ARTICLES

 

Are dialects markers of ethnic identity? The case of Setswana dialects and ethnic groups

 

 

Mompoloki Mmangaka Bagwasi

Department of English, University of Botswana, Private Bag 00703, Gaborone, Botswana E-mail: bagwasimm@mopipi.ub.bw

 

 

---

ABSTRACT

Many studies on dialects present language in neat, organised groupings that highlight similar language habits and linguistic features of people who belong to the same social, linguistic or regional group. In that way, social and regional groups are identified by the dialects that they speak, and vice versa. However, given the fluid and mobile nature of languages, dialects, and people, it is time that this relationship between language and identity was reviewed, and its complexity exemplified. The fluidity and dynamism of language makes it difficult to attach any linguistic features to any group of people or location. Using examples from Botswana, this paper argues that the relationship between Setswana dialects and Botswana ethnic and regional groups is non-representational and non-exclusive. Thus, the paper makes a distinction between Setswana ethnic groups and Setswana dialects, and challenges current perceptions of Setswana dialects which are based on ethnicity. The argument of the paper is based on historical claims, and translanguaging and levelling theories.

Keywords: dialect, identity, ethnicity, Botswana, Setswana, translanguaging, levelling

---

 

 

1. Objective of the paper

This paper, which is hypothetical and conjectural, has three objectives: to distinguish Setswana ethnic groups from Setswana dialects, to problematise the current perception of Setswana dialects that is based on ethnicity, and to argue for a linguistic approach that views language as fluid and, therefore, difficult to attach to any regional or ethnic group. The paper uses historical arguments, and translanguaging and dialect-levelling theories to refute the perceived connection between ethnic and linguistic identity in Setswana dialects.

 

2. Introduction

In Botswana, the relationship between ethnic and dialectal groups is perceived to be so close that each ethnic group has a dialect by which it is identified. For example, Otlogetswe (2012: 534) defines "Sekwena" as a dialect of Setswana spoken by the Bakwena, "Selete" as the dialect of Setswana spoken by the Balete ethnic group (Otlogetswe 2012: 537), and "Sengwaketse" as a dialect of Setswana spoken by the Bangwaketse ethnic group (Otlogetswe 2012: 540). Thus, labels for Setswana dialects are nouns that refer to a dialect as well as adjectives that describe a certain culture, practice, behaviour, object, and belief of this community. It seems that there is a distinct and defined way in which each ethnic group speaks (its dialect) and behaves. Furthermore, for an individual to belong to an ethnic group, they have to show their membership by speaking the dialect associated with the group and behaving in a way that is characteristic of that ethnic group. On the surface, the relationship between an ethnic group and its dialect seems straightforward. However, in reality, dialects are less definite, ethnic groups more multifarious, and the relationship between them most unpredictable.

A language is often described as a group of related dialects that constitute a single linguistic norm. Dialects develop when speakers of the same language become separated, geographically or socially, and begin speaking different variations of the same language. Southerland and Katamba (1996: 565) point out that these variations are found in vocabulary, grammar, and pronunciation. Unfortunately, there is no one linguistic criterion which specifies the degree of linguistic variation or similarity that gives rise to different dialects or languages. In cases where dialects exist in a geographical continuum or chain, at any point in the chain, speakers of a dialect can understand speakers of the other dialects who live in adjacent areas to them because there are many linguistic similarities between the two dialects. However, as the chain gets longer, those at the beginning of the chain may find it difficult to understand those who live further along the chain because there are many linguistic differences between the two dialects [see Crystal (1987)].

Thus, dividing the linguistic chain into different dialects and languages is problematic because the separation is not simply based on their degree of mutual intelligibility, linguistic similarities or differences. While some linguists base their categorisation of dialects on linguistic similarity or difference, others also consider social relations. There is therefore no straightforward and objective criterion for delineating clear dialect and language boundaries. The divisions are based on different kinds of factors: social, historical, linguistic, political, and/or ideological.

 

3. The major Setswana ethnic groups and their dialects

Ngcongco (1979: 25) argues that the Sotho Tswana group was not ethnically and linguistically well-defined nor fully differentiated when it crossed into southern Africa. It came in "small scale scattered movements of several segmentary lineage groups occurring slowly and gradually in many directions over a wide area". Later on, some of its members proceeded to Bechuanaland (now Botswana) to form a group of people who are now called "Batswana". The first Sotho Tswana group to come to Bechuanaland in the 14^th century is said to have been the Bakgalagadi who found the San or Basarwa people already settled in the country. The Bakgalagadi are said to have been followed by the Barolong and Tlhaping who defeated and displaced the Bakgalagadi and the Basarwa. The last and largest wave of Sotho Tswana people to come to Bechuanaland, around 1540, are those who later became the Bahurutshe, Bakwena, Bangwaketse, and Bangwato. The Bangwato later split to form a group called "Batawana" in 1795. The Bakgatla, Batlokwa, and Balete are said to have arrived much later in the 19^th century [see Anderson and Jason (1997), Ngcongco (1979), Schapera (1952, 1994), and Tlou and Campbell (1997)]. Each ethnic group is said to have expanded its territory and increased its membership through conquest of minority and less powerful groups such as the Bakgalagadi, Basarwa, Bakalanga, Babirwa, Batswapong, and the Bayeyi who were subjugated and became the subjects of their conquerors. The conquered groups often had their land confiscated and their identities subsumed under the dominant Tswana group that conquered them.

