Mean Length of Utterance: A study of early language development in four Southern Bantu languages.

Mean Length of Utterance (MLU) has been widely used to measure children's early language development in a variety of languages. This study investigates the utility of MLU to measure language development in four agglutinative and morphologically complex Southern Bantu languages. Using a variant of MLU, MLU3, based on the three longest sentences children produced, we analysed the utterances of 448 toddlers (16-32 months) collected using the MacArthur-Bates Communicative Development Inventory, a parent-report tool. MLU3, measured in words (MLU3-w) and morphemes (MLU3-m), significantly correlated with age and other indices of language growth (e.g., grammar and vocabulary). MLU3 measures also accounted for significant variance in language development particular morphosyntactic development. Our results suggest that MLU3-m is a more sensitive measure than MLU3-w. We conclude that MLU measured in morphemes provides a useful addition to other indices of language development in these kinds of morphologically complex languages.

Sociocultural Factors Affecting Vocabulary Development in Young South African Children.

Sociocultural influences on the development of child language skills have been widely studied, but the majority of the research findings were generated in Northern contexts. The current crosslinguistic, multisite study is the first of its kind in South Africa, considering the influence of a range of individual and sociocultural factors on expressive vocabulary size of young children. Caregivers of toddlers aged 16 to 32 months acquiring Afrikaans (n = 110), isiXhosa (n = 115), South African English (n = 105), or Xitsonga (n = 98) as home language completed a family background questionnaire and the MacArthur-Bates Communicative Development Inventory (CDI) about their children. Based on a revised version of Bronfenbrenner's (1977) ecological systems theory, information was obtained from the family background questionnaire on individual factors (the child's age and sex), microsystem-related factors (the number of other children and number of adults in the child's household, maternal level of education, and SES), and exosystem-related factors (home language and geographic area, namely rural or urban). All sociocultural and individual factors combined explained 25% of the variance in expressive vocabulary size. Partial correlations between these sociocultural factors and the toddlers' expressive vocabulary scores on 10 semantic domains yielded important insights into the impact of geographic area on the nature and size of children's expressive vocabulary. Unlike in previous studies, maternal level of education and SES did not play a significant role in predicting children's expressive vocabulary scores. These results indicate that there exists an interplay of sociocultural and individual influences on vocabulary development that requires a more complex ecological model of language development to understand the interaction between various sociocultural factors in diverse contexts.

Fast mapping of verbs in Afrikaans-speaking children from low and mid socioeconomic backgrounds and children with language impairment.

Children with LI are at a disadvantage because they typically have smaller vocabularies than their TD peers. However, children from low-SES households often also have smaller vocabularies and can thus be misdiagnosed with LI. The purpose of this study was to compare the fast-mapping (FM) skills of 3 groups of 4- to 9-year-olds: typically developing (TD) children with low socioeconomic status (SES) and mid SES, and mid-SES children with language impairment (LI), to ascertain whether FM is affected by SES. The FM items of the Diagnostic Evaluation of Language Variation were administered to 253 TD mid-SES children, 75 TD low-SES children, and 36 children with LI. On FM with novel verbs, the TD groups differed significantly (p = .001) as did the TD mid-SES and LI groups (p = .046), but not the TD low-SES and LI groups (p = 1.00). On a comparable task with real verbs, the TD mid-SES group also differed significantly from the TD low-SES (p < .001) and LI group (p = .018), with no significant difference between the latter two groups (p = 1.00). FM has been proposed as a non-SES-sensitive measure but, unlike previous studies, we found FM to be affected by SES, rendering it an insufficiently unbiased measure for our sample of children. Further thought should be given to measures that can successfully differentiate between children with LI, and children from low-SES backgrounds, possibly rendering scores adjusted for SES, so that targeted intervention can be offered.

Noun and verb knowledge in monolingual preschool children across 17 languages: Data from Cross-linguistic Lexical Tasks (LITMUS-CLT).

