninsin 2006:5; see also Ekeh 1975).

This is done either explicitly by exciting or mobilizing ethnic passions, or indi-rectly by insinuating ethnicity into otherwise purely policy matters. The former isparticularly effective at critical junctures such as the loosening of autocratic con-trol, as happened in Kenya’s 1992 general election (Oyugi 1997). By contrast, in-direct ethnic mobilization tends to occur during more settled times and especiallyin pluralistic systems where no party is dominant enough to win power on its own.In these cases, political elites employ ethnic wedge issues ‘to splinter the supportof a key opponent by an ethnic issue appeal to members of the opponent’s coali-tion’ (Gadjanova 2017:486). This contradiction of the caretaker turned offenderwas captured in Chinua Achebe’s The Trouble with Nigeria:

I was an eye-witness to that momentous occasion when Chief ObafemiAwolowo ‘stole’ the leadership of Western Nigeria from Dr NnamdiAzikiwe in broad daylight on the floor of the Western House of Assemblyand sent the great Zik scampering back to the Niger ‘whence [he] came’.Someday when we shall have outgrown tribal politics, or when our childrenshall have done so, sober historians of the Nigerian nation will see thatevent as the abortion of a pan-Nigerian vision. (Achebe 1984:5)

Ninsin (2006:2) is even harsher in his general condemnation of the political classin Africa. He decries their ‘iconoclastic propensities’, and blames them for the im-plosion of the ‘institutions of self-government that the people had built during thestruggle for independence’.

The object of Ninsin’s worry is what Lonsdale refers to as ‘political tribalism’.Since ethnicity as such does not necessarily entail inter-group tensions or intoler-ance, nor indeed does it lie at the root of all conflicts in Africa, it is actually‘political tribalism itself [which] always needs to be explained’ (Lons-dale 1994:132). Even though some forms of ethnic tension do find social expres-sion, they rarely achieve political salience unless they are capitalized on as aweapon in national struggles over resources. And since these struggles are wagedbetween national elites, it is clear that they are the ones who most need to takeheed of their own sermons on national integration, rather than directing them at‘the masses’.

Acknowledgements

Early versions of this paper were presented at seminars at the Institute for Ad-vanced Study in Toulouse (IAST) and at the Institute of African Studies at theUniversity of Ghana. I thank participants at these forums, especially colleaguesin the history section of IAST. For comments on earlier versions of this paper, I


20

am grateful to Kofi Blankson Ocansey, Edward Boateng, Slimane Dridi, MartinAgyemang, Esther Darku, Sheila Dzifa, and Christopher Muhoozi. I am gratefulto Eunice Darku for research assistance. The research for this paper was con-ducted while I was a research fellow at IAST. Support through the ANR LabexIAST is gratefully acknowledged.

Notes1 Wallerstein did not directly quote Banfield.2 Their list of salient identities includes ethnicity, religion, class/occupation, gender, and amiscellaneous category labelled ‘other’.3 This qualitative component of the study is being conducted using a theoretical samplingmethod which is marked by an iterative approach to data collection and analysis. The pro-cess commenced in a two-fold process: (1) follow-up interviews with participants from thesurvey and (2) interviews with prominent or influential public personalities such as journal-ists. As the study progresses, further data collection will be informed by emerging insightsfrom the initial set of data being collected. Thus far, about twelve interviews have beencompleted.4 The respondents’ names used in this paper are all pseudonyms.5 However, since the effects of partisanship are not statistically significant, we cannot saymuch about its effect beyond the observation that it provides a straw-in-the-wind proof forthe individualistic-collectivistic pattern which I discern in the data.6 The Ghana Freedom Party and the Great Consolidated Popular Party are insignificantparties with comical leaders, and we can assume that support for these parties indicates adismissive attitude towards the state.7 CPP, the Convention People’s Party, is the party of Kwame Nkrumah, Ghana’s firstpresident.8 I am grateful to Kofi Blankson Ocansey for this phrase.9 This was not unique to the Gold Coast. Throughout history, states and kingdoms havebeen enlarged through, for instance, marriage. Notions of belonging and identification havenever been fixed in terms of place.10 For two fascinating case studies, see Arhin (1971) and Kobo (2010).11 These groups were militarily demobilized during the colonial period but they retainedtheir associational functions.