At some stage, these different groups adopted a well-developed and distinguishable language called "Setswana" which is spoken in some southern African countries like South Africa, Botswana, Namibia, and Zimbabwe. In Botswana, Setswana is the dominant lingua franca, spoken by 70-80% of the population as a first language, and by approximately 20% as a second language (Anderson and Jason 1997: 21). Because of its dominance in the country, Setswana enjoys the status of national language. It is now used in all parts of the country, in official, semiofficial and everyday communication. The many uses and spread of Setswana in Botswana has resulted in various geographical and social dialects.

According to Nyati-Ramahobo (2008), the Botswana Chieftainship Act recognises eight major ethnic groups: Barolong, Bakwena, Bangwaketse, Balete, Bakgatla, Batlokwa, Bangwato and Batawana who speak eight mutually intelligible dialects which are collectively known as "the Setswana language". She further points out that the British organised Botswana on the philosophical and territorial concept of 'Tswanadom' which has led many observers to assume that Botswana is a mono-ethnic state. Batibo (1999: 5) also points out that Sekwena, Sengwaketse, Sengwato, Serolong, Setlharo, Setawana, Setlokwa, Sekgatla, and Selete are "what have been accepted as Setswana dialects". This suggests that there are eight or nine major Setswana dialects in Botswana, and that the dialects are clearly differentiated and spoken in specific parts of Botswana by certain ethnic communities. There is a need to review this categorisation, and to examine the extent to which the dialects are distinct from each other. There is also a need to review the extent to which dialects have remained markers of identity in a complex society born out of rapid mobility and fluidity of languages and people.

Schapera (1994: 3) claims that "most of the present Tswana tribes, both within and beyond the protectorate, are derived from one common stock", though it is not clear what "common stock" refers to, given that the groups arrived in Bechuanaland from different origins and at different times. Schapera (1994: 3) states that "in the course of time the stock broke up into many different groups, each of which again became subdivided into more and more separate tribes". There were many reasons for these break-ups: famine, and tribal and family wars caused splinter groups to move away to set up new and independent ethnic groups under a new leader. These new ethnic groups developed new dialects. This tendency of large groups to splinter away is believed to be the motivation for the name "Batswana", which Cole (1975: xxi) says is "derived from the reciprocal verb tswana (come out or go out from one another, separate) [...] reference being made either to the separation of the Tswana from the main Bantu or Sotho stock to which they originally belonged, or to the separation from one another of the various tribes which we know today". Though Otlogetswe (2014: 329) argues that there is no linguistic evidence to support this claim, there is sufficient historical evidence to support the view that secession was common among Tswana ethnic groups. Even Tlou and Campbell (1997: 96) describe Batswana as groups of people who were always splitting up and other groups were joining them.

The splitting and joining of different ethnic groups, before and after independence, make it very difficult to untangle and isolate the different ethnic groups, languages, and dialects that are found in Botswana. Schapera (1952: 28) admits that the groups are so mixed that when "dealing with the problem of classification[,] a judicious combination of different bonds such as historical, cultural, linguistic and geographical are used and that is why some Sotho groups have Nguni origin and some Nguni have Sotho origin". Batibo (1999: 10) also points out that "most speakers of Setswana in Botswana find it easier to make the distinction ethnically rather than linguistically".

 

4. Language and ethnicity in Botswana

An analysis and description of Setswana dialects in Botswana is complicated by the fact that the name for the dialect is derived from the name of the ethnic group, thus creating a misconception that one represents the other. For example, Otlogetswe (2012: 534) defines "Sekwena" as a dialect of Setswana spoken by Bakwena. This close association of names is misleading because it is often interpreted to mean that Sekwena is spoken by Bakwena and therefore *Bakwena are an ethnic group which speaks Sekwena. The perception that the Bangwato are people who speak Sengwato, the Batlokwa speak Setlokwa, or Batswana speak Setswana is quite common. Perhaps that is why the government of Botswana requires immigrants applying for Botswana citizenship to be proficient in Setswana in order to qualify for it. Furthermore, in their search for illegal immigrants from Zimbabwe, Botswana police officers also often rely on proficiency in Setswana to help them separate non-Batswana from Batswana. Needless to say, this procedure has not been effective in identifying illegal immigrants from Zimbabwe who are fluent in Setswana.

