Address inversion in Swahili: Usage patterns, cognitive motivation and cultural factors

3.1 Metonymy a concept for its opposite

In previous research, the phenomenon of address inversion in the kinship dyad was sometimes explained as framing a junior in terms of a senior (e.g., Rieschild 1998). From the cognitive perspective, this type of analysis would involve a metaphorical mapping between the source domain of a senior and the target domain of a junior. For certain marginal cases of address inversion, this kind of account might be accurate, for example, when a junior is believed to resemble a senior, or when the two persons share the same name. In some cultures, beliefs in reincarnation can support a metaphor-based analysis. For example, in the Igbo community in Nigeria a young girl might be called ‘grandma’ when she is believed to be an incarnation of her late grandmother. ^[4] However, all of these cases are distinct from the ones discussed in this article since they involve referential terms and a senior in the dyad might even be deceased. Typical cases of address inversion are hard to defend as metaphor, because the hypothetical source and target domains do not share convincing schemas for a mapping to take place.

Ubiquity of the inverted address among the languages of the world calls for a more adequate explanation. Building on Vosshagen (1999), I propose that address inversion, as well as other similar reversal strategies have cognitive motivation and can be formally analyzed by means of metonymy: a concept stands for its opposite.

Vosshagen (1999) observes that under certain conditions a word may have a figurative meaning opposite to its literal or basic meaning. For example, in some slang varieties of English, the adjective bad is used in the sense of ‘good’, or nasty as ‘excellent’. The motivation behind a reversal of the meaning in these examples lies in social and cultural attitudes (1999: 296). Vosshagen accounts for these changes by proposing a metonymy: a concept stands for its opposite on the basis that the relation between lexical as well as conceptual opposites involves close mental contiguity. ^[5] The evidence for such a connection comes from psycholinguistic experiments, substitution errors, language acquisition, irony and diachronic semantic change.

From a more general perspective, metonymy based on inversion relates to the notion of opposition understood as a binary construal of comparison (Paradis and Willners 2011). Conventionalized antonyms of a special lexical status, such as, good–bad, or heavy–light, which are understood as opposites, create a bounded configuration in conceptual space divided by a boundary. Paradis and Willners (2011) also demonstrate that the comparison of antonyms, in addition to their structural properties, should involve contentful meaning structures which makes it possible to account not only for the most typical pairings of opposites, but also for less ‘perfect’ ones which have lower frequency in textual co-occurrence and in psycholinguistic elicitation experiments.

A binary schema can be postulated for an asymmetric kinship dyad in which a senior and a junior are conceptualized as opposites to each other. Kin relations in such a dyad “go together” and are connected within one mental space just like antonyms are. A binary configuration linking a mother (father) with her (his) child, or a grandparent with a grandchild, is supported by fixed conventionalized expressions, as mother and child etc. In some languages, this sort of a dyad may even have a completely lexicalized form of an unanalyzable morpheme, as it happens in the linguistic area of Australia and the Western Pacific (Evans 2006). In Swahili and other languages which exhibit address inversion, one part of the conceptual space can be substituted for another by means of the abovementioned metonymy, which has its particular instantiation based on kin relations: a senior (relative) stands for a junior (relative). Two cross-linguistic types of address inversion require two slightly different formulations of specific metonymies. Thus, languages of the “Arabic type” follow the pattern: mother/father stands for daughter/son, grandfather/grandmother stands for granddaughter/grandson, etc. In languages of the “Swahili type”, the relation of opposition is modified with respect to the target in order to address a junior with a term appropriately marked for gender, hence only: mother stands for daughter, father stands for son, etc., but not: *mother stands for son, or *father stands for daughter. I assume that gender adjustment is imposed as an additional constraint guaranteeing that an address term will be congruent with the natural gender of an addressee. In formal terms, this can be analyzed as the operation of another metonymy substituting mother for father, etc., that is a person in a certain kinship relation for a person of the opposite sex in the same (or analogous) kinship relation.

A question can be asked, why metonymies operating in the domain of the kinship dyad are unidirectional in the sense that the source is always a senior and the target – a junior? One possible response to this question relies on the diachronic development of address inversion from baby talk (Braun 1988). Once this directionality has been originally established, it becomes conventionalized and does not change. But there are also cultural reasons which comply with this particular direction of mapping (cf. Section 6).

It is worth pointing out that address inversion patterns with other strategies employed for communicating positive emotions which involve either lexical or grammatical reversal of the linguistic norm, as discussed in the following Section 3.2.