This article investigates the cross-linguistic comparability of the newly developed lexical assessment tool Cross-linguistic Lexical Tasks (LITMUS-CLT). LITMUS-CLT is a part the Language Impairment Testing in Multilingual Settings (LITMUS) battery (Armon-Lotem, de Jong & Meir, 2015). Here we analyse results on receptive and expressive word knowledge tasks for nouns and verbs across 17 languages from eight different language families: Baltic (Lithuanian), Bantu (isiXhosa), Finnic (Finnish), Germanic (Afrikaans, British English, South African English, German, Luxembourgish, Norwegian, Swedish), Romance (Catalan, Italian), Semitic (Hebrew), Slavic (Polish, Serbian, Slovak) and Turkic (Turkish). The participants were 639 monolingual children aged 3;0-6;11 living in 15 different countries. Differences in vocabulary size were small between 16 of the languages; but isiXhosa-speaking children knew significantly fewer words than speakers of the other languages. There was a robust effect of word class: accuracy was higher for nouns than verbs. Furthermore, comprehension was more advanced than production. Results are discussed in the context of cross-linguistic comparisons of lexical development in monolingual and bilingual populations.

The comprehension and production of quantifiers in isiXhosa-speaking Grade 1 learners.

BACKGROUND: Quantifiers form part of the discourse-internal linguistic devices that children need to access and produce narratives and other classroom discourse. Little is known about the development - especially the prodiction - of quantifiers in child language, specifically in speakers of an African language. OBJECTIVES: The study aimed to ascertain how well Grade 1 isiXhosa first language (L1) learners perform at the beginning and at the end of Grade 1 on quantifier comprehension and production tasks. METHOD: Two low socioeconomic groups of L1 isiXhosa learners with either isiXhosa or English as language of learning and teaching (LOLT) were tested in February and November of their Grade 1 year with tasks targeting several quantifiers. RESULTS: The isiXhosa LOLT group comprehended no/none, any and all fully either in February or then in November of Grade 1, and they produced all assessed quantifiers in February of Grade 1. For the English LOLT group, neither the comprehension nor the production of quantifiers was mastered by the end of Grade 1, although there was a significant increase in both their comprehension and production scores. CONCLUSION: The English LOLT group made significant progress in comprehension and production of quantifiers, but still performed worse than peers who had their L1 as LOLT. Generally, children with no or very little prior knowledge of the LOLT need either, (1) more deliberate exposure to quantifier-rich language or, (2) longer exposure to general classroom language before quantifiers can be expected to be mastered sufficiently to allow access to quantifier-related curriculum content.

A comparison of proficiency levels in 4-year-old monolingual and trilingual speakers of Afrikaans, isiXhosa and South African English across SES boundaries, using LITMUS-CLT.

This study investigated how trilinguals fare on the cross-linguistic lexical tasks (CLT)-Afrikaans, -isiXhosa and -South African English (SAE) (cf. Haman et al., 2015) compared to monolingual controls, and whether the CLT-Afrikaans renders comparable results across socioeconomic status (SES) backgrounds. The LITMUS-CLTs were administered to 41 low SES 4-year-olds (11 trilinguals; 10 monolingual speakers of Afrikaans, isiXhosa and SAE) and the LITMUS-CLT-Afrikaans to 11 mid-SES 4-year-old monolinguals. Results (a) indicate that trilinguals' proficiency in their exposure-dominant language did not differ significantly from monolinguals' proficiency, but their proficiency in their additional two languages was significantly lower than monolinguals' proficiency; (b) reflect the extent, but not current amount, of exposure trilinguals had had over time to each of their languages; and (c) show that low and mid-SES monolinguals differed significantly on noun-related, but not verb-related, CLT measures. Possible reasons for and the clinical implications of these results are discussed.

Ratings of age of acquisition of 299 words across 25 languages: Is there a cross-linguistic order of words?