It is also interesting to note that, although dialects are often defined in terms of the ethnic groups that speak them ("Sengwato" as a dialect spoken by Bangwato, or "Selete" as a dialect spoken by Balete), the reverse does not apply. For example, the Bakwena are not necessarily people who speak the Sekwena dialect, or the Balete as people who speak Selete. Instead, ethnic groups are often defined by such factors as location, totem, chief, and culture. Kgasa and Tsonope (1996: 166), for example, define Bakgatla as an ethnic group whose totem is a baboon, while Otlogetswe defines them as people found in the Mochudi area. The latter author also defines Bangwato as an ethnic group found in the central part of the country, in Serowe [see Otlogetswe (2012)]. In other words, it seems that although a dialect can be defined by the ethnic group that speaks it, an ethnic group cannot be defined by the dialect that it speaks. There is, therefore, a need to elucidate this indirect relationship between language and social/ethnic identity. Accordingly, this paper argues that the relationship between the two is not binary, that one is actually capable of existing without the other, resulting in Batswana who do not speak Setswana, and non-Batswana who speak Setswana.

The main contention of this paper is that the relationship between a language and an ethnic or social group that speaks this language should not be interpreted directly and literally. Instead, it should be interpreted metaphorically and metonymically whereby something is used to represent a class because it has some typical qualities of that class. The requirement that immigrants applying for Botswana citizenship must speak Setswana suggests that there is a misconception of a direct relationship between language form and ethnic identity, and therefore a Mongwato should speak Sengwato, a Motlokwa Setlokwa, and a Motswana Setswana. This cannot be true because there are Batswana who do not speak Setswana, Bakalanga who do not speak Ikalanga, and Bangwato who do not speak Sengwato, yet they are full members of their respective communities. Schapera (1952: 30) points out that "the label Kalaka is commonly used by Batswana to refer to certain groups of people who speak Kalanga (a Shona language) although the group differs in origin; some belong to some branches of Shona and others to Sotho Tswana". Thus, although most research has extolled the pivotal role that language plays in the identification of ethnic identities, in reality, linguistic and social boundaries are imprecise, and the relationship between them is less predictable. Dialects have weak boundaries, and are therefore difficult to use as gatekeepers or badges of any ethnic group.

Schapera (1994), Batibo (2014), and Otlogetswe (2014) talk about the Bangwato being found in Serowe, the Bangwaketse in Kanye, Bakwena in Molepolole, Bakgatla in Mochudi, Balete in Ramotswa, and the Batlokwa in Tlokweng. Schapera even makes a distinction between "Mongwato proper" or "Mongwaketse proper" where "proper" refers to a real, original, or authentic Mongwato or Mongwaketse as opposed to a simulated one. He claims that in the Ngwato region, the "Ngwato proper" live in Serowe village while the outlying villages are inhabited by the Kgalagadi, Pedi, Kalaka, and Tswapong (Schapera 1994: 9). Schapera notes that "[i]n the Ngwaketse region, the "Ngwaketse proper" are concentrated around their chief in Kanye" (1994: 11). It is, however, unclear what Schapera means by "proper" Ngwaketse or "proper" Ngwato: is he perhaps referring to those closest to the chief or to the original members of the ethnic group? In any case, he does not relate his analysis to language.

Amongst Batswana, the authentic or proper members of an ethnic group are the ones who exhibit or epitomise certain unique features of the group, such as their history, lineage, colour, dress, dance, and even language. This group is made up of the original members as well as their descendants who have lived in the ethnic group the longest and embody all the characteristics of the group. They are the members who are also likely to speak or know what is considered the proper dialect of the group. They are also the members that language researchers would consider as the authentic speakers of the dialect from whom they can gather data. However, the extent to which one member's dialect is proper or more authentic than that of another member is highly subjective, and is a matter of personal opinion.

There are problems related to dialect authenticity. Each ethnic group contains within it many outsiders who sometimes outnumber the natives. For example, the Bakgatla natives constitute less than ^XA of the total population of the Bakgatla. About % of the Bangwato population consist of foreigners, and the relative proportion of foreigners to natives is said to be even greater among the Batawana [see Cole (1975) and Schapera (1994)]. There is also a high migration rate, and many intermarriages between the ethnic groups. There is a mitigating factor though: although the nuclear stock (original group) is small, it often has social and linguistic influence, and for that reason, it is more attractive to the foreign members who are under pressure to be accommodated. It is mainly through these outsiders and migration that dialect boundaries are destabilised: linguistic features are passed from one dialect or ethnic group to another resulting in the "proper" dialect being diluted. This therefore means that other instruments have to be used to determine ethnicity.