References

Achebe, Chinua. 1984. The Trouble with Nigeria. Oxford: Heinemann.

Adjei, Joseph K. 2012. ‘An African Model? Conflict Management in Ghana’s 2008 Presi-dential Elections’. In Managing Conflicts in Africa’s Democratic Transitions, ed.Akanmu G. Adebayo. Lanham, MD: Lexington Books, pp. 233–254.

Afrobarometer. 2015. Ghana Round 6 Data. Available at: https://afrobarometer.org/data/ghana-round-6-data-2015 (accessed 30 March 2017).

Anyimadu, Amos. 2006. ‘The State and Pain in Ghana: Toward a Long-term Analysis’.Hemispheres: Studies on Cultures and Societies 21: 5–16.

Arhin, Kwame. 1971. ‘Strangers and Hosts: A Study in the Political Organisation and His-tory of Atebubu Town’. Transactions of the Historical Society of Ghana 12: 63–82.

Arthur, Peter. 2009. ‘Ethnicity and Electoral Politics in Ghana’s Fourth Republic’. AfricaToday 56 (2): 44–73.


21

https://afrobarometer.org/data/ghana-round-6-data-2015

https://afrobarometer.org/data/ghana-round-6-data-2015

Badie, Bertrand and Pierre Birnbaum. 1983. The Sociology of the State, trans. ArthurGoldhammer. Chicago, IL: University of Chicago Press.

Balsvik, Randi Rønning. 1998. ‘Student Protest – University and State in Africa1960–1995’. Forum for Development Studies 25 (2): 301–325.

Banfield, Edward C. 1958. The Moral Basis of a Backward Society. New York: The FreePress.

Berry, Sarah. 1992. ‘Hegemony on a Shoestring: Indirect Rule and Access to AgriculturalLand’. Africa 62 (3): 327–355.

Chazan, Naomi. 1982. ‘Ethnicity and Politics in Ghana’. Political Science Quarterly 97 (3):461–485.

Cohen, Abner. 1969. Custom and Politics in Urban Africa: A Study of Hausa Migrants inYoruba Towns. London: Routledge and Kegan Paul.

Cooper, Frederick. 2005. Colonialism in Question: Theory, Knowledge, History. Berkeley,CA: University of California Press.

Durkheim, Emile. 1933. The Division of Labour in Society. New York: Free Press.Eifert, Benn, Edward Miguel and Daniel N. Posner. 2010. ‘Political Competition and

Ethnic Identification in Africa’. American Journal of Political Science 54 (2): 494–510.Ekeh, Peter P. 1975. ‘Colonialism and the two Publics in Africa: A Theoretical Statement’.

Comparative Studies in Society and History 17 (1): 91–112.Ekeh, Peter P. 1990. ‘Social Anthropology and Two Contrasting Uses of Tribalism in

Africa’. Comparative Studies in Society and History 32 (4): 660–700.Frimpong-Ansah, Jonathan H. 1991. The Vampire State in Africa: The Political Economy of

Decline in Ghana. Trenton, NJ: Africa World Press.Gadjanova, Elena. 2017. ‘Ethnic Wedge Issues in Electoral Campaigns in Africa’s Presi-

dential Regimes’. African Affairs 116 (464): 484–507.Godefroidt, Amélie, Arnim Langer and Bart Meuleman. 2016. ‘Towards Post-Modern

Identities in Africa? An Analysis of Citizenship Conceptualizations in Ghana’. CRPDWorking Paper No. 51. Available at: http://soc.kuleuven.be/crpd/files/working-papers/WP%2051%20Towards%20post-modern%20identities%20in%20Africa.pdf (accessed10 September 2017).

