Abstract
Introduction:
This paper investigates the structure and function of translinguistic apposition in Kinyarwanda news articles. In the structure, a Kinyarwanda noun phrase and a French / English noun phrase are juxtaposed as if the second clarified the first. Two major issues are identified, namely the possibility of translinguisticc apposition as practical action and its implicativeness relative to preceding and subsequent choices. These issues are identified against the background of two important discourse organisational principles, namely the principle of progressivity and that of the preference for same language discourse.
Methods:
Data were collected in two media blogs in Rwanda, namely Igihe.com and Umuseke.com using the method of theoretical sampling. They were analysed, in the tradition of the Ethnomethodology of Written Discourse, using a conversation-analysis-inspired qualitative analysis.
Results:
Translinguistic apposition is found to be possible as practical action because, organised as a parenthetical, it is shielded off from the reach of the preference for same medium discourse. It is also found to serve as a discourse-based language management strategy. Its implicativeness derives from this function.
Discussion:
Implications for the study of bilingual language use and for the practice of lexical development at the macro level are discussed.
Keywords
Introduction
Using the perspective of the ‘Ethnomethodology of Written Discourse’ (Padua, 2024), a perspective based on the view that written texts are ‘distant interactions’ (McHoul, 1978, p. 115), this paper investigates a discourse practice I have observed in Kinyarwanda news articles and termed translinguistic apposition (Gafaranga, 2015). For a gist of this practice, consider the following:
Mu igenzura ryakozwe na police, yasanze iyo modoka yarakoresheje (. . .) SP Ndushabandi yongeye kwibutsa abashoferi ko nubwo baba bafite ibyangombwa by’ (In their investigation the police found that the vehicle had passed the motor test and that it had a certificate valid until the end of August 2016. (. . .) SP Ndushabandi reminded drivers to check their breaks before undertaking a trip even if they have valid test certificates) Gusa, bivugwa ko yaje kugira ikibazo cyo kuva cyane nyuma y’uko kubyara abazwe, bituma abaganga bakuramo (. . .) akaba yarahise anamburwa amahirwe yo kuzigera abona undi mwana mu buzima bwe, kubera ko yakuwemo (it is reported that she developed a problem after the c-section which resulted in removal of her iterus (. . .) She has lost the chance to have children as her uterus has been removed)
As the highlighted elements in the extracts show, translinguistic apposition is a structure whereby two noun phrases (NPs) from two different languages (the Kinyarwanda ‘igenzura ry’ibinyabiziga’ vs its French equivalent ‘contrôle technique’ in Extract 1 and the Kinyarwanda ‘nyababyeyi’ and the English ‘uterus’ in Extract 2) are juxtaposed as if the second explained or clarified the first (Quirk et al., 1985).
In Kinyarwanda news articles, two different practices can involve the juxtaposition of two NPs from two different languages, and it is important to keep them apart. These are translanguistic apposition, as illustrated above, and the equivalent of what, in spoken discourse, is known as medium repair (Gafaranga, 2000, 2012), as in the examples below.
Imyambaro y’ingabo z’abanyamerika ubundi ikoranye ikoranabuhanga ryo ku rwego rwo hejuru ku buryo zikenera uburyo bwo kongera amashanyarazi muri za mudasobwa zikoresha ibyo byuma byose harimo na (The American army gear is very sophisticated such that they need ways to recharge the computers they use for different kinds of instruments including binoculars.) Ntitwabashije kuvugana na nyirimodoka kuko twasanze adahari, gusa ababibonye biba bavuga ko ashobora kuba yasize atayifunze na (We haven’t been able to talk to the car own, but witnesses said that they (the car owner) probably did not stabilise it using the hand break and that this is probably the reason it rolled)
Translinguistic apposition and medium repair are similar in that, by repeating the same content, they interrupt the discourse ‘progressivity’ (Schegloff [1979], Lerner [1996], Stivers and Robinson [2006], Stivers and Hayashi [2010], etc.). This is the idea that, in interaction, ‘each component progresses to the next relevant component immediately after (. . .) the prior component’ (Kitzinger, 2014, p. 239). Also, both translinguistic apposition and medium repair flout ‘the preference for same language’ discourse (Auer, 1984). This is the fact that, in monolingual discourse, it is preferred that adjacent units adopt the same language. However, they are also different. At the level of language choice, in medium repair, departure from the medium takes place before progressivity is interrupted while, in translinguistic apposition, departure from the medium coincides with the interruption of progressivity. Thus, in the case of medium repair, the question to ask is why the medium is interrupted while, in the case of translinguistic apposition, the question to ask is why progressivity, the primary organisational principle, is interrupted. And, based on what happens after the interruption, in the first case, the answer is that interruption was an ‘error’, hence repair, while, in the second, the answer is that interruption of progressivity is functional (e.g., clarification). Medium repair will not be discussed in this paper.