Schapera attempts to explain ethnic membership in Botswana, and his explanation supports Batibo's (1999: 10) observation (noted above) that "most speakers of Setswana find it easier to make the distinction ethnically rather than linguistically". Schapera (1994: x) explains that the process through which individuals become members of a tribe is by submitting to the authority of its chief. He argues that "it is primarily through their allegiance to the same chief that the different communities and individuals making up the tribe give expression to their unity. The chief is therefore not only the ruler of the tribe, but also the visible symbol of its cohesion and solidarity". Schapera (1994: x) goes on to explain that:

Normally, a man belongs to the tribe into which he was born. But he may transfer his allegiance to the chief of some other tribe. He may also go there as a fugitive from justice or oppression, or he may have been banished by his own chief, [or] he may have left for other reasons. Once accepted by the chief to whom he has appealed, he becomes a member of the latter's tribe and is allotted a place within the territory. Every tribe includes amongst its members people admitted in this way.

It is inconceivable to imagine that once an ethnic group has been formed in the manner described above, it would have a specified dialect or language features that distinguish it from other groups. Setswana ethnic groups and dialects are more diverse and fused together (linguistically, historically, and culturally) than previous studies have led us to believe. Schapera thus makes it clear that the identity of an ethnic group is determined by the chief of that group, who is a symbol and embodiment of its identity regardless of the group's language and origin. Each chief, called a "kgosi", had and still has a territory or jurisdiction within which he exercises his rule and power as well as broadens his identity. Nyati-Ramahobo (2008: 3) too points out that "land, identity and territory form the core of recognition and non-recognition of tribes in Botswana". She states that the Tribal Territories Act of 1933 created tribal territories which were drawn along ethnic boundaries. These territories have now been turned into administration districts which have maintained the name and the boundaries of the chief or ethnic group. For example, there are the Kgatleng, Kweneng, and Ngwaketse districts. Though chiefs no longer have any powers or rights over the land in these districts, as the administration of land has been given to the land boards, the chiefs still serve as symbols of identity and leadership in the districts. This suggests that independence did not dismantle tribal boundaries and identities - it affirmed them.

Furthermore, although Sekwena, Selete, Sengwaketse, and Sengwato are well-known linguistic dialects, they are hardly ever defined linguistically; rather, they are often defined socially, on the basis of the ethnic groups that speak them, as illustrated above. A linguistic definition or description of these dialects, instead of a social one, would be more useful. Otlogetswe (2014) and Batibo (1999) provide a short list of varied features such as [t /th/], [le], and vocabulary items such as "seswaa" or "kurwana" which they associate with dialects spoken by ethnic groups in the north. They also cite such features as [tl] or [tlh], [lo], and vocabulary items such as "loswao" and "moitaletsi" which they associate with dialects spoken by ethnic groups in the southern part of the country. Basically, the differences between the dialects are few [see Batibo (1999)]. However, in spite of that, traditional approaches which treat dialects as sets of autonomous skills, and strong ethnic sentiments of speakers for their dialects to be differentiated from one another have led to exaggerations and overstatements of the few linguistic differences between Setswana dialects. To date, there is no comprehensive empirical study that identifies all the essential features that really set apart each one of the Setswana dialects. An inventory of linguistic features which distinguish one Setswana dialect from another is needed to settle this debate. However, even if such an inventory existed, it would have no speakers because there are no pure speakers of any dialect. There are certainly some features which are indeed common or concentrated in some areas (in the north and south of Botswana, as indicated above), however, using them as linguistic markers of any ethnic group would be inappropriate as they do not exist in any speaker, dialect, or location exclusively, predictably, and systematically. For example, there is no study which has been done which shows that the linguistic features [nyena] or [lo] exist exclusively amongst the Bangwaketse. Further, as argued above, there is a reluctance by lexicographers and semanticists to define ethnic groups in terms of their dialects. Contrary to popular belief, location, ancestry, totem, behaviour, and cultural practices seem to be stronger ethnic markers than dialects. This makes sense in the context of globalisation, where linguistic elements are easily and constantly moved across locations and ethnic groups.

In their areas of concentration, the linguistic features are, to varying degrees, found in the speech of the pure stock and the elderly who are likely to have resisted external influences such as migration, integration, urbanisation, mass media, and education. Otherwise, the manner in which the 21^st-century Mokgatla, Mongwaketse, Mongwato, or Mokwena use linguistic features associated with their ethnic or regional group has become less systematic and less predictable. There is therefore a need for empirical research that can test the use of dialectal features by members of its ethnic group. Without such a study, our perception of a close relationship between dialect and ethnic group could simply be based on an ideological construction instead of reality. This calls for a different way of classifying linguistic features and the dialects in which they are supposed to reside. This paper contends that any approach to language that attaches linguistic features to particular groups or locations is restrictive; it fails to capture the fluidity and dynamism of the dialect and extend it beyond the essentialist territory of pure dialect or identity.

