Teacher Targets: A model for CLIL and ELF teacher education in polytechnic settings

The objective of this paper is to present an alternative model for CLIL/ELF teacher education, called the Teacher Target Model, which may help hard-science teachers undertake English-medium instruction more efficiently. Conceived as a visual educational trope for changing mentalities and practices, it derives from the needs analysis of a group of polytechnic teachers who have participated in the cycles of inservice training seminars given at the Polytechnic University of Madrid since 2009. Their notions, impressions, and video-recorded performances reveal that in general they hold a monolithic conception of classroom interactions and teaching procedures, with an ensuing impoverishment of genre and language repertoires. To bridge this gap, the model proposed here cross-weaves five discursive strands converging in university lectures, namely disciplinary discourse, (meta)discourse of the medium, embedded genres, lecture phase, and teaching style, thus fostering natural communication beyond the delivery of technical content and bringing ELF instruction closer to the CLIL dynamics.

78 language instructors to support their content colleagues in undertaking EMI, always welcomed and channelled through the Institute of Educational Sciences (Instituto de Ciencias de la Educación -ICE for short). One of those initiatives is the seminar "Preparación del docente para la enseñanza de contenidos técnicos en lengua inglesa dentro del EEES" ('Teacher-training for the EMI class within the EHEA'), which I have taught to over fifty engineering instructors (an average of 16 participants per course) throughout four successive editions between 2009 and 2012. The seminar sessions, which total 20 hours, cover a wide range of aspects within the linguistic and learning dimensions. The linguistic dimension gathers phonetic, intonation, lexicogrammatical, body language, and sociopragmatic guidelines. Among the latter, metadiscourse repertoires, politeness and conventions of academic genres, register features, and expressions for basic interpersonal communication skills (BICS) and cognitive academic language proficiency (CALP) -both terms coined by Cummins (1996) -are provided and explained. The learning dimension includes desirable classroom methodologies (e.g. interactive lectures, case studies, class discussions, team and project work), routines (e.g. question-answer sequences, preparatory readings) and tasks (e.g. realistic ones involving self-documentation, data representation and verbalization, problem solving and decision-making), together with their associated language. Special emphasis is laid on the phases and signalling of the lecture (Young 1994), reviewing the metadiscursive repertoire introduced in the linguistic dimension.
After a diagnostic performance (a reduced lecture up to 45 minutes long on an accessible concept from their disciplines) and a subsequent evaluation by peers and the teacher in a class discussion, the seminar participants watch different online videos of engineering lectures from the University of Stanford, available on YouTube (Stanford University Online). In pairs or small groups they take notes of teaching strategies (e.g. pace and tone, humour, asides, questions aimed at getting an answer, emphatic body language, interruptions allowed for students' questions and comments, read-outs simultaneous to blackboard calculations, etc.) to discuss their possible effectiveness in their own classes. Then, the language and teaching dimensions are taught gradually in each session by means of oral and written exercises, practical cases, discussions, and group evaluations of lecture exemplars from the MICASE corpus (The University of Michigan-English Language Institute) akin to their subject matters, given by speakers 79 with a similar academic status, native or non-native, to comparable audiences (undergraduate or graduate, junior, senior or mixed), and with diverse degrees of interactivity. The diagnostic evaluation of lectures mentioned above, made by students and teacher in discussion form, is normally accomplished with a rubric negotiated beforehand, which contains the parameters from the language and learning dimensions they deem most important, and will serve later to assess a second round of volunteer peer lectures. Trainees may also look at and comment on other didactic genres in the MICASE samples, like for example colloquia, seminars, study groups or laboratory demos, to incorporate them into their methodologies, either as full sessions or as embedded genres within the lecture.
The overall objective of the seminar is, in sum, to facilitate clil-ization; that is, the expansion of learning scopes beyond content in the attendees' future classes, making them more participative, cooperative, multimodal, varied, and oriented towards language and transversal skills. To that end, it tries to instil in the participants a methodological awareness as much as a metalinguistic one because, as some CLIL researchers have remarked (Fortanet-Gómez 2013, Klaasen and Räsänen 2006), general language proficiency and fluency may be indicators of being able to function in an educational context, but do not determine the qualification of staff, who must also reflect on teaching, be flexible to involve students in active learning, and select the pedagogical strategies and techniques that best fit the class's needs. Drawing on Smith and Simpson's (1995) catalogue of teaching competencies for higher education faculty members, the course pays detailed attention to presentation and communication skills, but also to interpersonal ones and indirectly to planning issues.