Gafaranga (2015) has examined translinguistic apposition in news texts in Rwanda, focusing on its possibility given the nature of the texts (‘most highly regulated’ – Sebba, 2013) in which it is observed and given the diversity of the Rwandan readership’s linguistic competence (see below). He accounted for the possibility of translinguistic apposition, a micro-level structure, by referring to the macro language policy. He commented that, in the particular case, ‘macro language policies (shape) micro language choice practices and, in turn, through micro language choice practices, macro language policies are renewed, that is, talked/written into being’ (Gafaranga, 2015, p. 108). The current paper supplements that earlier contribution. Two main issues are pursued, namely the possibility of translinguistic apposition as practical action and its sequential implicativeness. The first issue arises because of the preference for same language discourse (see above). It is that of how the contact between Kinyarwanda and French/English at the transition to and from transilinguistic apposition is managed. At each of these points, the preference for same language discourse is flouted. In the paper, it will be argued that translinguistic apposition is designed as a ‘parenthetical’ (Mazeland, 2007) and, therefore, that it is shielded off from the reach of the preference for the same language discourse.
As for the second issue, it is a strictly empirical one. As Extract 1 and Extract 2 above show, if the ideational content of the juxtaposed items is expressed in subsequent text, the Kinyarwanda formulation is adopted. However, there is also evidence that the French/English formulation can be adopted. Extract 5 is a good example of this second possibility.
Kimwe mu bintu bikunda kugora abashaka gutangira kwikorera kandi bafite imishinga myiza ni uguhitamo izina bazaha (. . .) Iyo ugiye guhitamo izina akenshi uba wifuza guhitamo izina rizaramba kandi rikubiyemo amahame ya (One of the challenges that those who want to be self-employed face even when they have good projects is the choice of the name for the business (. . .) You often want to choose a name that will last and, at the same time, captures the main aims and the originality of your business.)
Also, Extract 1, Extract 2, and Extract 5 suggest that translinguistic apposition is an initiating action, has or does not have consequences for choices to follow. However, in the following example, translinguistic apposition occurs after the Kinyarwanda item has already been used.
Mutimura avuga ko uku kumanuka kw’igiciro cy’imigabane ya Bralirwa n’ iya Crystal Telecom birimo guteza (Mutimura said that the drop in value of the Bralirwa and and Crystal Telecom shares is causing realized and unrealized losses to shareholders.(. . .) If that person does not sell and holds on to them (shares) they will be experiencing unrealized loss for they still have their shares but these are generating any benefits. (. . .) On the other hand, if they decide to sell, they will experience realised loss because . . .).
Schegloff (1993) has argued that ‘informal’ quantitative evidence can point to an underlying order. Informal distributional evidence can indicate ‘the likely presence of an element of order in the larger universe investigated’ (p. 119). In the case of translinguistic apposition, distributional evidence suggests that reproducing the Kinyarwanda item if the content of the items in the appositive structure is expressed again in subsequent text is normative, while using French/English is deviant. Likewise, distributional evidence suggests that translinguistic apposition is an initiating action and, therefore that deploying it when the Kinyarwanda item has already been used is deviant. In this paper, a functional account for this orderliness is proposed, namely that translinguistic apposition is a ‘discourse-based language management strategy’ (Nekvapil, 2016).
Context, data, and methods
The macro ‘declared language policy’ (Shohamy, 2006) of Rwanda, as stated in the country’s constitution, is:
The national language is Ikinyarwanda.
The official languages are Ikinyarwanda, English, and French.