3.2 Other cases of reversal in emotive language

Cross-linguistically, words of endearment draw on source domains associated with positive feelings or pleasant things. However, in some languages, affection may be expressed by means of lexemes which have ugly or even disgusting meanings in their literal usage. Particularly frequent words of endearment of this kind are found in Dutch, cf. bobbel ‘blackhead’, poepje ‘little shit’, scheetje ‘fart’, but also in other languages (Perlin and Milewska 2000). Although words of this kind do not constitute antonyms to any specific lexemes of positive connotation, they nevertheless establish a binary conceptual space of “pleasant/ugly things” along the axiological scale. The metonymy a concept stands for its opposite makes it possible to use a reversed counterpart, that is, a concept situated at the “ugly” end of the scale for a concept situated on the “pleasant” edge. Metonymic anchoring provides cognitive motivation for the use of these lexemes in the “opposite” meaning, while their unusual form draws attention and makes them highly expressive. A similar strategy has led to the conventionalization of English adverbs, such as terribly or awfully as intensifiers of positive notions (cf. Vosshagen 1999: 302).

Affection can also be conveyed by switching grammatical gender which takes place in address terms or verbal forms. In some languages, a switch applies bidirectionally: a man is addressed with feminine forms and a woman with masculine ones, in others, only one gender is associated with affectionate meaning and the change operates unidirectionally. Modern varieties of Arabic are known for a full range of grammar manipulation: bidirectional switches take place in verbs, nouns, adjectives and pronouns (e.g., Ferguson 1964). Bidirectional switches restricted to proper nouns are possible in Polish (e.g., Kraska-Szlenk 2010) or Marathi (Ferguson 1964). Unidirectional gender reversal is found in Modern Hebrew, where masculine forms can be used towards women (Tobin 2001). The opposite is true of Amharic, where verbs and pronouns of the feminine gender express affection directed to males (Wołk 2008: 89–90). Gender switch constitutes an exact parallel to lexical inversion and has a similar metonymic basis. Although it operates within an abstract category of grammatical gender, ‘masculine’ and ‘feminine’ features are conceptually united within a binary construal and understood as opposites to each other.

Examples brought to attention in this section have demonstrated that address inversion observed in kinship terms has many parallels in other linguistic behaviors and can be understood as metonymy operating within a domain of opposites. This strategy increases expressiveness, and therefore ideally complies with manifestation of affection.

5.1 The data and methodology

While kinship terms often occur as address formulas when they express real kinship relations holding between a speaker and an addressee, their use in the form of inverted address is very infrequent. ^[6] The setting of such uses is limited to a narrow circle of one’s family and to rather intimate situations. For these reasons, data is hard to obtain from real usage contexts. Likewise, it would be difficult to elicit data from an electronic corpus, because, without a larger context, relevant terms are indistinguishable from other uses of familial address. The research on inverted address in other languages was usually based on authors’ own intuition, if they were native speakers of described languages, or on questionnaires distributed among native speakers. Both of these methods have their drawbacks. The former introspective method might be biased toward the researcher’s own habits. In the case of the latter, language consultants tend to provide only a small set of prototypical examples, and sometimes find it difficult to interpret their own and other people’s emotions associated with the use of inverted address. In addition, both of these methodologies face the problem of significant inter-speaker variation which is characteristic for using words of endearment in general.

In order to overcome the above difficulties two distinct methodologies have been applied, based on the data drawn from a corpus of literary texts and from interviews with language consultants, respectively. The first method made it possible to determine a wide range of pragmatic contexts in which address inversion occurs. These will be discussed in Section 5.2. With this background, the data obtained from interviews, analyzed in Section 5.3, will help to determine sociolinguistic factors responsible for inter-speaker variation.

5.2 Pragmatic contexts

The data for analysis has been elicited from modern Swahili literary works, such as contemporary plays, novels and short stories written by Tanzanian writers (see Swahili Sources). All of these texts contain realistic dialogues from which address forms have been excerpted. Although such data constitutes only a substitute for natural conversations between Swahili interlocutors, for the purpose of the present study they provide even better source material than real speech acts, because an author creates not only a full situational context in which a particular linguistic form is used, but also describes characters’ emotions and intentions. Therefore, a researcher can determine with high probability what feelings are conveyed by a constructed address form – something not directly accessible to an observer of a real linguistic event – and can scrutinize other illocutionary and perlocutionary effects. Generally speaking, all contexts in which inverted address has been observed are marked for intimacy and positive emotions, whether already present in a social setting or created by its use. The collected data point to several typical situations and scenarios. Two address terms: baba – directed to a son, and mama – to a daughter, have turned out to be the most frequently used inverted forms in the literary texts. These will be discussed first.

In the Swahili-speaking environment, the most neutral way of addressing one’s own child is by the child’s first name, although some people of a conservative conduct find this form inappropriate when a child achieves maturity. A more cordial way is calling a child mwanangu ‘my child’. When compared to these two, the inverted address stands out as the most affective form of address.

Different address terms may co-occur interchangeably in one utterance, as illustrated in (1), which is said to an adult son by a mother who greets him when he pays a visit in his home village.

In the subsequent conversation which follows the opening greeting in (1), the mother uses baba eight times and mwanangu eleven times in total. These two address forms do not have random distribution: the initial high rate of baba slowly goes down, while the uses of mwanangu increase. Such distribution strongly correlates with the woman’s emotions calming down as the style and topics of the conversation gradually change.

