Abstract
Code-switching is a phenomenon integral to bilingual first-language acquisition (BFLA), the process whereby children simultaneously acquire two first languages from birth until reaching fluency in both. Broadly defined, code-switching refers to the alternation between two or more languages by speakers within or across sentence boundaries during spontaneous conversation. Such alternation typically arises to meet immediate communicative needs without necessarily resulting in the permanent borrowing of lexical items from one language to another.
Children may grow up acquiring either closely related or unrelated languages. Closely related languages, which are genetically related, tend to display a high degree of mutual intelligibility: speakers of the respective languages can understand one another with little prior exposure or special effort. Afrikaans and Dutch, for instance, represent such a closely related language pair, both belonging to the West Germanic branch and showing considerable overlap in vocabulary, grammar and phonology. In contrast, unrelated languages lack mutual or even partial intelligibility.
Over recent decades, research on BFLA has focused predominantly on children acquiring unrelated language pairs. While some studies have examined code-switching in children growing up with related languages or varieties, these remain fewer in number and often focus on school-aged children. To date there has been a clear gap in the literature concerning preschool children acquiring Afrikaans and Dutch simultaneously as first languages.
This article addresses this lacuna by exploring and describing the code-switching patterns of a group of preschool children in South Africa raised bilingually in Afrikaans and Dutch. The study included eight children aged between 2;3 and 6;4 (years;months). Six correspondents contributed data, with two of the correspondents each observing two children. Data was gathered primarily through diary entries using a structured template, yielding a final dataset of 120 utterances that were manually analysed.
The study takes as its starting point the characteristics of congruent lexicalisation (Muysken 2000) and the specific dynamics of code-switching between closely related languages to explore and describe the code-switching patterns found in the utterances of preschool Afrikaans–Dutch children during the BFLA process. All observed instances of code-switching are examples of intra-sentential code-switching, and in certain cases even intra-word switching. Among these, nouns accounted for the largest word class of switched items (n = 59; 49%), followed by verbs (n = 27; 23%).
The analysis identified several distinct patterns: switching involving non-identical cognates; morphological integration where affixes or inflectional endings from one language attach to stems from the other; the substitution of non-cognates; mixed collocations and idioms drawing from both languages; and switching that includes false friends, sometimes resulting in subtle shifts or mismatches in meaning. These patterns reflect the complex interplay of structural similarity and cognitive flexibility characteristics of code-switching in closely related languages.
From a cognitive perspective, children acquiring closely related languages have the advantage that there is much phonological and conceptual overlap, easing the process of acquiring corresponding forms in the languages. This may explain why switching involving non-identical cognates appeared frequently in the dataset. At times it became challenging to identify definitively whether code-switching had occurred, precisely because of the phonological and lexical closeness of the two languages.
The study’s findings suggest that code-switching between closely related languages like Afrikaans and Dutch cannot be fully explained by lexical or grammatical overlap alone; it is also shaped by cognitive processes and usage-based factors, such as the child’s dominant language and the language backgrounds of interlocutors. For example, Afrikaans lexical items appeared prominently in the dataset, likely reflecting the participants’ daily exposure to Afrikaans in South Africa. A comparative study involving children residing in the Low Countries might help determine whether different patterns emerge in distinct language contact settings.
Several limitations of the present study should be noted. First, the relatively small dataset, derived from diary entries rather than spontaneous speech recordings, limits the generalisability of findings. Second, the absence of audio data precluded analysis of phonological features such as stress, intonation and rhythm. As noted in the literature, where languages are morphophonemically similar, assigning an unambiguous language index to individual utterances can be challenging. For this study, we had to rely on the parent’s analysis of the communicative situation, bur future research could address this by collecting audiovisual data – such as sound or video recordings – to clarify instances involving cognates. Smartphone applications, as successfully used in other multilingual studies, could facilitate data collection by caregivers and reduce participant fatigue.
Looking ahead, future research could draw on larger and richer datasets, including spontaneous speech and audiovisual recordings, to explore in greater depth how specific word classes, such as nouns, verbs and function words, shape code-switching behaviour in young bilinguals. It could also investigate pragmatic factors, like politeness strategies and situational context, and how these influence children’s choice to switch languages in real time. In addition, examining the acquisition and use of more complex syntactic constructions – such as the double negation in Afrikaans – could provide valuable insight into whether and how structural similarities between closely related languages facilitate or constrain code-switching during early language development.
Comparative studies could also extend these findings by looking at other closely related Germanic language pairs, including Danish, Norwegian and Swedish, to see whether similar code-switching patterns and cognitive strategies emerge in different linguistic and cultural contexts. Beyond Germanic languages, future work might explore contact situations among South African languages with high mutual intelligibility, such as Tswana and Pedi or Zulu and Xhosa, to test whether structural relatedness similarly leads to patterns of congruent lexicalisation and morphological integration. Such cross-linguistic and cross-contextual research would help to refine theoretical models and clarify the extent to which the observations reported here reflect universal processes in BFLA or are shaped by specific language pairs and contact environments.
In conclusion, this study contributes to an underexplored area of BFLA by documenting the code-switching patterns of preschool children simultaneously acquiring Afrikaans and Dutch. The findings highlight the creative and dynamic ways in which young bilinguals use their multilingual repertoires. While limited in scope, the study offers a foundation for broader research that could deepen our understanding of code-switching between structurally similar languages.
Keywords: Afrikaans; bilingual first-language acquisition; code-switching; congruent lexicalisation; Dutch; false friends; mixed collocations; morphological integration; non-cognates; non-identical cognates
- This article’s featured image was created by Napong Rattanaraktiya and obtained from Vecteezy.