It is important that dialects and ethnic groups, as social constructs and categories of analysis, are examined within the framework of mobile communities in order to appreciate the mobility, transfer, and presence of features across ethnic and locational spaces, not as exceptions to the rule but as the norm. There is a need for studies on dialects to acknowledge not only the variations in language but also the co-existence of both shared and unshared features within single locations or groups. It is also quite ironic that dialects could be used as group markers when the concept of dialects itself emphasises variation, not uniformity and neat packaging of language. In situations where ethnic identity is a focal point, variations and discrepancies in language use are often ignored or deemphasised, and treated as exceptions in order to achieve a sense of unity and identity.

 

5. Translanguaging and dialect-levelling theories

The bases for the argument of this paper are the translanguaging and dialect-levelling theories which treat languages and their dialects as fluid and mobile entities, and are therefore difficult to maintain in certain boundaries, communities or individuals. These theories are premised on the phenomenon of super-diversity which emphasises increased mobility and expanding networks in "deterritorialized" communities [see Busch (2012)]. Lewis, Jones and Baker (2012) note that the concept of 'translanguaging' was developed in Welsh education circles in the early 1980s to refer to the practices of bilingual learners. Now, the concept has been extended from school to street, from pedagogical practices to everyday multilingual practices, from single language to multiple language behaviour. Makalela (2013) uses the translanguaging framework to describe the evolved forms of Iscamtho and Flaaitaal/Tsotsitaal (South African Black townships slang varieties) into a form that he refers to as "Kasi-taal" to demonstrate the weakening of boundaries between Sotho, Nguni, Afrikaans, and English. According to him, Kasi-taal is a hybrid and a new site of linguistic contact which challenges traditional conceptualisations of linguistic boundaries between South African languages as well as languages from neighbouring countries. He postulates that Kasi-taal language practices provide a window into future linguistic mergers and expanding linguistic identities.

Highly mobile populations, such as the ones found in Botswana, give rise to diffuse social networks which in turn promote permeable language chains. Every dialect and language is simply a linguistic melting pot with linguistic features from all other languages and dialects around it. Blommaert (2012: 2) argues that language in its actual reality only occurs in the shape of small features gathered from all kinds of languages. Such a view of language dislodges it from any society, and denies it autonomy. Likewise, Setswana dialects are intrinsically and intricately related to many other dialects and languages in a continuous linguistic tangle which does not fit into any linguistic category. Otlogetswe (2014: 327) agrees that Setswana dialects are a product of fissions and amalgamations of Tswana chiefdoms. These amalgamations call for a shift in the way that we view language, not as fixed static entities but as flexible mobile entities.

The linguistic situation in Botswana also presents an opportunity for a dialect-levelling investigation. Research on British dialects in the late 1960s and early 1980s showed that dialect levelling was quite widespread as a result of high degrees of contact between speakers commuting and relocating to and from rural and urban areas [see Trudgill (1990), Williams and Kerswill (1999), Milroy (2002), and Kerswill (2003)]. The studies also found dialect levelling to be the leading cause of loss of localised features in urban and rural English dialects. Dialect levelling occurs when individuals "interact with speakers of other varieties and in their efforts to accommodate to their interlocutors tend to avoid features that are unusual or markedly regional or which might lead to comprehension difficulties" (Trudgill 1986: 25). Such features end up being replaced with features found over a wide region, suggesting that language variation needs to be studied in a wider geographical and social context instead of narrow ethnic or geographical spaces. The studies cited above provide evidence that the contact and mobility of languages have weakened internal linguistic norms, and disrupted close-knit local speech communities.

Development, education, urbanisation, and electronic media have accelerated dialect contact in Botswana, and the differences between co-existing dialects such as Serolong and Selete or Sekgatla and Setlokwa have been weakened. Most of the so-called "major" Setswana dialects are found in Botswana's big villages (Molepolole, Kanye, Serowe, Mochudi, Tlokweng, Maun, and Ramotswa) which are comparatively modern, well-developed, heavily populated, and quite cosmopolitan. These villages (except Serowe and Maun) are also in close proximity, not more than 100km from each other, making it very difficult to isolate the speakers and the dialects from each other. Though most Batswana still identify themselves with their ancestral villages, and would prefer to continue to work and live in them, the job market and the Botswana government's policy on the allocation of tribal land are not bound by any ethnic boundaries. In any event, the Batswana are free to live and work anywhere in the country. There is thus unlimited contact between the dialects, which makes it easy for one group to have some knowledge of the language habits of the other groups. The Botswana linguistic situation supports Makoni and Mashiri's (2007) observation that there is increased movements of people between and within nation states in the 21^st century, which has correspondingly resulted in the movements of languages and the shifting of traditional boundaries.