We present a new set of subjective age-of-acquisition (AoA) ratings for 299 words (158 nouns, 141 verbs) in 25 languages from five language families (Afro-Asiatic: Semitic languages; Altaic: one Turkic language: Indo-European: Baltic, Celtic, Germanic, Hellenic, Slavic, and Romance languages; Niger-Congo: one Bantu language; Uralic: Finnic and Ugric languages). Adult native speakers reported the age at which they had learned each word. We present a comparison of the AoA ratings across all languages by contrasting them in pairs. This comparison shows a consistency in the orders of ratings across the 25 languages. The data were then analyzed (1) to ascertain how the demographic characteristics of the participants influenced AoA estimations and (2) to assess differences caused by the exact form of the target question (when did you learn vs. when do children learn this word); (3) to compare the ratings obtained in our study to those of previous studies; and (4) to assess the validity of our study by comparison with quasi-objective AoA norms derived from the MacArthur-Bates Communicative Development Inventories (MB-CDI). All 299 words were judged as being acquired early (mostly before the age of 6 years). AoA ratings were associated with the raters' social or language status, but not with the raters' age or education. Parents reported words as being learned earlier, and bilinguals reported learning them later. Estimations of the age at which children learn the words revealed significantly lower ratings of AoA. Finally, comparisons with previous AoA and MB-CDI norms support the validity of the present estimations. Our AoA ratings are available for research or other purposes.

The challenge of linguistic and cultural diversity: Does length of experience affect South African speech-language therapists' management of children with language impairment?

BACKGROUND: South African speech-language therapists (SLTs) currently do not reflect the country's linguistic and cultural diversity. The question arises as to who might be better equipped currently to provide services to multilingual populations: SLTs with more clinical experience in such contexts, or recently trained SLTs who are themselves linguistically and culturally diverse and whose training programmes deliberately focused on multilingualism and multiculturalism? AIMS: To investigate whether length of clinical experience influenced: number of bilingual children treated, languages spoken by these children, languages in which assessment and remediation can be offered, assessment instrument(s) favoured, and languages in which therapy material is required. METHOD: From questionnaires completed by 243 Health Professions Council of South Africa (HPCSA)-registered SLTs who treat children with language problems, two groups were drawn:71 more experienced (ME) respondents (20+ years of experience) and 79 less experienced (LE) respondents (maximum 5 years of experience). RESULTS: The groups did not differ significantly with regard to (1) number of children(monolingual or bilingual) with language difficulties seen, (2) number of respondents seeing child clients who have Afrikaans or an African language as home language, (3) number of respondents who can offer intervention in Afrikaans or English and (4) number of respondents who reported needing therapy material in Afrikaans or English. However, significantly more ME than LE respondents reported seeing first language child speakers of English, whereas significantly more LE than ME respondents could provide services, and required therapy material, in African languages. CONCLUSION: More LE than ME SLTs could offer remediation in an African language, but there were few other significant differences between the two groups. There is still an absence of appropriate assessment and remediation material for Afrikaans and African languages, but the increased number of African language speakers entering the profession may contribute to better service delivery to the diverse South African population.

Towards a dialect-neutral assessment instrument for the language skills of Afrikaans-speaking children: the role of socioeconomic status.

The aims of the study were to establish whether there is a correlation between the socioeconomic background of Afrikaans-speaking children and their performance on a dialect-neutral language test, and to ascertain whether the allowance the test currently makes for parental education level is sufficient. The Afrikaans version of the Diagnostic Evaluation of Language Variation (Seymour, Roeper & de Villiers, 2005a ) was administered to 231 Afrikaans-speaking children age 4 ; 0 to 9 ; 11 from various socioeconomic backgrounds. A positive correlation was found between the composite language scores as well as the scores for each of the language domains (syntax, pragmatics, semantics) and the primary female caregivers' highest level of education. Children with father figures present did not outperform those without. It appears that the original manner of accommodating parental education level in interpreting the children's language scores on the test is sufficient and need not be refined for the South African context.

Production of tense morphology by Afrikaans-speaking children with and without specific language impairment.