An organic law may add or remove an official language. 2
Official documents may be either in one, two, or all official languages.
(Constitution of the Republic of Rwanda, 2003 as revised in 2015, art. 8).
As it can be seen, relative to the other official languages, Kinyarwanda occupies a privileged position because it has been recognised the additional status of the national language. In turn, this is not surprising as the overwhelming majority of Rwandans are either bilingual in Kinyarwanda and one other language or simply monolingual in Kinyarwanda (Rosendal [2009] and Samuelson and Freedman [2010]). On the other hand, the privileged position of Kinyarwanda, as found in the declared policy, is reflected in the ‘perceived policy’ (Shohamy, 2006). In Rwanda, Kinyarwanda is associated with ‘Rwandan-ness’, with being Rwandan (Gafaranga et al., 2013). It is in this context that the preservation of the integrity of Kinyarwanda is very high on the political agenda. For example, the Rwandan Academy of Language and Culture (RALC) has been created with the explicit agenda of preserving and promoting Kinyarwanda, notably by protecting it against the so-called threat from the other official languages (www.ralc.gov.rw/index.php?id=140, accessed 29 March, 2016).
However, even though Kinyarwanda has been declared an official language, like many other languages in a similar situation in post-colonial contexts, it faces, among other challenges, that of the limited nature of its lexicon. Commenting on post-colonial language policies in Africa, Simpson (2008, p. 5) says that concerns about ‘the state of development of indigenous languages and their potential use in the same range of domains as European languages’ are one of the reasons why, post-independence, many African countries have opted to keep former colonial language policies intact. In this respect, opening the consultative meeting on Kinyarwanda lexical development, the chairperson of the Rwandan Academy of Language and Culture said:
‘Muri iki gihe birazwi ko abenshi mu bakoresha Ikinyarwanda bahura n’ikibazo cyo kuvuga inshoza z’ubumenyi, ikoranabuhanga, imyuga n’ubugeni mu rurimi kavukire. Babura amagambo yabugenewe yo gukoresha mu nzego zinyuranye, n’ariho ntibayamenye’. (http://ralc.gov.rw/fileadmin/user_upload/Documents/Raporo_y_Inama_nyunguranabitekerezo_ku_muga_y_Ikinyarwanda_13.06.2014_RALC_08.pdf, accessed 12-06-2017) (Today it is well-known that many of the users of Kinyarwanda struggle to express scientific, technological, professional (. . .) concepts in the native language. Not only appropriate terminologies are not available, but also those which are available are often not familiar to them (users).)
In other words, Kinyarwanda, although it is fortunate to have been recognised as a de jure official language, is not sufficiently equipped at the lexical level to function efficiently as an official language.
Aware of this problem, Rwandan authorities have undertaken lexical development activities, although on an intermittent basis. Notable at this level is the work accomplished by Urutonde. This project, in action in the 80’s and early 90’s, had been launched so as to address the needs of the then language-in-education policy which required the use of Kinyarwanda as the medium of instruction in all Rwandan schools. In the same time, some other public and private institutions produced their own lists of new words in Kinyarwanda, such as the one produced by the Department of Justice. Notable work was also accomplished at the then National University of Rwanda in the form of students’ theses. However, all this language development activity came to a halt following the political and security situation in the first half of the 90s (economic downturn, civil war and Genocide Against Tutsis). As soon as security and stability were restored, the Rwandan Academy of Language and Culture was instituted and its first ‘experts’ were sworn in on 25 August 2012 (http://ralc.gov.rw/index.php?id=103, accessed 14 June 2017) with, as we have seen, the explicit agenda of protecting and promoting Kinyarwanda. As soon as it was instituted, the Academy put itself to the task of lexical development. So far it has developed a Kinyarwanda ‘dictionary’ for the sciences (anatomy and biology) and work in other areas goes on. It has also reviewed and updated already existing lists. Crucially, the Academy has recognised its limitations and called on various professions and language users in general to contribute to the effort of developing Kinyarwanda. In this context, the role of the media has been highlighted. On one hand, the media have been blamed for contributing significantly to the demise of Kinyarwanda through inappropriate use of the language (https://www.nonaha.com/intiti-ziracyabona-abanyamakuru-ku-isonga-mu-bahohotera-ikinyarwanda, accessed 11 December 2017) and, on the other hand, they have been recognised as a potentially powerful channel for disseminating linguistic innovations, their own as well as those from other agents, to the wider public (www.bwiza.com/musanze-abanyamakuru-barasabwa-kuba-umuyoboro-mwiza-wo-kwigisha-ikinyarwanda/, accessed 11 December 2017).