Garcia (2009) and Makalela (2013) argue that it is difficult to determine where one language or dialect ends and where the next one begins. This is due to the fact that languages and their dialects are not discrete, and do not have clear boundaries between them. They exist in the form of an amorphous continuum in which speakers are constantly putting together sounds, words, and sentences from different languages or dialects that they know in order to achieve their communicative goals. Lemke (2002: 85) also believes that if languages were not politically and socially prevented from mixing, they would mix and dissolve into one another. But strong political and ideological pressures exist to keep them pure and separate. Languages do not have boundaries - they cross ethnic, historical, and social lines. Romaine (1994: 12) sees the classification of linguistic elements into discrete categories as superficial. She argues that "the very concept of discrete languages is probably a European cultural artefact fostered by procedures such as literacy and standardization but any attempt to count distinct languages is an artefact of classificatory procedures rather than a reflection of communicative practices". Dialects are simply social constructs, just like gender, whose boundaries are determined and defined by individual societies, place, and time. Dialects are neither fixed nor stable: they are socially constructed and negotiated. They therefore vary or shift with time, and with each individual. This point is also made by Blommaert (2006: 512), who sees "language as the result of ideological construction and therefore involves power, authority and control", and by Bakhtin [in Holquist (1981: 270)], who sees language "not as something given but is always in essence posited and at every moment of its life opposed to the realities of heteroglossia".

Batibo's (1999) comparison of the basic vocabulary of Setswana dialects indicates that they share more than 80% of their basic vocabulary, and that they have high mutual intelligibility of 95-99%. He argues that "the speaker's reaction to each other's dialects is normally not that of lack of inter-comprehension, but rather that of unfamiliarity in the pronunciation or use of certain words" (Batibo 1999: 7). He also agrees that the Setswana dialects:

[...] have been in close contact and that a lot of harmonization has taken place. This is largely true considering the high mobility among the speakers as well as the influences of [the] standardized form, particularly in the school system and in official media.

(Batibo 1999: 10)

Thus, although there has been extensive research and discussion of Setswana dialects, no major differences have been found between them, something that Batibo (1999: 5) says indicates that there are no major, distinct dialectal variations or groupings. Cole (1975: xix) agrees that it is surprising how uniform the language is throughout the Tswana field.

Though there are a few minor linguistic differences between some dialects of Setswana and some dialects of languages that co-exist with Setswana, very strong sentiments exist that are in favour of keeping the dialects and the languages apart. This is the case with some dialects of Setswana and Setswapong, and Shekgalagadi and Sebirwa, which seem to be converging [see Batibo (1998)]. In spite of the growing linguistic similarities and the high degree of intelligibility between the speakers of these languages, there are strong political and ideological pressures to keep them in different linguistic groups. For example, some dialects of Ikalanga (e.g. Dalaunde in Serowe), Shekgalagadi (in and around Molepolole), Sebirwa (in Bobonong), and Setswapong (in Maunatlala and Mogapi) could be considered dialects of Setswana. However, for historical reasons, emphasis on dialect differentiation is used to maintain the social and ethnic distance between them. During the period of territorial expansion, Sotho Tswana groups dominated these other groups, and suppressed their ethnic identities. Batibo (1998: 10) points out that there is a high lexical relationship between these languages and Setswana: 83% between Setswana and Shekgalagadi, 84% with Sebirwa, and 86% with Setswapong. Malepa (2008: viii) also admits that Sebirwa in Maunatlala and Mogapi has been assimilated by the Sengwato which is politically and numerically powerful. She states that:

though the Maunatlala and Mogapi people perceive themselves to be speaking Sebirwa, they can be best described to be speaking a version of Sengwato as they have accumulated a lot of Sengwato vocabulary. Lexical evidence shows that they are closer to Sengwato than other Sebirwa dialects. What is left among these groups is the Sebirwa pronunciation, otherwise the vocabulary is almost similar to Sengwato.

(Malepa 2008: 66)

Pronunciation is a distinguishing feature of a dialect. However, the extent to which it is rated against other distinguishing features such as vocabulary varies from one individual to another. Though the Setswapong, Sebirwa, and Sengwato dialects in the area have evolved and become close, the speakers still use whatever minor differences that exist between them not only to assert their autonomy but also to symbolise and dramatise their historical, ethnic, and social differences. During the period of territorial expansion and ethnic wars, Bakgalagadi, Bakalanga, Babirwa, and Batswapong were conquered and oppressed by the Setswana ethnic groups. Thus, historically and socially, there is a lack of amity between these groups and Setswana, and it is this lack of amity that has created resistance towards a common social identity in spite of the harmonisation of the languages. Bagwasi (2016: 10) argues that it is the desire to belong to:

any one language that constitutes the major reason for language isolation and compartmentalization and therefore for resistance by indigenous communities to break down language barriers. Speakers of languages often see their languages as fences protecting their identity; these fences separate and shield them from outsiders, and they are often not ready to weaken them.

Dialects are social constructs used to include people in or exclude people from certain social or regional groups.

