Public online presentations by Scientific Advisory Board members of the research programme Transdisciplinary Approaches to Learning, Acquisition, Multilingualism (TEAM: www.team-up.lu.se)

15.15 – 16:00 Luke Plonsky, Northern Arizona University: From quality to questionable: Exploring the methods-ethics interface in the language sciences 

It wasn’t until fairly recently that we as a field began to reflect on and systematically examine the quality of our research (Gass et al., 2021). Simultaneously, and probably not coincidentally, we have also seen methodological improvements on fronts ranging from sampling and design to instrumentation, data analysis, researcher training, open science, and replication (e.g., Arndt et al., 2023; McManus, 2024; Sudina, 2023). Another focus of this meta-science movement has involved an expanded view of research ethics (De Costa, 2016; Yaw et al., 2023). In this talk, I first propose a framework for defining the notion of ‘study quality’ that includes a broad understanding of the many micro-ethical considerations found throughout the research cycle (i.e., Kubanyiova, 2008). I will focus in particular on the notion of ‘questionable research practices’ (QRPs) as discussed in recent studies from within applied linguistics and beyond (Fanelli, 2009; Larsson et al., 2023). QRPs are decisions that may be considered more or less ethical depending on contextual factors (e.g., researcher intent, resources available). In addition to presenting results on the frequency and perceived severity of QRPs in applied linguistics, I will address—and present empirical findings from meta-research on—several moments in the research cycle at the methods-ethics interface. 

16:00 – 16.45 Raphael Berthele, University of Fribourg: Clear, simple, and wrong versus complex and unusable: Insights from multi-variable approaches to language learning aptitude 

Drawing on data from several aptitude studies involving both children and adults, I illustrate strategies for reducing dimensionality (e.g. factor analysis, partial least squares correlation) as well as for exploring complex patterns of association (e.g. graph-based analyses). Across age groups, the results reveal consistent clustering of variables: language-specific measures tend to group with general cognitive tasks, while motivational dispositions exhibit clear internal structure.  I conclude by discussing how such multi-variable approaches can contribute to a more nuanced understanding of additional language learning, including widely held assumptions such as the ‘multilingual boost’ associated with bi- and multilingual experience and the prominent role attributed to motivation in language education. 

17:00 – 17:45 Jeanine Treffers-Daller, University of Reading: Revisiting the headedness of nominal compounds in multilinguals 

This study aims to provide new insights into the variables that affect the internal structure of nominal compounds in multilinguals. If compounding rules for two languages differ, it is complex to keep the rules separate and to create mixed compounds. While in Greek-English language contact, the head of mixed compounds is invariably Greek (Alexiadou, 2020), in Malay-English contact, the headedness varies. Thus, for example, in kampung chicken “free range chicken”, the head is English, and it appears on the right, as is common in English compounds, while the Malay modifier kampung “village” appears on the left. By contrast, in babbab research “research chapters”, the head babbab “chapters” is Malay and appears on the left, while the modifier chapters is English. Thus, here the compound follows Malay word order in that it is left-headed. This then leads to the question which variables affect the headedness of compounds in Malay-English language contact.

In this study, we ask whether mixed and unmixed compounds in Malay-English language contact follow Malay or English rules for headedness.  The current study is novel, first of all, because we analyse the effect of the language of the clause in which the compounds appear on the internal structure of compounds, while in most studies compounds are studied in isolation from their context. Second, we study the impact of mutual information (MI) scores on the headedness of Malay, English and bilingual compounds. MI scores are often used in corpus linguistics to provide information about the strength of the association between both parts of formulaic expression (Wood, 2020), but they have not so far been used in the analysis of mixed and unmixed compounds.

We analyse data from a corpus of over a thousand utterances with intrasentential code-switching, collected by Majid (2019). We found that Malay compounds in the data were generally left-headed and English compounds right-headed, but the directionality of mixed and unmixed compounds also depended  on the language of the clause in which they appear: even monolingual English compounds can be left-headed if the remainder of the clause is Malay. In addition, collocational strength plays a role: Malay compounds that have an MI score above 3 (e.g. papan hitam “blackboard”) are left-headed even when inserted in an English clause.

We finish with a discussion of the ways in which depth of language contact affects variability in compounding and how studying compounds furthers our understanding of the separability of grammars in language contact.