Concerning the news media themselves, two points have to be mentioned in this background to the practice of translinguistic apposition. First of all, Rwandan news media have a relatively short history. The first publication of a newspaper in Rwanda dates back to the 30s (Kinyamateka was first published in 1933). Another milestone was the publication of Imvaho in 1959. Both newspapers published, and still do, in Kinyarwanda only. While Kinyamateka is owned by the Catholic Church, Imvaho is a Government-owned news outlet. Over time, the number of news outlets has increased such that, today, there are between 30 and 50 print media outlets (www.mhc.gov.rw/index.php?id=62, accessed 20 April 2016). With the arrival of the internet, many of these outlets started publishing online versions, and even internet-only news outlets have since emerged. The two blogs I have monitored for this study reflect this reality as Umuseke has a paper version while Igihe is published online only. Second, there are huge training needs for Rwandan journalists. Commenting on these needs, Thompson (2007, p. 25) writes that ‘newsrooms in Rwanda are populated for the most part by young journalists with little or no professional training’. In fact, this is not surprising as, up until 1996, there was not a single journalism training institution in Rwanda. Today, the situation can be assumed to have improved as there are four journalism training centres, including the School of Journalism and Communication of the University of Rwanda. It is hoped that empirical investigations of media practices such as the one presented in the following pages will contribute to filling those skills gaps.
The data for this paper come from news articles collected from two Rwandan multilingual news media blogs, namely Igihe and Umuseke. 3 While Igihe publishes in all three official languages of Rwanda, Umuseke publishes in Kinyarwanda and English only. In so doing, the two blogs follow the pattern described by Gafaranga (2015, p. 95) as combining features of both ‘complementary’ and ‘parallel texts’ (Sebba, 2013). For the purposes of this article, only the Kinyarwanda versions were investigated. Most data were collected in July 2016, but, as the news outlets continue to be accessible and that their archives can be searched there is no limit to what counts as data for this paper. Rather, the theoretical sampling method was adopted (ten Have, 1999). The blogs were (and continue to be) monitored for instances of translinguistic apposition and its consequences for previous and subsequent choices, and hypotheses as to its orderliness were developed and tested on each new instance until no new observation could be made (saturation). Analysis of these data, as indicated in the Introduction, has adopted the perspective of the Ethnomethodology of Written Discourse. This perspective, just like its more popular sister Ethnomethodological Conversation Analysis (EMCA), retains the ethnomethodological ontology and epistemology (Padua, 2024, pp. 3–4) and, therefore, can be referred to as Ethnomethodological Text Analysis or EMTA for short. More specifically, in this perspective, written discourse is seen as ‘distant interaction’ (McHoul, 1978, p. 115). Therefore, this paper asks the kind of questions which are typically asked in the analysis of face-to-face interactions and adapts and adopts some of the key concepts developed in Conversation Analysis. The findings were triangulated with media practitioners’ own interpretations obtained during a knowledge exchange event which was organised in Kigali (Rwanda) in September 2019.
Theoretical frameworks
Four key concepts form the theoretical background of this paper. The first concept is that of translinguistic apposition. The notion of apposition, well-known in stylistics, refers to ‘a syntactic structure whereby two noun phrases with a similar meaning are placed next to each other, the second functioning as an explanation, clarification, rename and so forth of the first’ (Gafaranga, 2015, p. 88 [referencing Quirk et al., 1985]). Two types of appositive structures exist, namely restrictive and non-restrictive appositives. As the name indicates, in a restrictive appositive structure, the second NP narrows the scope of the first while it does not in a non-restrictive appositive structure. Instead, in a non-restrictive appositive, the second component is seen as a restatement, reformulation, and so on of the first. As a result, a non-restrictive appositive structure is conventionally signalled by the use of brackets, quotation marks, dashes, and so on, while a restrictive appositive is not. Of interest for this paper is the non-restrictive type of apposition. The data we have looked at so far show these properties of non-restrictive appositives and, in addition, the structure makes use of two languages.