PURPOSE: To establish whether the predictions of the extended optional infinitive (EOI) hypothesis (Rice, Wexler, & Cleave, 1995) hold for the language of Afrikaans-speaking children with specific language impairment (SLI) and whether tense marking is a possible clinical marker of SLI in Afrikaans. METHOD: Production of tense morphology was examined in 3 groups of Afrikaans-speaking children-15 with SLI who were 6 years old, 15 typically developing (TD) 4-year-olds matched on mean length of utterance, and 15 TD 6-year-olds-using both elicited and spontaneously produced verb forms. RESULTS: On the sentence completion task, children with SLI fared on par with 4-year-olds and worse than age-matched peers. However, in terms of spontaneous production of morphemes pertaining to tense, children with SLI fared worse than both TD groups. Furthermore, children with SLI mostly made the same types of errors as 4-year-olds, although some errors were unique to the SLI group. Most errors entailed omissions, of modal and temporal auxiliaries as well as of copula be. CONCLUSION: The errors offer support for the EOI hypothesis. Tense marking has the potential to be a clinical marker of SLI in Afrikaans, but further research with larger groups of Afrikaans-speaking children, including children of other ages, is needed to confirm this.

Literacy development of English language learners: the outcomes of an intervention programme in grade R.

This study aims to contribute to the knowledge base on the status and development of emergent literacy skills of learners receiving formal education in their second or additional language. The focus is on young English language learners (ELLs), i.e. learners whose home language is not English but who have English as their language of teaching and learning. This article reports on a study that investigated ELLs' emergent literacy skills prior to entering grade 1 and then evaluated the effectiveness of an evidence-based stimulation programme on early literacy skills in the South African context. Using a quasi-experimental design, ELLs' emergent literacy skills were assessed with an adapted version of 8 of the subtests of the Emergent Literacy Assessment battery (Willenberg, 2004) and were compared to those of English first language (L1) and of ELL control groups, both before and after the 8-week purpose-designed programme. While learners showed significant improvement on 6 of the 8 subtests, the programme did not significantly improve ELLs' skills in comparison to those of the control groups. Possible independent variables contributing to the dearth of intervention effect include socio-economic status, learners' L1, and teacher- and classroom-specific characteristics, all of which were considered in this study. Clinical implications for speech-language therapists with regard to assessment, intervention, service delivery and outcome measures are highlighted.

Methodological issues in the calculation of mean length of utterance.

Mean length of utterance (MLU) is widely used as a diagnostic, monitoring and group matching measure. This study investigated methodological issues regarding the calculation of MLU. The aim was to establish whether different calculation procedures render different MLUs, and whether there is a high correlation between MLU measured in words (MLU-w) and in morphemes (MLU-m). Language samples from 15 Afrikaans-speaking 6-year-olds with and 15 with ot specific language impairment were analyzed. MLU was calculated eight times for each participant, varying sample size (50 or 100 utterances), unit counted (words or morphemes) and calculation method (traditional or alternate). Significant differences in resultant MLUs were due to the calculation method used, rather than sample size or unit counted. A high positive correlation (>0.96) between MLU-w and MLU-m was found. The results imply that researchers and clinicians should clearly state their MLU calculation procedures, otherwise reliable comparisons between MLU scores from different sources cannot be made. The results furthermore imply that, in order to generalize research results and make diagnostic decisions based on MLU, consistent procedures should be used, not only with regard to language sampling, but also to MLU calculation.

Comparison of conversation, freeplay, and story generation as methods of language sample elicitation.

The spontaneous language sample forms an important part of the language evaluation protocol (M. Dunn, J. Flax, M. Sliwinski, and D. Aram, 1996; J. L. Evans and H. K. Craig, 1992; L. E. Evans and J. Miller, 1999) because of the limitations of standardized language tests and their unavailability in certain languages, such as Afrikaans. This study examined 3 methods of language elicitation, namely conversation (CV), freeplay (FP), and story generation (SG), on the following 5 measures to determine which method is best for clinical practice: number of utterances, variety of syntactic structures, mean length of the utterance (MLU), number of syntactic errors, and proportion of complex syntactic utterances as elicited from ten 5-year-old, Afrikaans-speaking boys. FP elicited significantly more utterances than did SG but elicited a smaller proportion of complex syntactic structures than did CV and SG. Furthermore, SG elicited longer utterances than did CV or FP. It is recommended that SG be used in clinical practice with 5-year-olds if the clinician wishes to observe maximum behavior. Where typical behavior is to be evaluated, the clinician can select a language elicitation method that best suits the client's personality and communication style, bearing in mind that FP does elicit a larger language sample.