On the Interplay between Strategic Competence and Language Competence in Lecturing through English Findings from Italy

The present contribution is part of a wider research project whose overall goal is to define the construct of language-teaching methodology interface in lecturing through English. This paper aims at exploring the interplay between the use of pragmatic strategies and non-standardness at the morphosyntax level in lectures conducted through English as the academic lingua franca in an Italian public university. In pursuing this aim, the main differences between findings from previous research carried out in Sweden and those obtained in the setting under investigation will be documented. Results indicate that pragmatic strategies are adopted quite often during lectures and lecturers show higher degrees of non-standard morphosyntax in their spoken production, compared to what happens in the Swedish context.


1
Lecturing in the International University The Bologna process (1999) 1 has pushed forward the internationalisation of Higher Education Institutions in Europe (De Wit et al. 2015), acting as "a lever for forcing change in higher education pedagogy" (Dearden 2014, 24) and contributing to the rapid growth of English Medium Instruction (EMI), 2 as a global phenomenon (Dearden 2014;Macaro 2018). Internationalisation poses a series of challenges within the academic domain, among others the necessity for lecturers to teach through a language (i.e. English) in which they are generally not as competent as they are in their own L1. In addition to this, their students as well are usually non-native speakers of English, a circumstance that adds further complexity to the overall picture. As a consequence, the issue arises of the effectiveness of the teaching action of such lecturers, who are experts in their discipline(s) but are faced with the challenge of having to teach through a language that is not their native one, a situation that may turn out to be an impediment to good lecturing (Airey 2009(Airey , 2012Vinke 1995).
In agreement with the claim according to which "the distinction between what is linguistic and what is pedagogic is a difficult one to make" (Björkman 2011b, 961), and under the dual assumption that a) lecturing requires communicating effectively, b) communication is a linguistic action (but not only), language being the first and most important tool a lecturer needs, 3 it is hypothesized that the two dimensions, language and teaching methodology, do not have precise boundaries, neat and well-defined, yet represent two sides of the same construct -i.e. the language-teaching methodology interface.

Lecturing through English. A Review of the Literature
The background for the present study is represented by three main strands of research: a. studies that have examined situations in which lecturers are mainly native-speakers of English and have international audiences: in these studies, attention is focused on the prag- EL.LE e-ISSN 2280-6792 9(3), 2020, 345-366

348
This research aims at offering some new insights into the existing literature both by providing a comparison with findings from an influential study (i.e. Björkman's) carried out in a similar setting (i.e. an international university in Sweden), and by reporting findings from a country -Italy -where, by comparison with northern and central European countries, lecturing through English is a relatively recent phenomenon (Costa, Coleman 2013;Wächter, Maiworm 2014). Moreover, the novelty of this research project consists in trying to merge two different yet complementary perspectives: on the one hand, a distinctly linguistic perspective, represented by the aim of describing and analysing the oral production of lecturers who employ English as the academic lingua franca to teach their subjects. To our knowledge, few studies on this topic have been conducted in Italy so far, one of these being that by Basso (2012) at the Venice International University. On the other hand, a pedagogic perspective, characterized by a dual interest in 1) understanding whether and how the use of vehicular English in the academic context impacts on lecturers' teaching methodology, and 2) identifying ways to improve the effectiveness of lectures carried out in English. Research conducted in Italy on these issues is, once again, relatively little and the questionnaire survey seems to be the favourite format (Francomacaro 2011;Costa 2016;Helm, Guarda 2015;Guarda, Helm 2016. Two studies, Francomacaro's (2011) and Costa's (2016), matched survey and lecture observation: while in the latter the descriptive component of lecturers' language production is secondary with respect to questionnaire data, 7 the former is of particular interest as it combines the analysis of lecture structure with that of pragmatic strategies. Following Francomacaro's conclusions, according to which it is necessary to improve the standard of the English-medium lecture […]. The improvement might start by raising awareness, among all the participants to the lecturing event, of the fundamental function that language plays in any learning process. (2011, 157; emphasis added) The overall aim of the present contribution is to explore the interplay between the use of pragmatic strategies and non-standardness at the morphosyntax level in lectures conducted through English (Björkman 2008(Björkman , 2009(Björkman , 2011b(Björkman , 2013 as a preliminary step towards the definition of the construct of language-teaching methodology interface where, in our view, awareness of language as a tool for teaching is thought to play a pivotal pedagogic role (Bier 2019 In order to identify a point of departure from where to start and in keeping with the literature that has been briefly presented so far, it may be hypothesized that the language-teaching methodology interface in academic lecturing exists and its preliminary definition could be formulated as follows: all those actions in which language -as well as being the main vehicle for transmitting information -is employed as a teaching tool to support learners in the organization and cognitive management of subject-matter content, represent evidence of the 'activation' of such interface.