The dialogue can be divided into three parts which roughly correspond to three different pages of the book, cf. Table 2. At the very beginning, the mother and the son express mutual joy of seeing each other. This is the most affectionate part of the conversation, full of exclamations and conventionalized greetings. Four instances of baba and one of mwanangu appear in this section (page 27). Next, a casual talk follows, in which further greetings combine with exchange of family news. The proportion between baba and mwanangu begins to even up and the latter address becomes dominant once the conversation takes a different turn and a specific topic is introduced (the bottom of page 28). During an exchange of conventionalized questions about everybody’s wellbeing, the mother begins to give a report about a real health condition experienced by her son’s grandmother. The young man happens to be a medical doctor, and therefore the topic of his grandmother’s illness and her treatment is continued in some detail on the next page until the conversation is interrupted by a newcomer. The son’s professional, matter-of-fact style affects his mother who also calms down her emotions. In this last part (page 29), baba is said only once, while mwanangu as many as six times.

Descriptions of emotions distinguish temporarily felt affects, observable in people’s behavior in a particular situation, from a broader category of feelings (Besnier 1990: 421). While the aforementioned example provides an illustration of the former case, a positive correlation between the use of the inverted address and affection can be likewise observed when cordial feelings toward someone have a permanent character. An example in (2), coming from the same novel, demonstrates that the author deliberately constructs different address strategies to convey a message of unequal treatment of two sons by their father. He adores the younger son and this is precisely the one whom he calls baba, even without a special reason. But he does not approve of his elder son, whom he calls by his first name Mustaffa, by impartial bwana ‘Sir’, or does not use any title to address him, as in (2a) which sounds rather rude in the context of greeting a son unseen for a long time. Occasionally, however, the father can reward Mustaffa with the tender address baba, if the son “deserves” it, as shown in (2b), uttered when the father is suddenly pleased and surprised by the news delivered by his son. These lines are then followed by the narrator’s comment, cited in (2c), describing the young doctor’s feelings after being called baba.

The cases discussed thus far illustrate the illocutionary effects of the inverted address, used for the purpose of expressing one’s own emotions. At the same time, a performed speech act exhibits some perlocutionary force in the sense that it makes an addressee feel good, as explicitly stated in (2c). In other pragmatic contexts, this function of the inverted address becomes primary. Directives, such as requests, demonstrate this most clearly, but these will be discussed in the following Section 5.3, since examples of this kind were usually given by the language consultants. The literary sources, however, provide a variety of other contexts of perlocutionary acts, as discussed below.

In many contexts, address inversion accompanies words of compassion and support offered by a speaker to a junior facing a difficult situation, as in all examples in (3). The father’s words said to his daughter in (3a) look like a direct command, but they are mitigated by the title mama which could be interpreted here as the inverted address. The context makes it very clear that the father is full of compassion for the girl who is too little for the household chores imposed on her. The words in (3b) are uttered by a mother to her adult daughter, with whom she is reunited after many years of separation. In (3c), the inverted address has the form mjomba (literally ‘maternal uncle’) and is addressed to a nephew by his uncle who is the only family member attending the young man’s wedding. The uncle offers his support in the most parsimonious way by using two tender address forms: mwanangu ‘my child’ and the inverted familial title mjomba.

While perlocutionary effects in the examples of (3) involve influence on the interlocutors’ emotions, in other cases address inversion strategies help to stimulate desired actions. The words in (4a) are said by a mother to her grown-up son. The address title baba constitutes a part of gentle manipulation the woman uses to persuade her son to marry a suitable girl. In (4b), the father asks his son for forgiveness after he brought about a major family crisis and caused a lot of emotional pain to his son. The father has to find words strong enough. Not only does he call his son baba and mwanangu ‘my child’, but eventually he seeks for extraordinary, creative means involving an address inversion of the word mzazi ‘parent’. Apparently this unconventional form is meant to convey more expression and perlocutionary force than the conventionalized address inversion of the title baba.

To conclude this part, it is worthwhile to emphasize that in spite of low frequency, address inversion occurs in various pragmatic contexts. Most of them, however, are limited to very special occasions of expressing strong positive emotions towards an addressee.

5.3 Additional data and variation

First interviews with Swahili consultants were conducted with Tanzanian students living in Europe and took place on an occasional basis between 2012–2015. Several speakers – both males and females in their twenties or early thirties, were asked to share their opinions on the use of specific terms of address within their families while the interviewer took notes. In general, these interviews confirmed the usage patterns observed in the written sources. They also revealed a lot of inter-speaker variation in the contextual use of specific terms. Thus more systematic interviews devoted to the topic were conducted by the author during fieldwork in the town of Zanzibar. The location of the fieldwork was chosen on the basis of previous observations that the inverted form of address occurs mostly in coastal varieties of Swahili. The goal of the research was to determine how much variation was found in a geographically limited and culturally homogeneous culture, of course to the extent Swahili culture can be called homogeneous at all.

