Session 1A

In the past decade, several studies have shown the importance of out-of-school exposure or Extramural English in L2 English learning (Lindgren & Muñoz, 2013; Sylvèn & Sundqvist, 2012) even before the start of formal language learning (De Wilde et al., 2020). Most studies looking into young learners' L2 Extramural English learning have investigated receptive skills or receptive vocabulary learning but only little is known about these learners' productive skills. 
In this study we investigated lexical characteristics of young Flemish learners' L2 English writing at the start of formal English lessons. The participants (n = 3168) were all in the first year of secondary school and had received a maximum of 15 hours of English lessons. They all did a picture narration task. The tasks were given a holistic score (no output, pre-A1, A1, A2, above A2) and several measures of lexical complexity (diversity and sophistication), accuracy and fluency were calculated with the help of NLP tools (Kyle et al., 2018). 
The results showed large individual differences between learners' writing. Holistic scores ranged from 'no output' to 'above A2'-level. Regression analysis was used to investigate which lexical characteristics predicted learners' holistic proficiency score. The final model contained 9 variables and explained 50% of the variance. Similar to what was found in a previous study investigating young L2 English learners' writing (Verspoor et al., 2012), we found that a number of broad predictors impacted the proficiency score. These were lexical diversity (MTLD), word count, total number of spelling errors and percentage of English words used in the text. Additionally, four more fine-grained variables predicted the proficiency score: word frequency, trigram frequency, age of acquisition and imageability. The results show the added value of investigating a wide range of variables in order to shed light on the lexical factors that might predict holistic writing scores, even in beginner and pre-intermediate level L2 writing. The results further show that some variables such as frequency and age of acquisition have a different impact depending on learners' proficiency level. These findings will be discussed in detail during the presentation. The presentation will end with pedagogical implications and suggestions for future research. 


References:
-De Wilde, V., Brysbaert, M., & Eyckmans, J. (2020). Learning English through out-of-school exposure. Which levels of language proficiency are attained and which types of input are important? Bilingualism: Language and Cognition, 23(1), 171–185. 
-Kyle, K., Crossley, S., & Berger, C. (2018). The tool for the automatic analysis of lexical sophistication (TAALES): Version 2.0. Behavior Research Methods, 50(3), 1030–1046. 
-Lindgren, E., & Muñoz, C. (2013). The influence of exposure, parents, and linguistic distance on young European learners' foreign language comprehension. International Journal of Multilingualism, 10(1), 105–129. 
-Sylvén, L. K., & Sundqvist, P. (2012). Gaming as extramural English L2 learning and L2 proficiency among young learners. ReCALL, 24(03), 302–321. 
- Verspoor, M., Schmid, M. S., & Xu, X. (2012). A dynamic usage-based perspective on L2 writing. Journal of Second Language Writing, 21(3), 239–263. 

