Session 3F

Embedded in a Systemic Functional Linguistics (SFL) understanding of education as an initiation into structures of knowledge and specific sets of activities (e.g. Lemke, 1990; Martin & Rose, 2008; Polias, 2016), both of which are fundamentally mediated by language, this study addresses one of the critical concerns that have arisen around CLIL: that is, the quality of learning academic content through the L2 and the possible effects on students' academic language competence, both in the L2 and L1. To address this issue, Dalton-Puffer's (2013) construct of cognitive discourse functions (CDFs) was adopted because it simultaneously targets cognitive operations and their verbal realizations. The study examines how the CDF of Categorize, essential for the construction of specialist knowledge (see e.g., Mohan, 1986, Lemke, 1990), is realized in CLIL classrooms. Categorize involves acts of categorizing, classifying, comparing and contrasting facts, objects, phenomena, abstract ideas and concepts. Despite the fact that CDFs are essential for academic knowledge construction at school, empirical research on how they are realized in CLIL classrooms is still in its beginnings. To operationalize this CDF, we developed a conceptual-analytical model of Categorize and its two sub-functions (classify & compare) through an exploratory, data-driven analysis of an oral learner corpus of grade 6 students' productions in L2 English and L1 Spanish on science and history topics collected in one bilingual school in Madrid, Spain. We also use Systemic Functional Linguistics (SFL, Halliday & Matthiessen, 2014) tools to examine particular grammatical and lexical choices which CLIL learners employ to realize the CDF of Categorize across the two subjects and languages. The analysis reveals that students seem to encounter difficulties, both conceptual and linguistic, when forming complete and appropriate classifications and comparisons in both languages, although to a dissimilar degree and in different aspects. The results obtained across subjects reveal clear subject-specific tendencies in how categorizing is carried out: comparing seems to be a defining figure of thought in history while classifications are predominant in science. The study contributes to a better understanding of how students in CLIL classrooms realize categorizations in the L2 and L1 and suggests how teachers can scaffold students through the academic language in order to produce classifications and comparisons that are acceptable both in terms of the language and subject-specific content.

References
Dalton-Puffer, C. (2013). A construct of cognitive discourse functions for conceptualising content-language integration in CLIL and multilingual education. European Journal of Applied Linguistics, 1(2), 216–253.
Halliday, M. A. K., & Matthiessen, C. M. I. M. (2014). An introduction to functional grammar (3rd ed.). London: Routledge.
Lemke, J. L. (1990). Talking science. Language, learning, and values. Norwood, NJ: Ablex Publishing. 
Martin, J. R., & Rose, D. (2008). Genre relations: Mapping culture. London: Equinox.
Mohan, B. (1986). Language and content. Reading, MA: Addison-Wesley Publishing Company.
Polias, J. (2016). Apprenticing students into science: Doing talking & writing scientifically. Melbourne: Lexis Education.
Trimble, L. (1985). English for science and technology: A discourse approach. Cambridge: Cambridge University Press. 