III. THE STUDY
The study reported here combines qualitative and quantitative methods to depict the starting point of UPM instructors before committing to EMI. There are two overarching research questions: How participative do they think their teaching is? And how do their beliefs differ from their practices? In answering these two questions a third one emerges: Are the performances analysed discursively rich and engaged with student audiences? Put another way: How much do UPM content teachers resort to genre- Lastly, the engagement items examined, those of Hyland's (2005b) taxonomy, included the use of personal pronouns referring to the audience (as a sign of interactivity), asides, directives, expressions of shared knowledge, and questions. We should not forget, though, that metadiscourse in general, good lecture structuring, certain embedded genres such as demos, stories and anecdotes, and first-person pronouns referred to the speaker (as a sign of involvement and self-disclosure) may also be taken as engagement features. Furthermore, it is important to bear in mind that the engagement conveyed by the various kinds of questions (attending to Tsui's 1995 typology) differs in purpose: to arouse interest, expectations and reflection (rhetorical questions), to work as interactive reminders or knowledge-recall devices (display questions), and to elicit unknown information from the student (referential questions). Not every teacher knows this distinction and sometimes rhetorical and display questions are erroneously interpreted as symptoms of an interactive, dialogic lesson.

III.1.2. Participants' profile
My 18  Examinations. When asked to evaluate their abilities in the English language on a scale from 1 to 5, where 5 stands for the highest level of command, they reported being more competent in the receptive skills (written and oral comprehension, with mean scores of 4.2 and 3.8 respectively) and less in the productive ones -written expression, with a mean score of 3.7, and oral expression, which ranked last with 3.1. This disparity between types of skills, to the detriment of oral communication, makes them feel insecure under the so-called 'native speaker fallacy' (Klaasen and Räsänen 2006), which may bias students' judgments over the pedagogical competence and credibility of their instructors (Maum 2002). Interestingly, such generalized insecurity coexists in tension with the unanimous belief that it is not their job to teach language, which is in line with the findings of other investigations into teachers' attitudes about EMI (Airey 2012).

III.2. Perceptions and performances
Perceptions and performances are closely intertwined. Teachers' notions of participation and autonomy, their assumed responsibilities, classroom routines and interpretations of student behaviour coalesce with institutional policies and shape teaching practices.

III.2.1. Perceptions: Linguistic over strategic concerns
A vast majority of respondents (72.2%) found their classes participative enough and for slightly less than half (44.4%) increasing students' involvement was not necessary. This conviction had been stirred up by the apparent lack of interest of senior undergraduates in extra assignments, tasks and projects, as they frequently dodge or plagiarize them to get them out of the way quickly and obtain their degree, and by the fact that junior undergraduates lack the theoretical background that would enable them to take an active part in those activities. It is somehow surprising that what informants consider  Figure 1, where the vertical axis indicates the number of teachers), accompanied by a relatively high incidence of 'autonomous learning' and little group work. 'Autonomous learning', however, is too vague a term to draw solid conclusions about student participation and for that reason its constituents have been specified in Figure 2.
In it we can observe that autonomy is not sought through before-class readings at all, and that the predominant dynamics is the laboratory session, followed by visits to companies, institutions, centres, or facilities. Because both do require the presence of some monitor or tutor, who often monopolizes talk and attention by giving instructions or even presentations and mini-lectures, the attainment of full autonomy on the students' part is questionable. In contrast, multimedia learning and project works without teacher intervention, which confront learners with content, procedural and media hurdles, are well below the other two alternatives. If we now turn to proportions (Figure 3), we may wonder about the notion of 'participation' UPM teachers really have. The compound percentages quantify the degree of intervention by students (the first figure, before the hyphen) and by teachers (the second figure, after the hyphen). 6% of informants, for example, rule their class time completely (100%) with monologic lectures, leaving students no chance to participate. Actually, teachers control their lectures to a large extent (at least 75%) in 78% of cases, and the frequency of egalitarian ratios is scant (22%). A similar pattern was exhibited by the ideal participation rates they suggested: only 33% of them showed an accurately balanced share of control by student and teacher (50-50%), while the rest displayed teacher quotas ranging from 60 to 80%, smaller in practical subjects and bigger in theoretical or descriptive ones.   Methodology is certainly implicit in difficulties such as mixed-ability classes, students' low proficiency in English and the need to slow down teaching pace, but this link goes unnoticed to UPM teachers, who in a conferencing session asserted that the acquisition of BICS and CALP skills is the learner's responsibility and did not connect slowing down the teaching pace with methodological adjustments, but exclusively with repercussions on the fulfilment of the syllabus.

