2.3 Vocabulary acquisition and important considerations
2.3.8 The role of multilingual abilities on language acquisition
Developmentally, bilingual language learners exhibit differences in vocabulary development to monolingual language learners, for example, they are slower in each language (Bialystok et al., 2010 in Potgieter and Southwood, 2016: 2). When counting the number of items that the child produces or comprehends in both languages, however, it can be said that bilingual children develop language skills similarly to monolingual children (Junker and Stockman, 2002 and Pearson, Fernández and Oller, 1993 in Vogt, Mastin and Aussems, 2015: 2). Just as there is a need for monolingual assessments, there is a need for bilingual or even trilingual assessments to be created by the Southern African CDI team,14 since as Potgieter and Southwood (2016: 1) conjecture, “[bilingual children] are being misdiagnosed with language impairment and word finding disorder, when in reality they may be typically developing bilinguals who, given sufficient time and exposure, would catch up with their monolingual peers”.
Potgieter and Southwood (2016) test the performance of developing trilingual speakers and find that when SES and age are controlled for, trilingual children’s proficiency in their exposure-dominated language is not significantly different from the monolingual control group, but their proficiency in the additional two languages is behind that of the monolinguals. With regard to grammar, however, Law and Roy (2008) posit the grammatical abilities of bilingual children to be separately acquired in each language (as the US and Spanish version of the CDI find).
2.3.8.1 Implications for monolingual vocabulary assessments
August, Shanahan and Escamilla (2009: 436) note that “bilingualism, not monolingualism, is now the global norm” and argue that although this is a fact, monolingualism still tends to be privileged as if it were the norm. The issue of whether South African children are largely monolingual or multilingual is
14 See acknowledgements for the names and affiliations of the members of the team.
27 contested. For example, Heugh (2002) argues against Deumert’s (2000) view that South African schools are largely multilingual and contends that the vast majority of teachers are monolingual (Heugh, 2002), although she does not say the same of the children. In the same article, she concedes that children in South Africa have “multiple proficiencies” in a variety of languages (Heugh, 2002: 188).
I believe the multilingual contexts that South African children may find themselves in may have influence on monolingual language, as Deumert (2010: 20) deftly captures:
space organises and defines patterns of multilingualism, sociolinguistic norms, repertoires and identities. Locality and urban space thus have the potential for semiotisation and interact with other indicators of social stratification.
Brookes and Legkoro (2014) write that urban male youth language varieties (often called tsotsitaal) have become a widespread phenomenon of urban South African township life. I would expect, for example, the language of these male youth street social networks to form part of the language and lexicon that an urban isiXhosa-speaking child is exposed to, since scholars such as Ntshangase (2002) and Makhudu (2002) are of the view that these varieties are being taken up by the general population as emerging languages. These languages stem from the youth’s desire to create a social status and identity (Brookes, 2014) that supersedes their SES. It would therefore be misguided to disregard the role that exposure to languages and their varieties plays, whether stemming from SES or other exogenous factors. Despite this study being rural based, the above points to the important role that language exposure plays, and moreover it would be naïve to disregard that geographically mobile people from the cities return periodically to rural areas where their language will play an influencing role in shaping rural language and child input. Indeed, Deumert, Inder and Maitra (2005) suggest that rural-urban migrants maintain strong ties with their rural families through regular visits.
Calteaux (1996) notes that children may be exposed to lexical adoption from other African languages too, due to migration and intermarriage, with the result that a child’s home language can be that of either the mother or the father, although it generally ends up being the mother’s due to contact time spent with the child. Most often parents make a conscious decision about which language to use in the home, although sometimes middle class African language speakers use English as a lingua franca (Calteaux, 1996).
Bylund (2014) acknowledges that younger generations of isiXhosa speakers are growing up using more English than ever before. This means that infants and toddlers may be hearing more English (from
28 parents and older siblings or, for the more economically privileged, from English children’s books read out to them by caregivers). But the fact that, as Posel and Zeller (2016) report, English as a second language is acquired through education and labour markets rather than the home suggests that when English lexical items are used by young isiXhosa-speaking children, they are considered to be isiXhosa words and not English ones (like idrink for ‘drink’ and iruf for ‘roof’). Bylund’s observation may be true even for children who have had no exposure to English15 as a second language, and such words as iruf and idrink therefore need to be incorporated into the monolingual CDI (see Demuth, 2000 on the incorporation of loan words into noun classes). This process of language change due to language contact is by no means a new phenomenon16 (see Koopman, 1999), and Demuth (2003) draws attention to children’s role in solidifying the developments of language change.
Contemporary translanguaging theory promotes movement away from naming languages and compels linguistics to consider fluid ways in which language is used (García, 2013). From this point of view, using labels that assign certain lexical items to either ‘English’ or ‘isiXhosa’ can re-impose colonially defined language boundaries and restrict understanding of linguistic realities and the dynamic nature of languages (García, 2013; see also Dowling and Krause, 2018). As the adaptations of the CDIs stand, it would be tricky at this stage to do away with named languages, but I urge future adaptors and language pathologists to be aware of the multilingual reality of South African children and the fluid way in which language is used – that is, even if parents raise their children to be monolingual isiXhosa speaking, lexical items of other named languages may be incorporated into a ‘monolingual isiXhosa’ child’s lexicon.
When a child has lexical exposure to another language that is not used fluidly within the isiXhosa language to a frequent enough extent, a monolingual CDI unfortunately cannot account for this fact.
Thus just as there is a need for monolingual assessments, there is a need for bilingual or even trilingual assessments to be created by the Southern African CDI team. Notwithstanding the manifold linguistic skills observed in South African children, the goal of my current research is to create norms for monolingual children. By so doing I hope I can eventually combine monolingual CDIs of different
15 This is in line with Deumert’s (2010) finding that isiXhosa maintenance (monolingualism) is more strongly correlated to the concentration of isiXhosa speakers in a given area rather than to economic factors.
16 It is in fact the very process that lead to the adoption of clicks in isiXhosa from Khoisan languages (Demuth, 2003).
29 languages (for example isiXhosa and South African English) to create a mixed inventory and re-pilot the combined version to create CDIs for bilingual or, eventually, multilingual, children. A monolingual CDI is the first step in this process since even children who are going to become multilingual by school- going age may still be monolingual in the young age cohort that the CDI is designed to serve. Ultimately, the implication is that a monolingual CDI may need to include lexical items from other languages in order to align it with the kind of speech used by South African children.