The second key concept is that of sequential implicativeness. The notion of sequentiality (and hence of sequential analysis), foundational in Conversation Analysis, refers to the property talk actions have of making relevant, of implicating, other actions. Prototypical examples of this implicativeness can be found in the conversational structure known as adjacency pair (Schegloff & Sacks, 1973). For example, a request makes relevant its granting/denial, an invitation makes relevant an acceptance/rejection, an assertion makes relevant an agreement/disagreement, and so on. So, given a request, we can inspect subsequent talk for its response (granting/denial) and given what appears to be a response we can inspect previous talk to see what it is a response to. This is known as sequential analysis. These ideas of sequential implicativeness and of sequential analysis have been extended to the study of language choice in bilingual conversation under the label of ‘conversation analytic model of code-switching’ (Auer, 1988; Gafaranga, 2007, 2019). The idea behind sequential analyses of language choice is that, in bilingual conversation, language choice at one point is implicative of, has consequences for, choices to follow. This paper addresses the implicativeness of translinguistic apposition for preceding and subsequent choices.
The third key concept I will draw on is that of ‘discourse-based language management’ (Nekvapil, 2016), a concept we owe to Language Management Theory (LMT). Since its inception, language planning, including LMT, has been ‘seen as a way of solving “language” or “communication problems”’ (Daoust, 1998, p. 440). In turn, according to LMT, two types of problems or ‘inadequacies’ can be found, namely ‘discourse inadequacy’ and ‘system inadequacy’ (Neustupný, 1978, p. 250). The difference between the two is whether the inadequacy in question is identified and corrected at the level of or beyond ‘a particular discourse (act)’. Correspondingly, two types of language management are identified, namely ‘discourse-based language management’ and ‘organised language management’. However, as processes, discourse-based language management and organised management are not ‘intrinsically different’ (Kimura, 2005, p. 7). They both consist of four phases which Kimura (2014, p. 267) has captured in the following cycle (Figure 1):
Language management cycle (Kimura, 2014, p. 267).
Language management, whether discourse-based or organised, starts by noticing an inadequacy, usually in the form of a ‘deviation from a norm or expectations’. The deviation is then evaluated either negatively or positively. Positive evaluation leads to the interpretation that deviation was functional while negative evaluation calls for an adjustment. The adjustment is then implemented and the result is evaluated again (‘post-implementation evaluation’ [Kimura, 2014]), this new round of evaluation potentially leading to a new cycle (Jernudd & Neustupný, 1987, pp. 75–76, Nekvapil & Sherman, 2015, p. 7; Svelsch, 2015, p. 145, etc.). It is because of these basic similarities that, according to Neustupný (2005), ‘(. . .) organised management is (simply) a complicated version of the basic simple management process’ (cited in Kimura, 2005, p. 7).
Discourse-based language management has been strongly likened to conversational repair (see Nekvapil, 2016, p. 17; Nekvapil & Sherman, 2009, p. 185). Repair, one of the most pervasive conversational actions, is when an interactant ‘interrupts the ongoing course of action to attend to possible trouble in speaking, hearing or understanding the talk’ (Kitzinger, 2014, p. 229). Studies of bilingual conversation have amply demonstrated that troubles in the production and reception of discourse are often related to language choice. Through conversational repair, medium repair in particular, interactants do discourse-based language management, that is, address interactional inadequacies.
However, discourse-based language management, as Kaplan and Baldauf (2005) show, can also be found in written discourse if the writing process itself is observed. This paper argues that, in Kinyarwanda news articles, writers use translinguistic apposition as a discourse-based language management strategy in order to negotiate potential communication inadequacies which may result from lexical gaps in Kinyarwanda.