Given that the issue had already been researched and basic usage patterns had been revealed, a form of semi-structured qualitative interviews was chosen as optimal. The interviews took place in January-February 2017 and were carried out on the premises of the State University of Zanzibar (SUZA) or in workplaces or private houses of the interlocutors. All consultants were born either on Zanzibar or on the neighboring island of Pemba, and for the majority of them the Zanzibar archipelago has been home to their ancestors for many generations back. Few interviewed had roots in other parts of Tanzania and one person had Oman-Comoro ancestry. Most of the interviewed were either teachers, office workers or students at SUZA. Others were self-employed persons doing all kinds of business. The interviews lasted for 20–30 minutes and took place in the form of conversation conducted in Swahili, while the interviewer took notes. They were preceded by informal small talk in order to put the speakers at ease. The interviewer would explain her research asking the consultants to provide ways of address in their own family circles, but sometimes the interviewed added comments on expressions said by other people. Either by their own choice or prompted by the interviewer, they also indicated which terms of address were used more typically than others and which were spoken in specific situations or had particular connotations. The consultants were not requested to comment on what they felt when being addressed (or addressing someone else) with one term or another, but sometimes they made such comments, too. Seventeen persons between 19–61 years old were interviewed, eight women and nine men. In one case, a consultant was accompanied by her nine-year old daughter who also made some comments.

In general, the fieldwork research has confirmed that variation is very common among Zanzibarian speakers. While some freely admit using inverted address and provide examples as well as interpretations of their usage patterns, others deny ever using such forms. No apparent correlation has been found between the use of the inverted address and sociolinguistic factors, such as age, gender or occupation, and only the type of personality seemed to be a factor enhancing its use. Although not much data has been collected from other Swahili-speaking parts, the consultants from Dar es Salaam or mainland Tanzania gave examples and explanations very similar to those given by the Zanzibarians. Thus, all the data will be discussed jointly, indicating the source by giving the gender of the consultant (M, F), his/her age and the place of birth and living.

During the interviews, the use of the inverted address was discussed in a larger context of how family members addressed each other. All speakers have admitted that addressing a junior by his/her first name does not convey any particular connotation and can be used in various situations. ^[7] This was the first option which the speakers gave as a response to the question “How do you call your child?” or “How does your mother/father call you?”. In addition, the expressions mwanangu or mtoto wangu ‘my child’ was spontaneously given by the interlocutors, when requested to provide alternative terms. There was also an agreement as to the expression mwanangu being a more cordial form of address than the first name of the child. The consultants usually did not produce inverted address forms spontaneously, but when prompted, immediately confirmed that such forms did occur and demonstrate affection. The word upendo ‘love’ was most typically used to describe the expressed feeling. One woman (28, Zanzibar) made the following comment on how she felt when being called mwanangu or mama by her mother: unajisikia furaha, kama anakubeba, hata nikiwa na hasira nitakuwa mpole ‘you feel joy, as if she carried you, even if I am angry, I will become gentle’. A young man from Dar es Salaam recalled that when he was a teenager, he felt embarrassed and did not like to be called baba by his mother in the presence of his friends, which clearly points out that the inverted address is a very affectionate expression.

When asked to provide contexts in which the inverted address occurs between parents and children, the speakers spontaneously produced requests, as in the examples in (5), or apologies, with a typical example as: samahani, baba/mama ‘I am sorry, baba/mama’.

While perlocutionary speech acts were given as typical examples of address inversion directed to adult children, no specific explanation was looked for by the consultants when describing address inversion used by grandparents towards their little grandchildren. To many speakers of Swahili, affectionate bibi and babu did not need justification and was treated as a typical nominal address said to children by grandparents. When asked for contexts of use, the consultants would bring up greetings and other casual situations and rarely requests, as in (6c) directed to a grown-up grandchild.

Address inversion has also been used by the consultants with other kinship terms, such as, mjomba ‘maternal uncle’ and shangazi ‘paternal aunt’, as well as with terms characteristic of the Zanzibar archipelago: ami ‘paternal uncle’, haloo ‘maternal aunt’, mpuu ‘paternal aunt’ and chachi ‘paternal aunt’ (a rare synonym of shangazi). Out of all these terms, mjomba was most accepted by the consultants, to the extent that some persons fully acknowledged its usage in the inverted address while declining other terms. It was the only kinship term which was unanimously judged as acceptable in referential use of the inverted meaning and which could be used toward or about boys and girls, as well. Thus, the word mjomba, which in its basic sense translates as ‘maternal uncle’, has become fully polysemous and its additional meanings of ‘nephew’ and ‘niece’ are, at least in the speech of some Swahili speakers, no longer restricted in use as other forms of the inverted address. This semantic change of mjomba will be further elaborated in Section 7. As to other abovementioned kinship terms, fewer speakers accepted them in the inverted address, and if they did, they would make a point of a gender restriction, for example, that haloo could be said to a girl only and ami to a boy. Still fewer speakers accepted the referential use of these terms. Language consultants also confirmed that the inverted address could be used with other nominal address forms, such as, bosi ‘boss’, chifu ‘chief’, or shehu ‘sheikh’.