III.2.2. Performances: The inertia of low-risk choices
For their diagnostic lessons, all volunteers opted for a lecture-centred model in a condensed conference presentation format (the average length of the talks was 17.4 min), presumably because it is a 'safe genre' on which to be evaluated, more expositive than interactive and with very narrow margins for negotiating expert roles, as the presenter holds absolute authority. In other words, they chose an updated variant of the 'chalk-and-talk' class (Mason 1994) -based on PowerPoint slide shows, a monological and prepared speech (Flowerdew 1994), and therefore little BICS, their most feared obstacle. The language they used was informal (Giménez Moreno 2008), that is to say, relaxed, with emphatic gestures, personalizations, verbal contractions, and simple connectors -let us remember that metadiscourse was their second worry after BICS when teaching in English and they logically stuck to what they already knew and avoided experimenting with variation. Most speakers were visualizers (nine out of ten  Figure 6, where the vertical axis represents the number of lectures) and just one delivered content without them, although he resorted to blackboard notes. None did web-based teaching. Of the nine visualizers, seven combined this style with other strategies and also became co-thinkers (embarking on a joint problem-solving task) or operators (manipulating realia brought to class or objects found in the room and used as improvised realia).
There were no storytellers inserting narratives or anecdotes and only one (the teacher who did not use PowerPoint) came close to encoding knowledge in a catch-phrase or stereotyped structure (Example 1). However, it was not reiterated all throughout the lesson or in a listen-and-repeat pattern, as true verbalizers do (Ogborn et al. 1996).
(1) Ways of evaluating wine. In wine-testing there are four or five senses (digression). The rule is very simple. It is a 'four-S' rule: the first thing is SEE. So, the second is SMELL or SNIFF, the third is SWIRL, then is SIP and the last part is SUMMARIZE. (Speaker writes four capital 'Ss' and then the terms on the blackboard). [L10] Genre-embedding was scarce and fundamentally consisted in process or procedural descriptions (Figure 7). Conversational exchanges (Example 2) took place for the most part while contextualizing the lesson, in its introductory phase, or in those lectures which, in addition to a final round of questions, the audience could spontaneously interrupt to comment on or add content and have their doubts cleared.  Discussions, case studies and stories were absent in every performance, and tasks and problem-solving, disguised as joint ventures due to their enunciation with the collective and inclusive pronoun 'we', were in the end led solely by the speaker, far from promoting autonomous learning. Structurally, the lectures were fairly complete. Introductions were made outlining the points to be touched on during the talk, in a slide devoted to that purpose (eight cases), simultaneously reading or paraphrasing them (six presenters), jotting them down on the blackboard (one lecturer), or just mentioning them (one teacher). Two speakers made use of elicitation through direct referential questions, but brainstorming, citation and quotation were untapped resources. With the exception of two lecturers, who used impersonal syntactic subjects in combination with second-person pronouns, (Example 3.a) and the first-person plural pronoun (3.b), the rest stated the contents and objective of the talk by speaking in the first-person, thus making delivery more vivid (Example 3.c). (4) a. Summarizing: we have four properties which make the software complex: the complexity of the domain, the development of the software project process, we have a discrete system that makes it harder to understand than a continuous system, and the eternal flexibility of the software can make it harder to understand. [L6] b. So, in summary: I have described the present course project, I have presented you three possibilities to implement the basic and advanced practice, and what you call 'innovative practice', where the users -sorry -the students, provide a proposal for the work -the terminal work. I also made a description of the evaluation that we want to do during the course. [L8] (5) a. And I wish I would have more time to give you more details, but this is enough for that today. [L10] b. And that's all, thank you everybody! [L6] From a pedagogical standpoint, lecture structuring and signposting, and in particular summarizing, could be taken as macro-engagement devices that facilitate the processing of information. Micro-engagement ones would then be the linguistic and discursive Second, asides (found in only two lectures) did not add or clarify information but pursued complicity and rapport through humour (Examples 7.a-c).
(7) a. I hate this blackboard! [L2] b. The game is so easy to play that even your cat can play it (slide projection of a cat near an electric guitar) [L8] c. I'm sure that most of you will be able to distinguish between white wines and red wines, which is something quite easy, especially if you have your eyes open. [L10] Third, directives did not abound and many were covert in the few endophorics referring to the visuals, which centred the audience's gaze (Example 8.a). There were four overt directives -three cognitive (Example 8.b), and one urging them to take immediate physical action with realia (Example 8.c).
(8) a. And this is a photograph from England. And this is another photograph from Japan. And another use you can see here is as fence, for separating houses.
[L7] Fourth, in all lectures there was a fluctuation in pronoun use. 'I' marked the speaker role, mostly in class outlinings (Example 9.a), 'we' appeared in summaries, supposedly joint tasks, hypotheses and common perceptions and conditions (Example 9.b), and 'you' also in endophorics, in hypotheses, and in procedural descriptions (Example 9.c).
(9) a. I will give an introduction and some brief concepts. [L5] b. Because the sensing nerves that we have are in the tongue -inside our mouth. [L10] c. Imagine that you are in this class, OK? And that you don't have any treatment at all -no absorption treatment at all, OK? [L7] Fifth, shared knowledge may be expressed in subtle ways, verbally (Example 10) or visually, as was the case of the projected comic strips.
(10) Probably you have heard of pyrolisis through Arguiñano, who is advertising some pyrolitic ovens.
To conclude this analysis of performances, metadiscourse and academic functions pervaded every lecture (Figure 8), but their repertoires were somewhat poor. (11) If we heat something in the absence of oxygen, I mean, there is no oxidizing agent in the atmosphere, then, whatever we have is not going to burn, it's not going to combust, but it's going to degrade, it's going to pyrolize, it's going to decompose. So we are not going to have combustion -we are going to have pyrolisis. [L9] Related to repertoire limitations, the building of idiolects is an added source of communication barriers. Lectures 1, 2 and 3 showed an over-recurrence of metadiscursive items with more than one function, which may easily lead to confusion: 'then' as both sequencer and inferential (Lecture 1), 'this' not followed by any noun (e.g. picture, graph, diagram, part, etc.) as an ambiguous endophoric when pointing to visuals and also as a blurry antecedent encapsulating a previous proposition (Lecture 2), or 'so' as an inferential, topic-shifter and discourse-filler (Lecture 3).