The final key concept in this paper is that of ‘lexical gap’. 4 The notion of lexical gap is commonly understood as the absence of a word/fixed expression in a language conventionally associated with a specific concept or meaning. Such gaps can be found in every language. Note that the absence of a conventional expression for a specific concept in a language does not mean that the concept cannot be expressed. According to Rissman et al. (2023), the expressive power of language relies on two processes, namely ‘conventionality’ and ‘compositionality’. The latter refers to the ability to express non-lexicalised meanings drawing on available language resources (p. 412), for example, borrowing (nonce words), compounding, derivation, and so on. Given this, lexical development, such as the one undertaken by the Rwandan Academy of Language and Culture, can be seen as a lexicalisation effort, an effort to conventionalise what speakers normally do compositionally. In turn, this effort may be seen as undertaken on the grounds that ‘a conventional label may offer a uniquely effective way of communicating’ and that ‘the ability to rely on a conventional term is surprisingly important despite the flexibility languages offer to communicate about nonlexicalized categories’ (p. 412). In the case of multilingual communities such as that of Rwanda, lexical gaps may be found across languages. That is to say, concepts may be lexicalised in one language and not in the other. Thus, when faced with a lexical gap while using one of the languages, speakers apply the resources of the language and create ‘compositional meanings’. In addition, as the lexicalised item is available to them in the other language, they may juxtapose it for increased expressivity and, thus, produce a translinguistic apposition.
A structural and functional account of translinguistic apposition
As I have mentioned it in the Introduction, the first issue translinguistic apposition raises is that of its possibility as practical action. I have also said that the issue arises because of the preference for same language discourse. I have framed the issue as that of how the transition to and from translinguistic apposition is managed, given the fact that, at each transition point, this discourse organisational principle is flouted. Mazeland (2007) has investigated a conversational structure he refers to as ‘parenthetical sequences’ and it seems to be an appropriate description of the work writers do in transitioning to and from translinguistic apposition. In parenthesizing, ‘a speaker (. . .) halts an ongoing turn constructional unit (TCU) before its completion, inserts a short parenthetical remark into it, and then returns to the halted TCU’ (p. 1816). More relevantly, a speaker ‘anticipates a possible recipient problem with something he has just said (. . .) in the on-going turn’ (p. 1863) and ‘informs the recipient metacommunicatively and in real time how to listen to the turn in progress’ (p. 1816). Mazeland groups parentheticals into four main categories, one of which is ‘specifying / explicating appositive nominals’ (pp. 1827–1829). Translanguistic apposition appears to be a version of this type of parenthetical in ‘distant interactions’ (McHoul, 1978, p. 115). Mazeland lists three problems that an account of parentheticals must address, namely initiation problems, maintenance problems and return problems. The issue of transition in and from translinguistic apposition corresponds to the first and last problems. Mazeland provides a general answer to both problems saying that ‘the speaker does special work to demarcate the appositive unit from its surrounding in the clause. He (sic) displays the appositive construction as a unit that may be separated out from its environment’ (p. 1829). In face-to-face interaction, this is achieved, among other devices, through prosodic shifts. In the case of translinguistic apposition, as the data we have looked at shows, devices such parentheses are systematically used. Focusing on the transition back to the host discourse after a parenthetical, Mazeland identifies three main strategies, namely continuation, resumption and redesign, noting that continuation is the most common. In this strategy, ‘the speaker continues with a non-initial fragment that seamlessly fits into the syntactic frame begun in the pre-insert part of the host’. (p. 1852). Translinguistic apposition adopts this strategy. Consider Extract 7.
Amategeko n’amabwiriza agenga imitunganyirize y’umugi agena ko mbere yo gutangira gukora bagomba guhabwa (. . ..) Biteganyijwe ko ubwo ibisubizo bizatangwa n’ubugenzuzi bw’iki kigo kimwe n’ abagenzuzi bihariye b’Umujyi wa Kigali, iyi hoteli ishobora guhabwa (Planning regulations require that an occupation permit, issued after the building has been inspected, must be obtained before use. (. . .) Plans are that, after inspection by this institution and by the City of Kigali own inspectors, a permit will be issued to the hotel if it meets the requirements . . .)
In the extract, the post-apposition element (‘gi-tangwa’) agrees in class with the pre-apposition item (‘icyangombwa’). The point about this returning strategy is that, as Mazeland (p. 1863) says, it
‘
Thus, it is because translinguistic apposition is demarcated from the host discourse and that, through the continuation strategy, the French/English item is ‘deleted from the local structural memory’ (p. 1855) that transition into and from it is possible, that is, does not bump against the preference for same language discourse.
