Background

History & Development of Kaaps

Kaaps is a heavily contact-influenced matrilectal variety of Afrikaans which is spoken on the Cape Peninsula. It has also been referred to in the literature as 'Cape Vernacular Afrikaans' and 'Cape Afrikaans' (Hendricks 1978:13-26; Le Cordeur 2011:763-766), and more recently by the endonym 'Afrikaaps' (Valley 2010; Williams 2016).

The precise origins and development of Kaaps are difficult to trace (Kotzé 2016). Nevertheless, it is not a “new” variety, and though present-day Kaaps differs substantially from its predecessor(s), it emerged from one of the oldest varieties of Afrikaans, under socioculturally and sociolinguistically “superdiverse” conditions (Hendricks 2016:4). Ad hoc contact between the Khoekhoe and Dutch sailors who occasionally docked at Table Bay since around 1595 (Van Rensburg 2016:51) saw the development of a Dutch “foreigner talk” (Den Besten 1989) or a local koiné (Mufwene 2001; Chaudenson 2001). Fifty years later, with the presence of Dutch settlers, contact stabilized. Between 1665-1795, a yearly average of 2300 Malaysian-Dutch speaking members of the Dutch East India Company passed through the Cape. Slaves brought from Malaysia, Java, Indonesia, India, Madagascar, Angola, and Mozambique, spoke contact varieties from their own regions, e.g. Pasar-Malaysian and Creole Portuguese, and their respective home languages (Hendricks 2016: 4). In early 1800 Bo-Kaap & District Six (BK&DS) were designated slaves’ quarters (Le Cordeur 2016: 84). In the 1950s the Group Areas Act caused the forced removal of inhabitants from BK&DS to Mitchells Plain and other places on the Cape Flats. “That different groupings of the Afrikaans speaking community had lived apart… for so long, led to… divergences in linguistic practices” (Le Cordeur 2016:85; my translation).

Kaaps seems to have developed from what Den Besten (1989) calls Proto Afrikaans I, a “Creole Dutch” spoken by slaves and freemen, with its cradle in the superdiverse BK&DS. The variety of Afrikaans which eventually served as a major basis for standardisation developed from what Den Besten (1989) calls Proto Afrikaans II, a “Cape Dutch” spoken by settlers and slave traders. In BK&DS Afrikaans served communicative purposes at all levels of society, including the practice of religious rites, and medium of instruction in Cape Muslim schools. An orthography based on Arabic script was developed for these functions, and was the first written form of Afrikaans (Ponelis 1994:110-111; Botha 1989; Davids 1991; 2011). During Afrikaner nationalism, Kaaps was circumvented as a resource for standardising Afrikaans (Ponelis 1998; Odendaal 2012), a further example of how Kaaps was marginalised throughout South Africa’s history.

​

Present-day Kaaps

Speaking of ‘Kaaps’ could create the misimpression that the discussion centres on a single (relatively) homogeneous system. In fact, to the extent that we could consider an idiolect also to be a variety, and expect to find high degrees of both inter- and intra-speaker variation even within the same variety, Kaaps could be considered a ‘cluster’ of varieties where socioculture and community of practice are important factors (Van Rensburg 1989, Klopper 1983, De Vries 2006). Since individuals always belong to various communities of practice, even a hypothetical ‘monolingual’ speaker would hold within her linguistic repertoire various varieties of Kaaps. In this sense, Kaaps speakers (like speakers in general) are ‘multidialectal’ (i.e. in possession of linguistic resources that have high degrees of similarity / mutual intelligibility). Many Kaaps speakers are also proficient in standardised Afrikaans, through formal schooling. Additionally, most Kaaps speakers are also proficient in English and other languages. In this sense, Kaaps speakers in general are also multilingual (i.e. in possession of linguistic resources that have low degrees of similarity / mutual intelligibility outside of the contextualised multilingual practice).

Heavy influence of English is often remarked as a striking property of Kaaps (Ponelis 2009; Hendricks 2016). Some examples provided by Hendricks (2016:7) include lexical ‘borrowings’ (1), which typically inflect and derive on the basis of productive Afrikaans morphology (2). Other hybrid forms include English-Afrikaans compounds (3) and particle verbs (4). In (1-4) the Afrikaans is in ordinary text and English items are underlined.

(1) Loop,  en  moenie    iets            try nie, want      dan  finish ek self  die job.

    walk   and must-not something try neg because then finish  I   self the job

    ‘Leave, and don’t try anything, cause then I’ll finish the job myself.’

​

(Son, “Oom Sonnie”, 17/4/15)

 (2)      (a) unsettledgeid (lit.: unsettled-ness)

(Snyders: 1983:55)

            (b) geleave (lit.: PARTICIPLE-leave; “left”)

(Small: 1964:81)

(3)   hawkerkinners (lit.: hawker-children)

(Small 1975:64)

(4)   ytgewear (lit.: out-PARTICIPLE-wear; “worn out”)

(Snyders, Die Burger Landelik, 10/5/2001; from Hendricks 2016:22)

Many features, however, do not seem to result from English influence. Hendricks (2016:21-25) lists reduplication of plural, comparative, and possessive morphology (5), as well as adposition doubling; heightened tendency for non-emphatic, non-phonologically conditioned adjectival inflection (6); vir-marking of human objects, reminiscent of differential object marking (7); regularised inflection of auxiliaries (8); heightened tendency to extrapose adverbials and object arguments, and to use of resumption with subjects (9).