III.2.3. Discussion: When versions do not quite tally
We have seen that my informants were accurate in their perceptions of BICS, metadiscourse and CALP, which turned out to be the weak points in their performances.
Likewise, their self-reported teaching practices were found in most diagnostic lessons, but their ideas of autonomy and participative interaction, two cardinal issues in contentbased instruction, were not realistic and differed notably from the stagnated didactic strategies they conducted. Due to feelings of insecurity caused by their linguistic abilities, the participants remained anchored in a genre (the monologic lecture, modernized with visuals as a conference presentation) without being aware of the huge To break this inertial teaching, it is necessary to know that language and didactics go hand in hand, and so certain structural and signposting functions involve CALP and, in turn, the use of CALP may affect text structure: for instance, stage-labelling favours (or may be indicative of) summaries, while citations, quotations and deductive starts are useful tools for introductions and, some progressions, such as argumentation and problem-solution, call for CALP as well. Analogously, BICS has a bearing on structure through elicitation and brainstorming to introduce lessons and start them inductively, and in embedded tasks it may change pace or act as a closure strategy. In answer to my initial question, once teachers can cope with these implications, above phonetic inhibitions, they will be ready to meet the CLIL challenge with excellence. In the next section I suggest a model -an educational trope -for teacher-training, to help them relate and retain all the crucial aspects and resources that should be taken into account in participative EMI classes.