Guinjoan, Marc and Toni Rodon. 2015. ‘A Scrutiny of the Linz-Moreno Question’.Publius: The Journal of Federalism 46 (1): 128–142.

Haas, Karl J. 2017. ‘AView from the Periphery: A Re-Assessment of Asante-DagbambaRelations in the 18th Century’. International Journal of African Historical Studies50 (2): 205–224.

Hobsbawm, Eric. and Terence O. Ranger. 1983. The Invention of Tradition. Cambridge andNew York: Cambridge University Press.

Hodgkin, Thomas. 1956. Nationalism in Colonial Africa. London: Muller.Hunter, Emma, ed. 2016a. Citizenship, Belonging, and Political Community in Africa:

Dialogues between Past and Present. Athens, OH: Ohio University Press.Hunter, Emma. 2016b. ‘Introduction’. In Citizenship, Belonging, and Political Community

in Africa: Dialogues between Past and Present, ed. Emma Hunter. Athens, OH: OhioUniversity Press.

Hussin, Iza. 2007. ‘The Pursuit of the Perak Regalia: Islam, Law, and the Politics ofAuthority in the Colonial State’. Law & Social Inquiry 32 (3): 759–788.

Jockers, Heinz, Dirk Kohnert and Paul Nugent. 2010. ‘The Successful Ghana Election of2008: A Convenient Myth?’ The Journal of Modern African Studies 48 (1): 95–115.

Kobo, Ousman. 2010. ‘“We Are Citizens Too”: The Politics of Citizenship in IndependentGhana’. The Journal of Modern African Studies 48 (1): 67–94.


22

http://soc.kuleuven.be/crpd/files/working-papers/WP%2051%20Towards%20post-modern%20identities%20in%20Africa.pdf

http://soc.kuleuven.be/crpd/files/working-papers/WP%2051%20Towards%20post-modern%20identities%20in%20Africa.pdf

Kwabena Bonne, N. 1953. Milestones in the History of the Gold Coast: Autobiography ofNii Kwabena Bonne III, Osu Alata Mantse, Also Nana Owusu Akenten III, Oyokoheneof Techiman, Ashanti. Accra: Diplomatist Publications.

Lentz, Carola. 1995. ‘“Tribalism” and Ethnicity in Africa: A Review of Four Decades ofAnglophone Research’, trans. Pamela Selwyn. Cahiers des Sciences Humaines 31 (2):303–328.

Lentz, Carola. 2000. ‘Colonial Constructions and African Initiatives: The History ofEthnicity in Northwestern Ghana’. Ethnos 65 (1): 107–136.

Levstik, Linda S. and Jeanette Groth. 2005. ‘Ruled by Our Own People: Ghanaian Adoles-cents’ Conceptions of Citizenship’. Teachers College Record 107 (4): 563–586.

Lindberg, Steffan I. and Minion K. Morrison. 2008. ‘Are African Voters Really Ethnic orClientelistic? Survey Evidence from Ghana’. Political Science Quarterly 123 (1):95–122.

Lonsdale, John. 1994. ‘Moral Ethnicity and Political Tribalism’. Occasional Paper No. 11.Inventions and Boundaries: Historical and Anthropological Approaches to the Study ofEthnicity and Nationalism, ed. Preben Kaarsholm and Jan Hultin. Roskilde, Denmark:Institute for Development Studies, Roskilde University, pp. 131–150.

Mamdani, Mahmood. 1996. Citizen and Subject: Contemporary Africa and the Legacy ofLate Colonialism. Princeton, NJ: Princeton University Press.

Mustapha, A. Raufu. 1986. ‘The National Question and Radical Politics in Nigeria’.Review of African Political Economy 13 (37): 81–96.

Ninsin, Kwame A. 2006. ‘Introduction: The Contradictions and Ironies of Elections inAfrica’. African Development 31 (3): 1–10.