A plethora of studies has been conducted over the last forty years on the use of technologies to integrate telecollaborative exchanges into the foreign language classroom (Author a, 2017; Author b, 2017; Guth & Helm, 2010). The implementation of these exchanges worldwide entails engaging students in international communication and collaboration with partners of different cultures and in distant locations with the aim of developing both language skills and intercultural competence (Schenker, 2014). For virtual partnerships to remain effective and rewarding, affective and interpersonal issues need to be considered in participants’ virtual interaction since lack of participation and prompt response, misunderstandings and communicative incidents can affect the partners’ trust in each other thus threatening collaboration (O’Dowd, Shannon & Spector-Cohen, 2019). For this reason, the main objective of this study is to analyse how virtual partners used attitudinal language (Martin, 2000; Martin & White’s Appraisal Theory, 2005) in two unrelated bilingual (English and Spanish) telecollaborative encounters organized between university students in Madrid and New York (Study 1) and university students in Valencia and Maryland (Study 2). The main research question guiding this study was: What are the similarities or differences in the use of attitude by participants in two unrelated VEs? In order to answer this question, the participants’ virtual interaction in both studies was transcribed and tagged according to the attitudinal component in the Appraisal framework. By analyzing quantitatively (descriptive statistics, relative frequencies, frequencies per 100 words, chi-square test) and qualitatively the lexico-grammatical items (adverbs, adjectives, verbs, and nominalizations) found in the interaction, the authors examined the extent to which learners shared emotional responses (Affect), evaluated human behavior with respect to social norms (Judgment), and evaluated objects and products by reference to aesthetic principles and other systems of social value (Appreciation). Preliminary results (with an inter-rater reliability coefficient of 83.3%) indicate that, despite the differences in the design of the encounters (topics for discussion, communicative tool and tasks), participants presented similar behaviors in terms of Judgment, with higher instances of social esteem (Normality, Capacity, and Tenacity) over social Sanction (Veracity and Propriety) and in terms of Appreciation, with a more frequent use of Reaction (Impact and Quality) over Composition and Valuation. These similarities suggest that, when interacting virtually, participants from different countries tend to mirror and align with each other’s pragma-linguistic discourse patterns in order to collaborate more effectively as previous studies have suggested (Belz, 2003; Author a, 2018; Author b, 2018). In addition to these similarities, differences were observed in the use of Affect appraisals, with participants from Study 1, who interacted longer (2 months) and discussed a wider range of topics, using them more abundantly. This finding would indicate that the use of Affect by participants in telecollaborative encounters develops over time rather than being a given in this type of interaction and topic and task type also influence this variable. References Belz, J. A. (2003). Linguistic perspectives on the development of intercultural competence in telecollaboration. Language Learning & Technology, 7(2), 68–99. Guth, S. and Helm, F. (2010). Telecollaboration 2.0: Language, Literacies and Intercultural Learning in the 21st Century. Bern: Peter Lang. Martin, J. R. (2000). Beyond exchange: Appraisal systems in English. In S. Hunston & G. Thompson (Eds.), Evaluation in text (pp. 142-175). Oxford, England: Oxford University Press. Martin, J. R., & White, P. R. (2005). The Language of Evaluation: Appraisal in English. New York, NY: Palgrave. Schenker, T. (2014). The effects of virtual exchange on students’ interest in learning about culture. Foreign Language Annals, 46(2), 491-507. O’Dowd, R., Sauro, S., & Spector-Cohen, A. (2019). The Role of Pedagogical Mentoring in Virtual Exchange. Tesol Quarterly. DOI: 10.1002/tesq.543.

Second language (L2) writing processes have received growing attention from researchers for the past three decades. Yet, few studies have examined the processes involved in non-alphabetic language writing. There is also a need to broaden the research on the relationships among individual capacities, the temporal dimension of L2 writing and text quality. To help fill these gaps, this study examined the relationship between L2 proficiency and real-time writing processes, and the extent to which proficiency modulate the links between writing processes and text quality in L2 Chinese, jointly adopting Kellogg's (1996) and Rijlaarsdam and Van Den Bergh's (1996) writing models as the theoretical frameworks. Thirty-two L2 writers of Chinese performed two argumentative and two narrative writing tasks on a computer using the Pinyin input method. Their behaviours during writing (i.e. speed fluency, pausing and revision) were captured by Translog 2.0. Participants' proficiency in L2 Chinese was measured by a cloze test adapted from Test of Chinese as a Foreign Language. Text quality was determined via holistic ratings by two independent raters. To capture the real-time processes during writing, the whole session for each task was segmented into five equal stages. Linear mixed-effects regression analyses revealed that for the whole session, the increase of L2 proficiency led to greater fluency, fewer within-word pauses and more revisions above the word level. Due to the dual steps involved in typing Chinese characters, more proficient L2 writers also stopped less frequently to search for characters among homophones, while revised characters more often. The relationship of L2 proficiency to two revision measures tended to be mediated by stages of writing. L2 proficiency also was found to modulate the links between writing behaviours and text quality throughout the whole and at the different stages of the writing process. The findings will be discussed with reference to previous research on L2 writers of alphabetic languages. References: Kellogg, R. T. (1996). A Model of Working Memory in Writing. In C. M. Levy & S. Ransdell (Eds.), The science of writing: Theories, methods, individual differences, and applications (pp. 57–72). New Jersey: Lawrence Erlbaum Associates. Rijlaarsdam, G., & Van Den Bergh, G. (1996). The Dynamic of Composing—An Agenda for Research into an Interactive Compensatory Model of Writing: Many Questions, Some Answers. In C. M. Levy & S. Ransdell (Eds.), The science of writing: Theories, methods, individual differences, and applications (pp. 107–126). New Jersey: Lawrence Erlbaum Associates.