However, not all speakers admitted using inverted terms of address as their habit. Some categorically denied producing any such forms, while acknowledging their use in the speech of other people. For example, a 58-year old man from Pemba has said that neither he nor his wife would call their children by mwanangu or baba/mama and would not say bibi or babu to their grandchildren. Asked for the reasons, he described his wife as mkali ‘strict’ and added that children should not be spoilt by words of endearment. When pressed to specify, how he would show affection towards his children, he responded that it was by action (e.g., by providing money and other support), not by words. Sometimes the consultants commented on differences in behavior within their families, for example, by saying that one parent or grandparent would tend to use the inverted address more than the other parent/grandparent.

The consultants were also asked to provide other words of endearment used in their families, with a special focus on address forms directed to spouses and small children. This helped to determine whether the observed variation in the use of address inversion correlated with personality type and general verbal behaviors of the interviewed. Some speakers gladly gave examples of tender titles of address, while others denied using any such expressions making a clear statement about their preference for neutral or even reserved ways of calling their family members. Examples of endearment words most commonly given by the consultants are those based on the verbal root pend ‘love’, e.g., mpendwa wangu ‘my dear’, kipenzi changu lit. ‘my beloved-dim.’, mpenzi wangu wa moyo lit. ‘my beloved of heart’, but other expressions have also been mentioned, e.g., furaha ya moyo wangu ‘joy of my heart’, nyota yangu lit. ‘my star’, mwangaza wangu ‘my light’. English loanwords, such as darling, sweetie or honey, were mentioned by some speakers as endearment words used to spouses, even in public. Other speakers, on the contrary, claimed that they would not go beyond purely conventionalized forms of address and did not use endearment words in calling their spouses and children.

A lot of variation was observed especially in ways of calling spouses. While many Swahili speakers of younger, but also middle-aged generations find no problem in addressing their spouses by their first names in private and in public, some never do it and consider it disrespectful to call any adult by their first name. While talking about their more conservative kin, the Zanzibarian consultants gave an example of a religious man shehe who would call his wife bibi ‘lady’ and who would be addressed with the term bwana ‘Sir’ by his wife and children. A distant, respectful form of address may also include professional titles, e.g., mwalimu Ali ‘teacher Ali’ said to a husband. For other speakers it was acceptable to call a wife by her first name, but not a husband for whom a more respectful address form should be used.

In general, the interviews have confirmed a positive correlation between using the inverted address and using other words of endearment. Those who admitted using address inversion, also provided examples of other tender expressions used by them within their families. Likewise, those who restrained from using endearment words, would not employ the inverted address in particular. These facts are naturally interpreted to the effect that some persons are either more cordial than others or more outspoken in expressing positive emotions. However, some of those interviewed preferred formal terms of address among adults, but were open to cordial address terms directed at children. Thus a third type of personality emerges from the data: a person whose behaviors are rather reserved by imposition of the conservative Swahili culture and manifest themselves in distant forms of address with a spouse and adults, but who is at the same time able to reveal his/her emotions in addressing their own children and grandchildren, or juniors in general, with cordial forms of the inverted address and other words of endearment.

5.4 Non-canonical patterns of inversion

The patterns analyzed in Sections 5.2 and 5.3 constitute a canonical model of inversion based on a mirror-image reversal in the kinship dyad. They have very low token frequency, when compared to other forms of address (cf. Section 6). Even lower frequency of use characterizes affective address terms which are in a sense “deviant” modifications of symmetrical address inversion or involve some other “misplacement” of kinship terms. Such forms have been found in literary sources and produced by language consultants, too.

The most popular example among such “misplaced” kinship terms given by the Zanzibarian speakers was dada lit. ‘sister’ said by parents or grandparents to children or adults. One woman (28, Zanzibar) said that her mother preferred to call her dada rather than mama and commented that a person called like that by her mother felt kama amemtunza ‘as if she kept her/took care of her’. Another example often given by the speakers was mke wangu lit. ‘my wife’ said by a grandfather to a granddaughter and mume wangu lit. ‘my husband’ directed to a grandson by a grandmother. The speakers considered this term of address as both tender and joking. One woman (28, Zanzibar) added that such address was used in the case of a special bond, for example, when the child was the only grandchild or when a grandparent had given a name to the child. Instead of the kinship term, grandparents can use the name of their spouse. One man (25, Dar es Salaam) admitted that his grandmother would call him mume wangu lit. ‘my husband’, or Aaron which is the first name of her husband. Similar examples have been given by other consultants.

Departure from the canonical model was observed with respect to the gender of the kinship term used in inverted address. Three consultants from Zanzibar asserted that baba could be said by a man to his daughter and babu to a granddaughter, while also providing examples of the usual Swahili pattern with gender adjustment. Apparently, the principles governing address inversion in these consultants’ families represent a mixed variant in which Arabic and Swahili traditions occur in free variation.

