Heritage Bilingualism and the Acquisition of English as an Additional Language (L3) in the German Context: A Longitudinal Study

Abstract

This dissertation investigated the links between heritage bilingualism and the longitudinal development of L3 English in the German secondary education context by employing a longitudinal design. In addition to examining L3 English longitudinal growth trajectories and growth rate in heritage bilingual students, the dissertation research examined which linguistic (i.e., heritage language proficiency, majority language proficiency, and L1 background) and extra-linguistic (i.e., cognitive ability, socioeconomic status, and school type) factors exerted an impact on L3 English development. The participants were 301 heritage bilingual students (Russian-German=123, Turkish-German=178) sampled from 72 schools across eight federal states in Germany. The students participated in the first three waves of the MEZ project (MEZ—2014–2019), which examined the multilingual development of secondary school students in the German educational context. The longitudinal data initially were examined by using a set of descriptive and inferential statistics (e.g., analysis of difference and correlational analysis). The main analyses were conducted by using the Latent Growth Curve Modelling methodology in the context of Structural Equation Modelling. The latent growth curve modelling results showed that the baseline, the initial level of L3 English, was significantly positive. Additionally, a slight yet statistically significant variation was detected at the baseline in L3 English scores. The mean of the linear slope, the rate of growth, was found to be slightly negative, with scores decreasing by 0.061, on average, across three measurement time points. Moreover, the results demonstrated that heritage bilingual students did not vary considerably in terms of their L3 English growth trajectories over time. Furthermore, it was found that heritage bilingual students’ initial L3 English scores did not necessarily predict their L3 English growth across three measurement time points. Concerning the links between linguistic and extra-linguistic variables and L3 English growth, the analyses revealed that, notably the effect of the majority language proficiency, German, was significant and positive across all three MTPs for the baseline. The effect of heritage language proficiency was significant and positive at MTP1, negative and not significant at MTP2, and negative and significant at MTP3. The impact of school type was statistically significant, meaning that attending the Gymnasium was strongly associated with a higher L3 English proficiency initial level. The effect of socioeconomic background was negative but statistically not significant while the effect of cognitive ability was positive and statistically significant. When it comes to L3 English growth, the results indicated that both heritage language proficiency and the majority language proficiency did not seem to predict linear growth over time. The impact of German proficiency on growth was positive at MPT2 and MPT3, while the impact of heritage language proficiency was negative across all MPTs, yet the associated p values did not reach statistical significance. School type remained a marginally significant predictor for L3 English growth. Socioeconomic background contributed positively to L3 English longitudinal growth whereas cognitive ability was not found to impact the growth. Language background was a statistically significant predictor for L3 English growth, particularly indicating an advantage for the Russian-German group compared to the Turkish-German group. Overall, the results did not indicate a clear bilingual advantage due to a lack of a positive pattern in L3 growth trajectories and growth rate. The lack of linear growth may stem from the influence of extra-linguistic, i.e., contextual factors, participants’ age, educational level, and bilingual profiles, and the status of heritage languages in educational, societal, and home settings.