This linguistic situation described above is not unique to Botswana. Crystal (1987: 284) describes a similar situation involving the Norwegians, Swedes and Danes whose languages are linguistically close and, therefore, to a great extent, mutually intelligible. However, for non-linguistic reasons, the language varieties are assigned to different national identities, so that Norwegians speak Norwegian, Danes speak Danish, and Swedes speak Swedish in order to merge political and linguistic identities. Crystal (1987: 284) also notes that there are many other such cases (Hindi versus Urdu, Twi versus Fante, Xhosa versus Zulu) where political, ethnic, religious, literary or other factors force a division yet, linguistically, there are very few differences. These political and ideological boundaries fail to account for the shared features and close language relations that arise from origin, constant mobility, and exchange of speakers and linguistic elements.

Given the high mobility of Setswana dialects, and the lack of major differences between them, this paper suggests that it would be more useful to approach the language and dialect situation of Botswana from the perspectives of translanguaging and language levelling. These perspectives highlight convergence and diffusion of languages and dialects rather than their divergence. All the existing research on Setswana dialects has focused on divergence and delineation of the dialects [see Batibo (1999) and Otlogetswe (2012)]. The uniformity, harmonisation, and high mutual intelligibility among the Setswana dialects offer a window for linguistic mergers that would be interesting to explore. Such a treatment of Setswana dialects in Botswana will not only allow for the dialects to be studied from a broader geographical and social context, but will also offer a different perspective to understanding dialects. This paper does not in any way suggest complete dialect levelling. Complete levelling is rare, and is often met with resistance by speakers who wish to use their dialects as symbols of identity. Therefore, there will always be some noticeable, albeit minor, linguistic differences in the lexicon and pronunciation amongst speakers which will be retained and used as markers of identity.

 

6. Conclusion

Whilst previous studies have elucidated clear divisions between Setswana dialects, and have portrayed a link between dialects and ethnic groups, this paper has proposed a review of such relations. It has laid the background for an alternative approach to the treatment of dialects. The paper recommends that a comprehensive and empirical study of Setswana dialects be carried out. This proposed study should investigate all the essential and necessary linguistic features that distinguish each dialect from one another, and all the features that these dialects share. A linguistic instead of a social-differentiation approach is needed. Batibo (1998: 3) agrees that "the early studies on Setswana dialectal variation have normally been based on the historical splits between the different groups". There is also a need to review the perceived relationship between language forms and social categories, especially in the context of mobile societies and fluid identities. The relationship between language forms and social categories has been so extolled that it is possible to overlook mismatches when they occur.

This paper calls for a different approach to the study of dialects, an approach that represents the complexities that have arisen from the fluidity of languages, and the rapid mobility of the identities that are found in Botswana. The paper thus recommends translanguaging and dialect-levelling approaches in which languages and dialects are not seen as definitive and fixed entities, but rather as active and fluid systems that defy compartmentalisation and confinement. It argues that linguistic features need to be studied in a broad context, everywhere they exist, and by whomever uses them. Accordingly, variations within Setswana would best be captured if the variations are studied in a broader context involving all other linguistic and social situations in which they exist. Studies that confine themselves to insular locations and specific communities fail to capture the fluid nature of language, as it transcends ethnic and geographical boundaries. The lack of major dialectal differences between Setswana dialects and other languages in Botswana is evidence that language transcends ethnic and geographical boundaries.

In this era of globalisation, there is a need for descriptions of language and dialect groups that emphasise the mobility instead of fixity of language, that present language as an active and mobile object "no longer tied to an organic, specific community, residing in a particular space but moving around such places and communities in intensive ways, on the rhythm of globalizing flows of commodities, people, messages, and meanings" [(Blommaert 2012: 2); also quoted by Bagwasi (2016: 205)].

 

References

Anderson, L.G. and T. Jason. 1997. Languages in Botswana. Gaborone: Longman Botswana.         [ Links ]

Bagwasi, M.M. 2016. A critique of Botswana's language policy from a translanguaging perspective. Current Issues in Language Planning 18(2): 199-214. https://doi.org/10.1080/14664208.2016.1246840        [ Links ]

Bakhtin, M.M. 1981. Discourse in the novel. In M. Holquist (ed.) The dialogic imagination. Austin: University of Texas Press. pp. 259-422.         [ Links ]

Batibo, H.M. 1998. A lexicostatistical survey of the Bantu languages of Botswana. South African Journal of African Languages 18(1): 22-28. https://doi.org/10.1080/02572117.1998.10587183        [ Links ]

Batibo, H.M. 1999. A lexicostatistical survey of the Setswana dialects spoken in Botswana. South African Journal of African Languages 19(1): 2-11.         [ Links ]

Batibo, H.M. 2014. Plurilingualism in Southern Africa and the emergence of the Setswana language. Boleswa Journal of Theology, Religion and Philosophy 4(2): 275-292.         [ Links ]

Blommaert, J. 2006. Language ideology. In T. Todorov (ed.) Handbook of language and linguistics Vol. 16. Chicago: Elsevier. pp. 510-523.         [ Links ]

Blommaert, J. 2012. Sociolinguistics and English language studies. Working Papers in Urban Language, Paper 85. Available online: http://www.kcl.ac.uk/sspp/departments/ (Accessed 20 February 2019).