(5) onse       way van doen   (<ons)

      our-POS  way of   do

     ‘our way of doing (things)’

(De Vries 2010:56)

(6) korte      wit   skirtjie    (<kort)

      short-E white skirt

     ‘short white skirt.’

(De Vries 2010:28)

(7) Hy’t      vir Kietie  liefgehad.

     He-had for Kietie love-had

    ‘He loved Kietie.’

(Small 1964:51)

(8) Het gekan (<kon)

     has GE-can

     ‘could’

(Small: 1964:81; 54)

(9) Kanna hy’t      nou baie    geld     gemaak.

     Kanna he-had now much money made

    ‘Now Kanna, he made a lot of money.’

(Small 1964:60)

​

A Sociolinguistic Perspective on Multilingualism

Turns in sociolinguistics have problematised categories like ‘dialect’, ‘language’, ‘standard’, ‘mother-tongue’, etc. as reinforcing harmful socio-politically motivated language ideologies (Weber & Horner 2011), and Dyers (2016) points out the particular relevance of this for Kaaps. Thus, in accordance with current work in sociolinguistics (e.g. Blommaert 2010; Hornberger & Link 2012; Stroud 2015;  Dyers 2015a,b; Williams 2016b; Antia 2018), multilingualism – and language more generally – is conceptualised as dynamic, hybrid, and innovative in nature, characterised by high degrees of non-discreet inter- and intra-speaker variation. This work strongly emphasises that a simplistic characterisation of multilingualism as a plurality of isolated monolingualisms is not only inaccurate, but perpetuates harmful and outdated language ideologies. Instead, multilingualism is characterised by intense heteroglossia, which entails linguistic hybridity that intersects language varieties, social and cultural contexts (Bakhtin 1981; Bhatt 2008). The linguistic practices of multilingual, multidialectal individuals, societies, and cultures performing identity by drawing on sets of linguistic resources in linguistic repertoires is described as ‘translanguaging’ (Jørgensen & Lytra 2008) or ‘languaging’ (Pietikäinen et al. 2008; Creese & Blackledge 2010).

​

A socio-syntactic perspective on grammatical hybridisation and innovation

One vastly underdiscussed way in which speakers exercise their sociolinguistic agency is by exploiting grammatical patterns through combining properties of systems in their repertoires or through grammatical innovation (patterns created ‘from scratch’). E.g. Work on Germanic urban youth varieties – e.g. Kiezdeutsch, Danish Urban Vernacular, Norwegian Urban Vernacular, and Swedish Urban Vernacular – has shown that speakers who are proficient in both the urban and standard varieties of the dominant language have innovated a V3 (‘verb-third’) pattern in the urban variety (cf. discussion summarised in Walkden 2017). V3, which entails finite verbs in main clauses taking third position, is a reorganisation of the V2 pattern in the respective standard varieties, facilitating more transparent information-structuring for narratives (cf. discussion in Problem Statement on superficial V1 in Kaaps, and how this innovation relates to structuring discourse). Another kind of example of sociolinguistic agency exercised through syntax, as Cornips (2014) discusses, involves conscious deviation from a pattern in the codified variety: in relation to the Dutch gender system for common/neuter nouns, speakers of an urban variety in Rotterdam deviated consciously from the “correct” pattern in codified Dutch to mark group identity (cf. also Nortier and Dorleijn 2008: 132). It was also found that bi-dialectal children acquired relevant grammatical patterns in both varieties at a younger age than either monolingual or bilingual children did, suggesting not only that being bi-dialectal is qualitatively different from being bilingual, but also that being bi-dialectal carries an advantage for language acquisition.

​

Neo-emergentist generative grammar and syntactic ecology

We understand from the discussion above that multilingual communication is essentially hybrid and innovative, yet that grammatical patterns in these emerging systems are not a random ‘anything goes’. It is well understood in generative grammar (Chomsky 1995, 2000, 2005) that acquirers impose a special kind of order on linguistic input. They create complex, coherent systems whose inner working could not have been discerned from the input. This ‘logical problem of language acquisition’ is pertinent in multilingual contexts, where input is highly mixed and variable, as in the case of Kaaps. Neo-emergentist generative theory (Biberauer 2017a,b) offers the ideal lens through which to make sense of the complex facts, and therefore also stands to benefit through integrating the linguistic facts of Kaaps. Tenets are that acquirers (i) pay attention to very particular cues in the input (= systematic departures from arbitrary, Saussure-style form-meaning mappings), (ii) postulate new grammatical features only when the existing features will not suffice (= a general cognitive mechanism to maximise minimal means, which also leads speakers of hybrid systems to exploit all the linguistic means at their disposal), and (iii) follow a general cognitive learning strategy in regulating the function of grammatical features already present in the system, and in deciding when to postulate new ones. These factors, in tandem with an impoverished Universal Grammar, see acquirers creating integrated grammatical systems (from even highly variable input) which then evolve and diversify on individual and societal levels (Mufwene 2001; 2005; 2009, Aboh 2015).

When grammatical properties from different systems combine, or when patterns are innovated, changes respect the ‘syntactic ecology’ of the systems in play. It has been understood for a while that grammatical properties of a system are integrated and form clusters that determine the pre-existing conditions of a (potentially mixed) grammatical system such that a change in one part of the system could trigger almost simultaneous changes and ‘cascading effects’ in seemingly unrelated parts of the grammar (Biberauer & Roberts 2008 et seq.). This is what Adger (2017) calls a 'syntactic ecology'. The complexity and interconnectedness of these systems are typically unobservable until the nature and identity of co-varying grammatical properties are probed, one of the project’s objectives.