IV. OUTCOME: THE TEACHER-TARGET MODEL
Taking advantage of the fact that hard-science professionals are keen visualizers, I have designed an operative visual metaphor to integrate discursive and linguistic elements and serve as a comprehensive mnemonic. Shaped like a shooting target (Figure 9), its concentric circles represent the different strands of discourse that converge in university lectures, and its crosshairs embodies the reversible discursive control, depending on the genre and the situation, by teachers or students. In a highly interactive class, students would ideally be supposed to take the lead and control the flow of the lesson (horizontal movement/axis) with their expositions, questions, comments, and peer-work exchanges.
The teacher might interrupt that flow (vertical movement/axis) with observations, answers, questions, glosses, cautions, and any other necessary intervention.
This model goes beyond the applications of educational tropes proposed to date, which have been essentially descriptive of teachers' and learners' perceptions of their identity roles, the learning process, and the transmission of information, all of them aspects studied cognitively by Cameron (2003) and Cortazzi and Jin (1999), or of their impact didactically and linguistically, at least roughly (see Table 2). In participative lessons, the instructor does not need to plan every phase linguistically, just to provide keywords for presenting and monitoring tasks in those stages led by the students. Whatever the style of teaching, carrying out the teacher-target model entails providing instructors with a variety of functional repertoires (phase-, genre-, medium-and pedagogy-bound) that add to the specialized terminology of the discipline and offer a number of advantages, namely: 1) Increase the number of learning stimuli and foster occasions for natural communication other than the transmission of technical content, approaching EFL classes to CLIL.
2) Make teachers envisage the lecture as a hybrid genre in which text types (narrative, description, exposition and argumentation), genres (demos, stories, 3) Make teachers realize that those lecture constituents mould one another: some disciplines and lecture phases encourage certain media and genres, which do affect the way of teaching. And conversely, in a specific subject matter and at a given point of the lecture, personal teaching preferences involve choosing some genres and media over others.
4) Make teachers plan their class dynamics: the rhetoric of the matrix and embedded genres, the discourse associated with learning boosters, and the engagement strategies with their audience.
TEACHING STYLE (Ogborn at al. 1996): Metadiscourse, BICS and formal and informal registers to handle different strategies and degrees of learning autonomy, control over the lesson and involvement with the students (e.g. questioning and feedback tactics, task-based teaching, project work, peer work, etc.). (Bhatia 2012): Moves, steps, typical metadiscourse and specialized phraseology of online genres (chats, forums, blogs, mobile learning, videos, e-portfolios, e-mail tutorials, twitters, etc.), stories, anecdotes and jokes, case studies, seminars, commentaries of graphics, demos and experiments, reports, oral presentations, instructions, etc., with their associated or admissible progressions to sequence contents (inductive, deductive, chronological, spatial, contrastive, problem-solution, known-unknown, general-particular). (Hewings 2012): Registers for face-to-face and online teaching, endophoric and evidential metadiscourse for visuals, expression of perception, reporting, calculation, analysis, and action verbs linked to TICs and CMC (click, cut and paste, delete, drag, log on/off, etc.).