Ninsin, Kwame A. 2008. Professor Kwame Ninsin on the Challenge of Tribalism in Ghana.Available at: https://www.youtube.com/watch?v=cJPI81G4ulU (accessed 9 October2018).

Olson, Mancur. 1993. ‘Dictatorship, Democracy, and Development’. American PoliticalScience Review 87 (3): 567–576.

Owusu, Maxwell. 1989. ‘Rebellion, Revolution, and Tradition: Reinterpreting Coups inGhana’. Comparative Studies in Society and History 31 (2): 372–397.

Oyugi, Walter O. 1997. ‘Ethnicity in the Electoral Process: The 1992 General Elections inKenya’. African Journal of Political Science/Revue Africaine de Science Politique 2 (1):41–69.

Parsons, Talcott. 1991 [1951]. The Social System. London: Psychology Press.Pope, Whitney and Barclay D. Johnson. 1983. ‘Inside Organic solidarity’. American

Sociological Review 48 (5): 681–692.Ranger, Terence. 1983. ‘The Invention of Tradition in Colonial Africa’. In The Invention of

Tradition, ed. Eric Hobsbawm and Terence Ranger. Cambridge: Cambridge UniversityPress, 211–262.

Ranger, Terence. 1993. ‘The Invention of Tradition Revisited: The Case of ColonialAfrica’. In Legitimacy and the State in Twentieth-century Africa, ed. Terence Rangerand Olufemi Vaughan. London: Palgrave Macmillan, pp. 62–111.

Rathbone, Richard. 1996. ‘Defining Akyemfo: The Construction of Citizenship in AkyemAbuakwa, Ghana, 1700–1939’. Africa 66 (4): 506–525.

Robinson, Amanda L. 2014. ‘National versus Ethnic Identification in Africa: Moderniza-tion, Colonial Legacy, and the Origins of Territorial Nationalism’. World Politics 66(4): 709–746.

Rostow, Walt W. 1959. ‘The Stages of Economic Growth’. The Economic History Review12 (1): 1–16.


23

https://www.youtube.com/watch?v=cJPI81G4ulU

Rouch, Jean. 1956. ‘Migrations au Ghana’ [Migrations to Ghana]. Journal de la Sociétédes Africanistes [Journal of the Society of Africanists] 26 (1): 33–196.

Sklar, Richard L. 1960. ‘The Contribution of Tribalism to Nationalism in Western Nigeria’.Journal of Human Relations 8 (3/4):407–418.

Sklar, Richard L. 1967. ‘Political Science and National Integration: A Radical Approach’.The Journal of Modern African Studies 5 (1): 1–11.

Spear, Thomas. 2003. ‘Neo-traditionalism and the Limits of Invention in British ColonialAfrica’. The Journal of African History 44 (1): 3–27.

Takyi, Baffour K. 2001. ‘Marital Instability in an African Society: Exploring the Factorsthat Influence Divorce Processes in Ghana’. Sociological Focus 34 (1): 77–96.

Tilly, Charles 2002 [1985]. ‘War Making and State Making as Organized Crime’. In Vio-lence: A Reader, ed. Catherine Besteman. New York: New York University Press,pp. 35–60.

Tsikata, Dzodzi and Wayo Seini. 2004. ‘Identities, Inequality and Conflicts in Ghana’.CRISE Working Paper No. 5. Oxford: Queen Elizabeth House, Oxford University.

Valenzuela, J. Samuel and Arturo Valenzuela. 1978. ‘Modernization and Dependency:Alternative Perspectives in the Study of Latin American Underdevelopment’. Compar-ative Politics 10 (4): 535–557.

Wallerstein, Immanuel. 1960. ‘Ethnicity and National Integration in West Africa’. Cahiersd’Études Africaines [Journal of African Studies] 1 (3): 129–139.

Young, Crawford. 1979. The Politics of Cultural Pluralism. Madison, WI: University ofWisconsin Press.


24

individualistic and collectivistic orientations: examining

Documents