Babu lit. ‘grandfather’ is the term which deserves special attention because of its multiple uses, extending beyond the genealogical and inverted kinship. As a term of tender address, it can be used among spouses, or among women. In the latter case, it became so popular that even the entry for babu in the Swahili-English Dictionary (2001) includes as one of its meanings ‘a call used among women friends’. Apparently, the word was reinterpreted as a general word of endearment first, but then, due to high frequency of use, it underwent semantic bleaching and lost its affectionate connotation. Examples of babu in varied contexts excerpted from the literary sources demonstrate a gradual change of this lexeme, evolving from a word of endearment to a discourse marker. ^[8] In (7a), said by a husband to his beloved wife, babu is a cordial form of address. In (7b), it represents an example of “girls’ talk” and serves to create an atmosphere of intimacy. The semantic bleaching of babu has proceeded even further in (7c), where it functions as a discourse marker or even a word of exclamation.

None of the language consultants have confirmed the use of the terms babu or bibi as endearment words directed to a spouse, while their use as a discourse marker has been fully accepted. One woman (38, Zanzibar) commented that in some contexts babu was not neno la kubembeleza ‘word of endearment’ but hata neno la kudharau ‘even an offensive word’.

6.1 Introduction

Kinship terminology certainly reflects the culture of language users. Heath (2006: 2015) gives the example of modern American society organized in nuclear families which correlates with the fact that relatively few differentiated terms denoting more distant relations are used by speakers of American English. Another of his examples is the practice of levirate ^[9] in some societies which apparently enhances the practice of using the same word for ‘father’ and ‘parental uncle’. The correlation between language and culture becomes even more striking in figurative uses of kinship terms. For example, Sharifian (2011: 57–59 and passim) demonstrates how the metaphor land is kin – present in many original Australian cultures – shapes Aboriginal English and extends the use of kinship terms far beyond the domain of people, when they refer to concepts of ‘land’, ‘country’ or specific landmarks.

While address inversion occurs in many languages, there are reasons to believe that it may comply with some cultures more than with others. A favorable socio-cultural setting enhances conventionalization of address inversion practices and makes it stable in passing from one generation to another. In this section, I will argue that indeed there are such favorable factors in the cultural values and social practices of the Swahili-speaking community.

From a more general perspective, if there are vivid cultural or social reasons triggering or supporting certain linguistic structures, we expect to find diverse linguistic data demonstrating that influence. Such an approach is in line with the cognitive linguistic assumption of shared cultural models existing in a speech community. This thought was articulated at the very beginning of the cognitive linguistics program (Lakoff and Johnson 1980; Lakoff 1987; Langacker 1987; Holland and Quinn 1987), and became a key factor in contrastive and typological research on emotions, embodiment and other issues involving the interaction of language with culture, philosophy or social practices (e.g., Kövecses 2005; Kraska-Szlenk 2014; Sharifian 2011; Yu 2009). Also today, many cognitive linguists take the stand that language structure and usage patterns are intrinsically tied up with the social and cultural needs of a speech community, and should be investigated taking the latter into account. ^[10] On the other hand, this line of research remains relatively unpopular, as recently observed by Dąbrowska:

while there is no doubt that language structure is strongly constrained by our perceptual and conceptual abilities, social and cultural factors play an equally important role; and although few cognitive linguists would dispute this claim, in practice, at least in the first thirty years or so of its development, cognitive linguists have paid little attention to the social factors that influence language structure. (2016: 485).

In a similar vein, Schmid gives good reason “why Cognitive Linguistics should embrace the social, pragmatic, and sociopragmatic dimensions of language more seriously” (2016: 545).

I will attempt to provide convincing evidence that the phenomenon of address inversion represents a fragment of a larger sociolinguistic picture. Taken piece by piece, the data might not provide a strong argument for the socio-cultural conditioning of specific linguistic forms, but taken as a whole, it demonstrates a coherent model of interaction between linguistic and social behaviors within the Swahili community. I will first focus on the high value of family ties in the Swahili society which represents a major cultural factor responsible for a range of linguistic forms and usage patterns, including the practice of inverted address. In the final part of this section, additional cultural aspects will be brought to attention, as well.

6.2 Family is good schema

There is probably no human society in which the concept of ‘family’, understood either in a very narrow or extended sense, would not have axiological positive marking. But it also appears a truism that societies significantly differ as to the importance of family ties, which affect cultural traditions and behaviors, including obligations with respect to people’s kin, and which also shape their attitudes and systems of values. The cultural Swahili community living alongside the East African coast was in the past, and still is, very much oriented towards maintaining close relations not only in the small circle of a nuclear family, but also with distant relations within extended families. Swahili people have numerous obligations towards their kin, respect family elders, follow genealogies through many generations and generally lead their lives with and for their families. Similar appreciation of one’s extended family also characterizes a larger population of Swahili-speaking people of different ethnic backgrounds in the interior of East Africa. Therefore, if a positively marked family-oriented cultural model was originally present in the Swahili language, it was naturally adopted and entrenched when the Swahili language expanded over larger communities of East Africa. In this section, I will draw on linguistic data to demonstrate how the high value of family ties is reflected in Swahili everyday language practice. In cognitive linguistics, these kinds of cases can be interpreted as instantiations of the propositional schema family is good, keeping up with a tradition of expressing interdependencies between language and social behaviors by means of schemas (e.g., up is good) or cultural metaphors (e.g., time is money), as proposed by Lakoff and Johnson (1980) and elaborated in subsequent works.

