https://orcid.org/0000-0002-6441-9746
![Sample our Education journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5932/TOC-Subject-Sample-2021-Education.jpg"})
ABSTRACT
Languages differ in their preferences for particular intensifying constructions. While intensifying adjectival compounds (IACs) (e.g. ijskoud, ice-cold) are productively used to express intensification in Dutch and English, in French this construction is hardly productive. Consequently, French-speaking learners may encounter difficulties acquiring IACs in Dutch/English. As part of a research project on CLIL in French-speaking Belgium, we explore the effect of CLIL on the acquisition of IACs in the target language (TL) Dutch/English through a multiple-choice test. The results confirm that CLIL students (learning English/Dutch) develop greater receptive knowledge of these constructions. Furthermore, the more frequent IACs are more likely to be recognized by the learners. Moreover, even when the CLIL effect is considered alongside other factors, such as the students' extracurricular exposure to the TL and their overall vocabulary, CLIL is still an important predictor of the learners' receptive knowledge of English IACs, in addition to productive and receptive vocabulary. By contrast, current informal contact with the TL and receptive vocabulary are significant predictors of learners' receptive knowledge of Dutch IACs, but CLIL does not significantly contribute to the regression model for the latter language.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 This work was supported by a Coordinated Research Project grant (ARC 14/19-061) from the Université catholique de Louvain (UCLouvain) and the Université de Namur (UNamur), awarded to Philippe Hiligsmann (spokesperson; UCLouvain), Benoît Galand (UCLouvain), Laurence Mettewie (UNamur), Fanny Meunier (UCLouvain), Arnaud Szmalec (UCLouvain), and Kristel Van Goethem (F.R.S.-FNRS & UCLouvain). We would like to thank Amélie Bulon, Audrey De Smet, Luk Van Mensel, and Morgane Simonis for their assistance in data collection, as well as Emmanuella Ahishakiye, Natacha Buntinx, Mathilde Dumont, Laura Nolevaux, Heloise Pierret and Léticia Raemdonck for helping us organize and transcribe the data. We also warmly thank the schools and students that took part in the study.
2 This is different from the situation in other countries, e.g. Germany (Rumlich Citation2016) or the Netherlands (Maljers Citation2007).
3 Socio-affective variables and cognitive variables may also have an impact on the learners’ knowledge of IACs, but considering these variables goes beyond the scope of this article. However, these variables are analyzed in detail in (De Smet et al. Citation2019 and Simonis et al. Citationforthcoming a in this special issue).
4 Note that the studies by Arnaud and Savignon (Citation1997) and Bahns and Eldaw (Citation1993) pre-date the corpus revolution. The classification of ‘collocations’ in these studies may be less reliable than the later studies by Koya (Citation2005) and Gyllstad (Citation2007), as there was no access to bigrams yet.
5 The index, established in 2011 by the education authorities, is based on a number of socio-economic criteria related to the neighborhood in which the students of a particular school reside. URL: www.gallilex.cfwb.be/document/pdf/36474_000.pdf.
6 LimeSurvey: https://www.limesurvey.org/about-limesurvey/team-contributors (team: Schmitz, Gac, Härstedt, Flür, Kondziella, Chirițoiu).
9 We used a test of receptive vocabulary of Standard American English, while the reference corpus is based on British English. However, we checked the IACs of our English test in British and American dictionaries and made sure that they occurred in both varieties. The IACs in the Dutch test also occur in both Flemish Dutch and Netherlandic Dutch. We were unable to control for the variety of Dutch or English (if only one) that was taught to the students.
10 The target-language input is not limited to teaching materials, and also includes the oral productions of the teachers(s) and other students, and even target-language contact outside the classroom (see Van Mensel et al. Citationforthcoming). Considering all these types of input and their relative effect on target-language acquisition of intensification goes beyond the scope of this study.
11 Koizumi (Citation2012) found that MTLD was least affected by text length compared with TTR and other recent indices (e.g. Guiraud index and D), when used with texts of at least 100 tokens.
12 We used SPSS version 24 for the statistical analysis.
13 First, the predictors showed some variation in value, which is in line with the assumption of non-zero variance. Second, there was no perfect multicollinearity between two or more of the predictors (VIF scores are > 1< 10). Third, the assumption of independent errors was fulfilled. The residuals are only slightly positively correlated in the regression models for TL English (Durbin-Watson = 1.835) and TL Dutch (Durbin-Watson = 1.636). Fourth, we plotted the regression standardized residual against the regression standardized predicted value in a scatter plot. The points are randomly and evenly dispersed throughout the plot, hence the assumptions of linearity and homoscedasticity are met. Fifth, the residuals are normally distributed. Sixth, the assumption of independence is also met – each value of the outcome variable comes from a different student. Seventh, the assumption of linearity was fulfilled because the scatter plot showed a linear relationship rather than any other.
14 Note that there is little research on TL input in CLIL. To our knowledge, Moller (Citation2017) is the only study so far that has analyzed CLIL teaching materials.
15 We included spelling variations in the searches, such as brandnew and brand new.
Additional information
Notes on contributors
Isa Hendrikx
Isa Hendrikx is invited lecturer at the University of Liège. She obtained a PhD in 2019 from the Université catholique de Louvain (Belgium), where she participated in a multidisciplinary research project on Content and Language Integrated Learning (CLIL). Her main areas of research are second language acquisition and learner corpus research, specifically the acquisition of intensifying constructions in L2 Dutch and English by French-speaking learners, drawing on the theoretical framework of Construction Grammar.
Kristel Van Goethem
Kristel Van Goethem is Research Associate of the Belgian National Research Fund (F.R.S.-FNRS) at the Université catholique de Louvain. Her research focuses on morphology, contrastive linguistics, Construction Grammar and corpus linguistics. She was one of the promotors of the research project "Assessing Content and Language Integrated Learning (CLIL): Linguistic, cognitive and educational perspectives” (2014–2019).