Traditionally, research on the role of emotions in L2 learning and acquisition focuses on the impact of emotion on learning conditions and motivation in guided L2 learning. After having explored the effect of negative emotions such as anxiety (cf. Horwitz et al., 1986), it is today mainly interested in positive emotions such as the pleasure of learning a L2 (see Dewaele et al., 2019 for a recent review). Recently, the direct impact of emotion on L2 learning and acquisition is being studied in experimental conditions. Miller and Godfroid have explored the effects of negative emotions on vocabulary learning (Miller et al., 2018) and positive and negative emotions on incidental learning of grammar (Miller & Godfroid, 2020). In this presentation, we focus on L2 performance and evaluate the influence of positive and negative emotion on L2 French spelling.
The effect of emotion on spelling in L1 primary school children learning to read and write has been investigated in different experimental settings. Both under dictation conditions where the negative and positive emotions are induced by the text itself (Cuisinier et al., 2010 ; Fartoukh et al., 2014 ; Tornare et al., 2016) and free writing where the emotions are induced by music (Largy et al., 2018), a negative impact of negative emotions, and to a lesser extent positive emotions, has been observed. These studies seem to confirm the Resource Allocation Model (Ellis & Moore, 1999), which postulates that emotions mobilise cognitive resources. Here the handling of emotions seems to use cognitive resources needed to control spelling.
The effect of emotion on L2 spelling is, to our knowledge, rather unexplored. In the only study we know of, Gunnarsson-Largy and Largy (2018) found a negative effect of a sad/negative emotion, induced by music, on grammatical spelling in dictated phrases (subject-verb agreement). In the present study, the effect of emotion on spelling in a free writing task was investigated. The participants were presented to a text extracted from a novel with a neutral, happy/positive or sad/negative emotional valence. The texts had been chosen from a battery of 15 texts (5 for each emotion) advised by experts (librarians). In a pre-test, 20 L2 French learners (B2 level), read the texts and indicated their emotional state, on a 7-point Likert scale, before and after reading each text. The texts generating the less emotion (neutral), the saddest emotion and the happiest emotion were chosen. In the main experimentation, the participants were asked to continue writing the story after having read the extract. 36 L2 French learners, with different L1s (18 B2 level and 18 C1-C2 level), all different from those in the text-selection pre-test, participated in the study. All participants wrote all 3 texts in a counterbalanced order. The results show and overall effect of positive and negative emotions on L2 French spelling. Furthermore, we observed a difference between the effect of positive and negative emotions, with a stronger effect for negative emotions. No interaction between L2 level, emotion and spelling was found.

Formulaic language (FL) plays a critical role in academic writing as it expresses common language functions. However, few studies have investigated how FL, particularly productive knowledge, is successfully learned. One possible approach is to assess how the use of context may strengthen components of lexical knowledge that can be derived from the context (Nation, 2013). A recent study with first language learners has suggested that a cohesive context is more effective than unconnected sentences for developing the productive knowledge of words (Chilton & Ehri, 2015). However, the effectiveness of using these types of contexts for acquiring FL is yet to be investigated. Moreover, whether the use of context is equally beneficial for learning words and FL has not been examined.
The present study addresses these gaps by comparing the effect of using two types of contexts (unconnected and cohesive) on developing learners' receptive and productive knowledge of words and formulaic sequences (FS). The objectives were: 1) to investigate if the pedagogical intervention was sufficient to increase receptive and productive knowledge of words and FS; 2) to compare the impact of unconnected and cohesive conditions on learning words and FS; 3) to identify whether words and FS were learned at similar rates. 
Fifty-six advanced second language postgraduate students participated in the study. The participants were assigned to one of three conditions: unconnected, cohesive, control. A pre-test-intervention-post-test design was adopted. Thirty-four words and 30 FS were taught in seven hourly lessons. After initially encountering the target items and their meanings, the unconnected group (n=20) practised the vocabulary in unrelated sentences depicting individual events, while the cohesive group (n=20) practised the vocabulary in sentences related by theme, which shared semantic elements. The control group (n=16) completed only the pre-tests and post-tests. The pre-tests and post-tests assessed knowledge of form recognition (multiple-choice) and form recall (gap-filling).
Data for the post-tests was analysed using a two-by-two between-groups ANCOVA to examine the effect of group on the development of words and FS, after controlling for the pre-test scores. For the form recall test, results showed a significant main effect of group and a significant interaction between group and vocabulary type. As expected, scores were higher in the experimental groups than the control group. Words were significantly easier to learn than FS in the two experimental groups, but no significant difference was found between words and FS in the control group. For the meaning recognition test, results showed significant main effects of group and vocabulary type, but no significant interaction between group and vocabulary type. Post-test scores were higher for the experimental groups than the control group with an overall advantage of words over FS. Pedagogical implications of findings will be discussed.
References

Chilton, M., & Ehri, L. (2015). Vocabulary learning: Sentence contexts linked by events in scenarios facilitate third graders' memory for verb meanings. Reading Research Quarterly, 50(4), 439-458.
Nation, P. (2013). Learning vocabulary in another language  (2nd ed.). Cambridge University Press.