Swahili language users speak about their own and others’ family literally every day. One reason for that is that “family news” is included into conventionalized greetings which take the form of several questions and answers (e.g., Omar 1991; Yahya-Othman 1995). Habari za nyumbani? ‘What is the news from home?’, or Hawajambo nyumbani? ‘Are they well at home?’ – are typical examples. It is also polite to ask specifically about the wellbeing of particular persons mentioned either by name or his/her relation to the interlocutor, e.g., Wazazi hawajambo? ‘Are [your] parents well?’ Questions of this kind represent conventionalized greetings and are responded to in a conventionalized way which does not provide any meaningful content. However, the mere fact that such formulas are so common, indeed almost obligatory, in everyday socialization undoubtedly points out that one’s family is treated as an inseparable and important part of one’s identity in the Swahili-speaking community.

In everyday language usage consanguinal and affinal terms are used as cordial and mildly honorific address forms. The practice is very common in all Swahili-speaking area and can be considered a respectful, unmarked manner of addressing one’s kin (cf. Habwe 2010). In the Swahili-speaking community, addressing a family member with a term indicating real consanguinal or affinal ties is common and polite. Depending on the context and habits within one’s family, it may, but it does not have to entail particular closeness or affection.

Familial titles mama ‘mother’ and baba ‘father’ are encountered everyday by Swahili speakers due to their frequent use in the so-called teknonyms (from Old Greek teknon ‘child’ and onoma ‘name). This habit, recognized in many cultures of the world, communicates the high social value of parenting, but also joy and appreciation of one’s children. A widely known example is that of Arabic kunya which refers to a genitive construction of the form: abū X ‘father of X’ or umm X ‘mother of X’, in which X denotes the name of the eldest child.

In Swahili, teknonyms are likewise used with the name of the eldest child, hence, if X denotes the name of one’s eldest child, a woman is addressed mama X and a man baba X (also abbreviated as Ma X and Ba X). This practice occurs in all of Tanzania, but is more common in the interior than along the coast. Occasionally, teknonyms can be used with the name of any of one’s children who is meant to be honored in this way.

The above discussion has brought to attention three contexts in which Swahili language users encounter kinship terms on a daily basis: greetings, familial address and teknonyms. All of these examples have two important things in common. Firstly, in all these contexts, kinship terms have the positive connotation of being honorific and cordial forms at the same time. Secondly, token frequency of such uses is extremely high which means that these positive connotations get well entrenched into lexical representations of kinship terms, in accordance with findings of usage approaches to language learning, processing and retrieval (e.g., Langacker 1987, 2009). In other words, social practices such as those mentioned above add to positive axiological marking of kinship terms in their mental representations. This, in turn, makes kinship terms susceptible to developing other extended meanings with positive connotations. Inverted address constitutes one example. Another one comes from the practice of the so-called fictive kinship, briefly discussed below.

Fictive kinship refers to a situation when a non-related person is metaphorically framed as kin (a person is kin) and is addressed by a familial title. ^[11] Many societies in which the schema family is good is highly valued use the fictive kinship strategy for affective purposes. In industrialized societies, as for example, in modern Europe or the USA, we can observe that this practice, once also present, started to fade away as cultural values changed and individualism became more important than a social network. Although similar cultural changes affect the modern Swahili-speaking communities, the sense of family, in-group belonging and social interaction is still very strong, while the opposite behavior is negatively valued. As Yahya-Othman says, “‘[k]eeping oneself to oneself’ is […] something that one would try very hard to avoid” (1994: 145). In this cultural context, the habit of fictive kinship naming has stayed vivid and popular on the East African Coast, as well as in the continental parts of the Swahili-speaking area. In fact, it is widespread and frequent in language usage. Various familial titles used towards one’s family members can be directed to non-kin without restrictions. Their choice depends only on the relative age of the interlocutors, so that an established fictive relation reflects a possible family relation. It is generally agreed that these sorts of address strategies are polite and to some degree affectionate (cf. Habwe 2010; Kraska-Szlenk 2009; Podobińska 1997). Because of their positive connotation, they may also be employed for manipulative purposes, similarly to some uses of the inverted address.

Swahili literary sources provide many examples of kinship terms used as fictive address in various pragmatic contexts. For the purpose of this article, the most interesting cases are those in which the metaphor of fictive kinship combines with the practice of the address inversion. This is found in contexts in which strong positive emotions are expressed to a person who is unrelated to a speaker, but who is very close to him/her. The closeness may also have a temporary character as when a speaker wants to show compassion and to offer support in a difficult moment, as in the examples in (8), after Kraska-Szlenk (2009). In (8a), a young man is addressed baba by a man who loves him like his own son and expected him to marry his daughter, but the girl looked down on him, which became a source of great distress to both men. In (8b), tender words of support, including the address forms mwanangu and mama, are said by an older female servant to the daughter of her employer.