The second issue that translinguistic apposition raises is that of its sequential implicativeness. Distributional evidence suggests that it is normative for the Kinyarwanda item in an appositive structure to be reproduced if the same content is expressed in subsequent text. Informal quantitative evidence also suggests that it is normative for translinguistic apposition to be deployed as an initiating action. Both aspects are easily accounted for if translinguistic apposition is seen as a discourse-based language management strategy through which writers address the issue of lexical gaps in Kinyarwanda. To be sure, two strategies for addressing such gaps can be thought of. First, as in the case of spoken discourse, lexical gaps can be addressed through code-switching as in Extract 8.
Kompanyi y’imodoka ibarizwa mu ruganda rwa General Motors, Cadillac, yashyize hanze ubwoko bushya bw’iyitwa Escalade ishobora kwitwara mu mihanda migari bidasabye ko umushoferi agenzura akoresheje (Cadillac, a car manufacturer within the General Motors confederate, has unveiled a model called Escalade which can auto-drive without the need for the driver to direct the car using the steering wheel or to reduce or increases the speed using the relevant pedals)
However, this solution is not common, presumably due to the fact that news articles are ‘most highly regulated text’ (Sebba, 2013). In addition, directly inserting French/English items in a Kinyarwanda medium runs against the preference for same medium discourse. To get around this problem, strategies, for example, the use of inverted commas, for shielding the French/English item off from the reach of the preference for same language discourse are used (see above).
Translanguistic apposition appears to be an alternative and more common strategy for dealing with lexical gaps in Kinyarwanda. As indicated in the Context section, a group of media editors representing 24 news outlets in Rwanda gathered for a knowledge exchange event in Kigali (Rwanda) in September 2019. In the event, they concurred that they use translinguistic apposition in order to negotiate lexical gaps in Kinyarwanda. Indeed, in many cases, translinguistic apposition involves NPs which belong to a technical/scientific register (e.g. ‘uterus’ in Extract 2, ‘realised / unrealised loss’ in extract 6, ‘occupation permit’ in Extract 7, etc.). That is, translinguistic apposition is drawn upon in order to express meanings which are not traditionally lexicalised in Kinyarwanda. For practitioners themselves, translinguistic apposition is a discourse-based language management strategy. Consider Extract 9 below.
Guvernoma y’uRwanda ibinyujije muri Banki Nkuru y’Igihugu (BNR)
5
yashyize ku isoko (. . ..) Bavuga ko ari ubwa mbere hatanzwe igihe cyo kwishyura kingana n’ imyaka 15 kuva aho gahunda yo kugurisha (The Government of Rwanda, via the National Bank of Rwanda (BNR) has put on sale treasury bonds worth . . . (. . .) It is reported that this is first time that a repayment period of 15 years has been offered since the programme of selling treasury bonds started in 2008.
To put it in so many words, the writer notices an inadequacy in the form of a lexical gap. The concept of treasury bond is (assumed) not (to be) lexicalised in Kinyarwanda. This inadequacy is evaluated negatively, given the preference for progressivity. Following this negative evaluation, an adjustment (‘impapuro mpeshwamwenda
Also, because translinguistic apposition is a discourse-based language management strategy, deploying it after the Kinyarwanda item has already been used is, to say the least, an unnecessary disruption of progressivity. Since the Kinyarwanda item has successfully been used in previous text, there is no inadequacy to negotiate. Other forms of deviance too can be accounted for. Consider the following:
. . . ndavuga nti dukeneye uburyo mu duce tumwe na tumwe turimo ubucucike, by’umwihariko nka Nyabugogo hashyirwaho (. . .) Nkurikije ibiganiro nagiranye na bamwe mu banyamaguru bakoresha imihanda ya Nyabugogo mu buryo buhoraho, nasanze (I told myself that in overpopulated areas, especially in areas like Nyabugogo, we could have pedestrian bridges people can use to cross without having to use usual ways of crossing such as zebra-crossings. (. . .) Based on the conversations I had had with pedestrians who use Nyabugogo Roads on a permanent basis, I found out that pedestrian crossings are a solution they had always wanted)
Here, after the first instance of translinguistic apposition, either the inadequacy has been removed or it has not. If it has been removed, the second instance is unnecessary. On the other hand, if translinguistic apposition has not removed the inadequacy, here again, it is unnecessary as it cannot be expected to succeed this second time around.
