Languages are a means of communication, but also barriers to communication. Epistemic writing in English as a Lingua Franca (ELF) can be seen as a genre to transcend language barriers. For Japanese students it is easier to express their localized findings in Japanese, not only because of their language ability in the mother tongue (although questionable with regard to writing), but especially with reference to culturally-related concepts. Expressing these concepts in written ELF requires the ability to negotiate the meaning of such concepts: a process that is related to the central construct of "mediation" in the 2018 version of the Common European Framework for Reference (CEFR). However, not only the usage of key (cultural) concepts, but also the writing structure itself tends to follow complex patterns, which are described in the CEFR scales for writing. Using these scales not only as a means to enhance the level of students, but as description of language ability for non-native English users shows the potential of the CEFR to shape classroom language policy for learners ("learner-centered") as well as practitioners (teachers and researchers). The second order observations on the CEFR are using CEFR data (descriptors), with a different focus: not as a standard, but as a thick description of (English) language use.
However, the CEFR-scales for writing are only to a limited extent able to describe the dynamics of writing development, because they are not text-focused, but mainly user-focused. The CEFR descriptors show what abilities a specific user has in order to archive a defined task, text-production would be one task. North (2020:14) states that the CEFR-descriptors can't be complex enough to provide text types for all languages. Therefore, it is obviously not possible to teach writing with the CEFR, something we did not try to do. But is it possible to observe writing with the CEFR? This question is very difficult to answer, because the central term "writing" has to be defined. If writing is the Production of Texts, some text-focused scales could be used to provide evaluation criteria for given texts, especially the scale "Reports and Essay" (among others) would be useful is in this context. However, it wouldn't be a very deep analysis of the texts. If the production circumstances are known written texts, can be used to evaluate or assess the text production skills of the writer. The then produced texts are produced for assessment purposes. We did an assessment of learner-text (written for this purpose) for our judgement of the learner levels (between B1 and B2, not C levels) and also for presentation of our research project to the learners (see the lesson URLs of the three years for more details). But why is the basic reference to the CEFR ("CEFR-informed Second order Observations") are still in place? And here the CEFR as document is important, because the CEFR is more than it scales. Especially the concept of "domains" is important, because the CEFR gives credits to the sphere or domain, in which the language is used. Although the CEFR is not related to one domain, the reflection of language use in domains is central to our data, because the essential question for the data evaluation is: what genre do the text of our learners belong to? This question is indeed very difficult to answer, not only using the CEFR. The first answer is easy: the student did write a final thesis, but what is a "thesis"? Thompson (2016) gives some insights and is referring to large corpora, which could lead to "robust generalizations". This optimism, however, is not shared by us. Because it could be argued, that not the size of the corpus, but its diversity is crucial. A very large corpus, but only from one University is less valid for generalizations, than smaller corpora, from different settings. It is then of course difficult to combine the corpora, retrieved from total different settings and that is exact the reason, why a "thick definition" of the backgrounds is essential. Notwithstanding the allegedly ethnographic background of the notion, there are not only ethnographic methods needed (see Paltridge/Starfield 2016 for ethnographic research in English for Academic Purposes), hermeneutic methods are in times of "Digital Humanities" (see Sinclair/Rockwell 2016) very effective. Our presentation (confined to only three minutes) will hopefully be able to give an impression of the efficiency of visual tools for the purpose of complexity-mapping. Here another factor comes in: language or script for this matter. Languages might be a "pernicious myth", but if some reader (or writer) is not trained to use the give script, than words like 文章 (bunsho, text) are simply unreadable. Here Mediation (also used here) is central, the problem is of course, that in the given language (English) several other languages have to be made visible. This is not telling something about the "existence" of languages, but about the existence of different educational settings. For this research the educational setting is made visible, by the poster and by links, for instance to the teaching setting. It is indeed difficult to give a thick description of the data and their contexts, but the tools and the purpose should be mentioned. The tool is the MAXQDA software, developed for showing, "coding" in the language of the software, complexity, using different parameters. And the purpose is leading back to CEFR, the purpose is an enhancement of the framework (the CEFR), pointing to the integration to other domains in the CEFR. For example the integration of scientific domain in the CEFR would not demand the integration of several new text-types, but a more general reflection about texts, a process, which already started with the Mode of Mediation. Of course, the here presented research can only be a first step.
Link to Poster & References: http://www.linguamoodle.net/Uni/Research/Aila2021.html