In some cases, the practice of fictive kinship became generalized and led to the bleaching and lexicalization of certain terms in particular contexts (cf. Section 7). These are address titles, such as dada ‘sister’ and kaka ‘brother’ spoken among youths who do not even have to know each other. In some families, the term dada is customarily used to call female domestic servants, which is considered more polite than using a name alone and, as one language consultant has observed, gives the girls some respect. The title shangazi ‘aunt’, associated with tenderness and concern, can be given to women who provide care for children, but also to “caring” hosts of television programs or authors of newspaper columns, as for example, Shangazi Sizarina ‘Aunt Sizarina’ in the Kenyan newspaper Taifa Leo. Baba ‘father’ can be an address term used for God or religious leaders, or even for the president of the country (cf. Habwe 2010).

To conclude, numerous examples brought to attention here demonstrate that address inversion as a strategy of expressing affection correlates with other uses of kinship terms in their positively charged senses, constructed in various contexts. The high value of kinship ties in the Swahili society is also frequently emphasized in texts of culture. Swahili proverbs and clichés point to the importance of family, obligations toward one’s kin and love of one’s kin, whether in a direct or a figurative manner. The same message can be found not only in conventionalized expressions, but also in novel, creative forms in literature, in poetry and in prose.

6.3 Other social and cultural factors

Swahili culture values indirect strategies in verbal communication in general and is very well known for its appreciation of metaphor and mafumbo ‘enigmas’, as well as sharp vijembe ‘innuendos’ (e.g., Yahya-Othman 1994; Vierke 2012). Indirect speech also takes the form of addressing a third party instead of communicating a message straightforwardly to those who are meant to receive it. Such measures belong to politeness strategies which help the speaker to maintain a positive ‘face’; “in using indirect address, the speaker conforms to what is considered appropriate in the avoidance of shame and embarrassment, and consequently enhances their own status within the community, which contributes to their positive face” (Yahya-Othman 1994: 159).

In general, the ‘face’ of an individual, but also of one’s family and the community, constitutes a highly ranked value among all Swahili-speaking people, although there might be regional and ethnic differences as to the particular rules of proper and improper behavior. The metaphorical ‘face’ requires maintaining one’s heshima ‘honor’ with simultaneous avoidance of aibu ‘shame’. These two concepts are crucial in the Swahili culture and social interaction, and it is remarkable that each of the two words has numerous near synonyms: twelve for heshima and thirteen for aibu in Mohamed and Mohamed’s (2004) dictionary.

With respect to affectionate language, Swahili speakers are reserved in naming positive emotions, although they can very well describe them in a less direct way, for example, in a poetic form. Being too outspoken about such delicate matters could be judged as aibu or losing some of heshima. Address inversion makes it possible to express emotions and at the same time to avoid being direct and overtly affectionate, which ultimately enables a person keeping his/her ‘face’. Culturally recommended indirectness in speech and reserved manner in expressing emotions ideally comply with the practice of address inversion in the Swahili community, whatever the original source of the phenomenon could be.

The last part of this section will be devoted to these socio-cultural factors which the Swahili consultants themselves have pointed out as causes of address inversion.

The consultants interviewed on the island of Zanzibar often made a point about the same name reoccurring within the same family across different generations. The most commonly given example was that of a grandchild and a grandparent, since children are often named after their grandparents. Therefore, some consultants claimed that the reason for calling a child baba or mama was the fact that a parent’s father or mother bore an identical name. The same explanation was given with respect to other kinship terms used in the inverted address, such as dada, babu or bibi. Even when confronted with the argument that not all children called by the inverted address bear the same name as a senior bearing an appropriate title, some were quick to find an explanation in the polysemy of words like baba meaning ‘father’, but also ‘paternal uncle’, etc. Consequently, they would claim that Swahili families are so large that one always finds some uncle or grandfather of the same name. The name-sharing practice indeed seems to constitute an additional cultural factor which preserves address inversion in the community.

In mainland Tanzania, children naming practices are often based on other traditions, hence the language consultants coming from these parts of the country have not come up with the name-sharing idea as a possible explanation of address inversion. On the contrary, they seemed rather puzzled after having realized the occurrence of address inversion practices in their families. One young man mentioned that he once asked his father why he was never called baba by him, while his younger brother was. He received a response that there was resemblance between the man’s younger son and his own father and that was the reason for his using the term baba with respect to that particular son. More data is needed to verify whether such explanations are shared by other language speakers. It has been observed during the present research that the use of address inversion in the interior is not as common as in the coastal areas. This could mean that address inversion was adopted by other communities in East Africa as part of Swahili language practices. But it is also possible that attitudes toward address inversion in mainland Tanzania are influenced by local ethnic cultures.