Discussion and conclusion
To summarise, translinguistic apposition in Kinyarwanda news articles, a practice whereby an appositive structure draws on two different languages, initially appeared to be disorderly at many levels. This article has focused on two of these, namely the possibility of translinguistic apposition as practical action and its implicativeness for preceding and subsequent choices. The first issue was framed as that of the management of the transition to and from translinguistic apposition, given the preference for same language discourse. Observation of the data revealed that writers deploy various strategies, for example the use of brackets, in order to demarcate translinguistic apposition from the host discourse, to shield it from the reach of the preference for same language discourse. As for the second issue, I solved it referring to the function of the practice. Media practitioners themselves say that they use translinguistic apposition in order to overcome the problem of lexical gaps in Kinyarwanda. That is, they report using translinguistic apposition as a discourse-based language management strategy. Analysis of the data confirmed this understanding. In the paper, this function of translinguistic apposition was demonstrated to be consequential to the shape of the practice. Because of this function, only translinguistic apposition as an initiating action is normative and the use of Kinyarwanda in subsequent text after translinguistic apposition is preferred.
This investigation of the structure and function of translinguistc apposition has many theoretical and applied implications, but, for the lack of space, only two (one theoretical and one applied) can be mentioned here. First of all, studies of bilingualism, and of language alternation in particular, have been based on a number idealisations (idealisation of the community, idealisation of speakers’ competence, etc.). For our concerns, we can note the idealisation of language as a perfect tool of communication. A common view in Linguistics is that any language is capable of expressing all of its users’ needs. While this is generally true, it is also true that not all languages can express all of their users’ needs equally efficiently. To express all of its users’ needs, each language, as we have seen, draws both on conventionality and on compositionality. It follows that, in bilingual contexts, conventionality and compositionality do not necessarily map onto each other across languages. Concepts may be lexicalised in one language and not in the other. We have also seen that, according to the literature, communication is more efficient when concepts are lexicalised than when recourse must be made to compositionality. Studies of bilingual language use seem to have overlooked this aspect. They seem to have been based on the assumption that both bilinguals’ languages can express all of their needs equally efficiently, i.e., they have assumed that conventionality and compositionality match onto each across the bilingual’s languages. For example, Myers-Scotton’s (1993) maintains, and rightly so, that, in bilingual communities, language choice is socially motivated. But it is also true that, not all language choices can be explained in terms of ‘social motivations’. Some choices, for example, translinguistic apposition in Kinyarwanda news articles, may be motivated by the need to communicate most efficiently, that is, to compensate for lexical gaps in one of the languages in contact. Likewise, Auer (1984) identifies two categories of language alternation, one which tells us something about the organisation of discourse (discourse-related code alternation) and one which tells us something about participants (participant-related code alternation). The possibility for language alternation to tell us something about the code itself is not envisaged. Briefly, this paper points to the need for studies, such as the on-going, which focus on what, echoing Auer (1984), one might call code-related language alternation.
Second, there is an applied implication. As we have seen, the Rwandan Government, through the Rwandan Academy of Language and Culture, has been involved in organised language management and they have recognised the crucial role the media can play in the process. On the other hand, scholars tell us that there is a ‘dialectical or reflexive’ relationship between discourse-based language management (micro) and organised language management (macro; Nekvapil, 2016, p. 16). They tell us that organised language management needs discourse-based language management. According to Nekvapil, ‘ ideally, organized management is based on the occurrence of simple management, that is, it resonates with speakers’ (own) (. . .) effort to use adequate means to remove (language and communication) problems’ (p. 16). That media practitioners themselves engage in discourse-based language management should be a source of comfort for macro-level language planners. Not only it gives the hope that management efforts at the macro level will be positively received at the micro level, but it also points to the existence of native strategies for language management that macro level agents can draw on for sustainable lexical development. This article has described one such strategy, and further investigation may reveal many more.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