Busch, B. 2012. The linguistic repertoire revisited. Applied Linguistics 33(5): 503-523. https://doi.org/10.1093/applin/ams056        [ Links ]

Cole, D.T. 1975. An introduction to Tswana grammar. Johannesburg: Longman.         [ Links ]

Crystal, D. 1987. The Cambridge encyclopaedia of language. Cambridge: Cambridge University Press.         [ Links ]

Garcia, O. 2009. Education, multilingualism and translanguaging in the 21^st century. In A. Mohanty, M. Panda, R. Phillipson and T. Skutnabb-Kangas (eds.) Multilingual education for social justice: Globalising the local. New Delhi: Orient Blackswan (Former Orient Longman). pp. 128-145. https://doi.org/10.21832/9781847691910-011        [ Links ]

Kerswill, P. 2003. Dialect levelling and geographical diffusion in British English. In D. Britain and J. Cheshire (eds.) Social dialectology: In honour of Peter Trudgill. Amsterdam: Benjamins. pp. 223-243. https://doi.org/10.1075/impact.16.16ker        [ Links ]

Kgasa, M.L.A. and J. Tsonope. 1996. Thanodi ya Setswana. Gaborone: Longman Botswana.         [ Links ]

Lemke, J. 2002. Language, development and identity: Multiple timescales in the social ecology of learning. In C. Kramsch (ed.) Language acquisition and language socialization. London: Continuum. pp. 68-87.         [ Links ]

Lewis, G., B. Jones and C. Baker. 2012. Translanguaging: Origins and development from school to street and beyond. Educational Research and Evaluation: An International Journal of Theory and Practice Vol. 18(7): 641-654. https://doi.org/10.1080/13803611.2012.718488        [ Links ]

Makalela, L. 2013. Translanguaging in Kasi-taal: Rethinking old language boundaries for new language planning. Stellenbosch Papers in Linguistics Plus 42: 111-125. https://doi.org/10.5842/42-0-164        [ Links ]

Makoni, S. and P. Mashiri. 2007. Critical historiography: Does language planning in Africa need a construct of language as part of its theoretical apparatus? In S. Makoni and A. Pennycook (eds.) Disinventing and reconstituting languages. Clevedon, UK: Multilingual Matters. pp. 62-89. https://doi.org/10.21832/9781853599255-005        [ Links ]

Malepa, M. 2008. A Lexical Study of Sebirwa Language Varieties Found in Botswana. Unpublished Master's thesis, University of Botswana.         [ Links ]

Milroy, L. 2002. Introduction: Mobility, contact and language change - Working with contemporary speech communities. Journal of Sociolinguistics 6(1): 3-15. https://doi.org/10.1111/1467-9481.00174        [ Links ]

Ngcongco, L.D. 1979. Origins of the Tswana. Pula Journal of African Studies 1(2): 21-46.         [ Links ]

Nyati-Ramahobo, L. 2008. Minority tribes in Botswana: The politics of recognition. London: Minority Rights Group International. Available online: https://www.refworld.org/pdfid/496dc0c82.pdf (Accessed 14 May 2019).

Otlogetswe, T.J. 2012. Tlhalosi ya Medi ya Setswana. Gaborone: Medi Publishing. https://doi.org/10.5788/23-1-1228        [ Links ]

Otlogetswe, T.J. 2014. Dialectal variations among Tswana peoples of South Africa and Botswana: A historical perspective. Boleswa Journal of Theology, Religion and Philosophy 4(2): 326-346.         [ Links ]

Romaine, S. 1994. Language in society: An introduction to sociolinguistics. Oxford: Oxford University Press.         [ Links ]

Schapera, I. 1952. The ethnic composition of Tswana tribes. London: Berg Publishers.         [ Links ]

Schapera, I. 1994. A handbook of Tswana law and custom. Oxford: James Currey Ltd.         [ Links ]

Southerland, R.H. and F. Katamba. 1996. Language in social contexts. In W. O'Grady, M. Dobrovolsky and F. Katamba (eds.) Contemporary linguistics: An introduction. London and New York: Longman. pp. 540-590.         [ Links ]

Tlou, T. and A.C. Campbell. 1997. History of Botswana. Gaborone: Macmillan Botswana Publishing Company.         [ Links ]

Trudgill, P. 1986. Dialects in contact. Oxford, UK: Blackwell.         [ Links ]

Trudgill, P. 1990. The dialects of England. Oxford, UK: Blackwell.         [ Links ]

Williams, A. and P. Kerswill. 1999. Dialect levelling: Change and continuity change in Milton Keynes, Reading and Hull. In P. Foulkes and G.J. Docherty (eds.) Urban voices: Accent studies in the British Isles. London: Arnold. pp. 141-162.         [ Links ]