The importance of syntactic complexity in second language (L2) writing research and pedagogy has long been recognized (Ortega, 2003). Studies have shown that syntactic complexity in L2 writing may be affected by various learner-, task-, and context-related factors (e.g., Ellis & Yuan, 2004), and one relatively understudied factor is learners' first language (L1). Meanwhile, there is a growing body of literature on L1-related differences in various aspects of L2 writing (e.g., idea generation, lexical bundle). Taken together, findings from these lines of research point to the need for close scrutiny of potential L1-related differences in L2 writing syntactic complexity. To the best of our knowledge, the only study to date that has systematically examined a range of L1s with a variety of measures is Lu and Ai (2015). However, their datasets were not fully controlled for potential confounding variables, and proficiency ratings of learner groups and the syntactic complexity measures were derived from the same writing samples. 
The present study aims to replicate Lu and Ai (2015) and systematically examine L1-related differences in the syntactic complexity in English writing, using a rigorously controlled corpus, and including multiple learner groups at the same proficiency level as determined by independent measures of L2 proficiency. We sampled 870 argumentative essays on the topic of whether smoking should be banned at restaurants from the Written Essay Module (Version 2.4) of International Corpus Network of Asian Learners of English (ICNALE; Ishikawa, 2013), with 145 essays from the native speaker (NS) group and from each of the five non-native speaker (NNS) groups at CEFR B1 proficiency level (i.e., L1 Chinese, Indonesian, Japanese, Korean, and Thai). These essays were analyzed using the fourteen syntactic complexity measures provided by the L2 Syntactic Complexity Analyzer (Lu, 2010). When learners of different L1 backgrounds were treated as single group, results from independent samples t tests showed statistically significant differences between the combined NNS group and the NS group in all fourteen measures. The magnitude of the differences, as reflected by Cohen's d, ranged from small to large. This finding was different from that in Lu and Ai (2015), where only three out of the fourteen measures showed significant differences. One potential explanation is that the rigorously controlled nature of the current corpus has reduced confounding effects and thus made the true between-group differences more observable. Comparing the five L1 groups and the NS group using one-way ANOVAs, statistically significant differences were revealed in all fourteen measures. Post hoc pairwise comparisons with Dunnett's tests showed that NNS-NS differences were statistically significant in all but three cases (L1 Indonesian on MLC and CN/C, and L1 Chinese on CN/C), with varied patterns regarding the magnitude of differences. Implications of the findings for L2 writing research and pedagogy are considered, and limitations of the study are pointed out.


References
Ellis, R., & Yuan, F. (2004). The effects of planning on fluency, complexity, and accuracy in second language narrative writing. Studies in Second Language Acquisition, 26, 59–84.
Ishikawa, S. (2013). The ICNALE and sophisticated contrastive interlanguage analysis of Asian learners of English. In S. Ishikawa (Ed.), Learner corpus studies in Asia and the world, 1 (pp. 91–118). Kobe, Japan: Kobe University.
Lu, X. (2010). Automatic analysis of syntactic complexity in second language writing. International Journal of Corpus Linguistics, 15(4), 474–496.
Lu, X., & Ai, H. (2015). Syntactic complexity in college-level English writing: Differences among writers with diverse L1 backgrounds. Journal of Second Language Writing, 29, 16–27.
Ortega, L. (2003). Syntactic complexity measures and their relationship to L2 proficiency: A research synthesis of college-level L2 writing. Applied Linguistics, 24(4), 492–518.