Research Aims and Research Questions
In the present research project, our overall goal is to work towards a comprehensive definition of the language-teaching methodology interface, providing both evidence of its existence and a thorough explanation of what is underneath such an interface, as 'switched on' (or off) by subject lecturers teaching through English in academic contexts. As far as the present contribution is concerned, two preliminary aims are pursued. First, an exploration of lecturers' strategic competence, operationalized in terms of frequency of pragmatic strategy use, is carried out: we identified these strategies, verified whether and to what extent our research participants adopted them while teaching, and compared our results with those obtained by Björkman (2010Björkman ( , 2013. Second, an investigation of lecturers' language competence, operationalized in terms of frequency of non-standard morphosyntactic features in their spoken production, is conducted: in line with findings from the ELF literature, wherein the native-speaker norm is rejected in favour of a less rigid benchmark 8 -thus making the adoption of the CEFR of little use when wishing to assess lecturers' performance (Pilkinton-Pikho 2013) -we both ascertained the frequency of non-standard morphosyntactic features characterizing our research participants' spoken production and compared our findings with Björkman's (2013).
This paper aims at answering the following questions. Research Setting and Method The present investigation has been carried out in a public university in northern Italy, where the number of English-taught programmes is gradually increasing at all levels (BA, MA and PhD). Eight lecturers, all Italian, who were involved in English-medium teaching, were contacted personally via e-mail by the researcher and were asked whether they wished to participate in this research project on the use of English as a medium of instruction: seven out of eight agreed and gave their permission to be observed and audio recorded while lecturing, and to be interviewed. A concise description of our group of participants is provided below [tab. 1]. 9 The whole corpus of collected data comprises lectures (20 lectures, 10 27 hours and 36 minutes) and semi-structured interviews with lecturers (7 interviews, 2 hours and 53 minutes). As far as lectures are concerned, the researcher herself attended all of them as a non-obtrusive observer, audio recorded them and took notes on the non-verbal behaviour of participants (e.g. students' non-verbal backchannelling, gesturing and use of ICTs by the lecturer, etc.). For both lectures and 9 To guarantee lecturers' anonymity, their names have been substituted with identification codes.
10 Three lectures for each participant were attended. The only exception to this is lecturer FL_01 of whom two lectures, instead of three, were attended.
The analytic approach that has been adopted is characterized by the distinction between extensive and intensive, as the whole corpus has been investigated extensively whereas only a small portion of it has been analysed intensively; this is explained by the fact that the present research has been managed by a single researcher, which made it practically impossible for all the lectures recorded to be wholly transcribed. Therefore, a combination of an extensive exploration of the whole corpus and an intensive analysis of a subset of it was thought to be more manageable. However, for the purposes of the present contribution, reference will be made to the intensive analyses of transcribed lectures only (i.e. one lecture for each participant): the main figures of the corpus subset that has been considered are summarized here [tab. 2]. The corpus subset that has just been described was carefully investigated with a view to identifying types and frequencies of both the pragmatic strategies adopted by the seven lecturers and the nonstandard morphosyntactic features characterizing their oral delivery. As already mentioned, Björkman's study (2013) served as the main point of reference for the present research, as it functioned as a useful blueprint to conduct a thorough exploration of those dimensions -i.e. strategic competence and language competence -which are of paramount importance to better understand the English-medium teaching phenomenon: it was thought therefore that starting this research endeavour with an investigation of such dimensions might contribute to our overall goal of uncovering and defining the construct of the language-teaching methodology interface (Bier 2019). As for types of pragmatic strategies, in keeping with Björkman's study, 12 the following have been contemplated: comment on terms and concepts, prospective and retrospective signalling of discourse structure, signalling importance, comment on intent, labelling a 11 https://www.helsinki.fi/sites/default/files/atoms/files/elfa_transcrip-tion_guide.pdf.
12 As for pragmatic strategies, Björkman used Penz' taxonomy (2008)  speech act, comment on details and on content of task, comment on common ground, backchannelling, self-repair (language form vs. content), other-correction, repetition for emphasis (vs. caused by disfluency), repetition of other's utterances, head and pre-dislocation, tail and post-dislocation. To these, a few others have been added: enumerating, involving the audience, immediate reformulation of concept/key-term/question, use of questions and rhetorical questions, repetition of problematic item with question.
The non-standard morphosyntactic features that have been investigated are these: not marking the plural of the noun, article usage, double comparatives/superlatives (noun phrase level); subject-verb agreement, tense and aspect issues, passive voice (verb phrase level); non-standard question formulation, word order, negation (clause level); non-standard word forms, non-standard analytic comparative, non-standard plural forms/countability (morphology). The only feature that was not found in Björkman's study and has been added here is dropping subject pronouns (verb phrase level).
As far as data analysis is concerned, each lecture transcription was examined twice: first, pragmatic strategies were identified, highlighted and counted; then, the same procedure was followed for non-standard morphosyntactic features. To ensure reliability of results, 13 the following procedure was adopted: both analyses -pragmatic strategies and non-standard forms -were repeated two weeks apart, without making any reference to the results previously established. This procedure had the double advantage of 1) allowing a close check of discrepant instances, and 2) stimulating careful reflection on the same instances.

Results. Pragmatic Strategies and Non-Standard Morphosyntax in Lecturing
Prior to presenting and describing findings related to pragmatic strategy use and frequency of non-standard morphosyntactic forms, rate of speech, which is a feature that has a bearing on comprehensibility and therefore on lecturing effectiveness, will briefly be illustrated. As can be seen in the table below [tab. 3], the average rate of speech has been calculated dividing the total number of words by the total duration (in minutes) of each transcribed lecture. These results will be commented on later, in the discussion section.  138.6 * From beginning to end of lecture. ** Stretch of talk comprised between two interruptions (by students). The only exception to this is FL_01 whose lecturing flow is never interrupted and therefore the two extracts represent the opening and the closing of the lecture.
As for pragmatic strategies and non-standard morphosyntactic features, since, as has just been shown, there are differences in rate of speech among the lecturers, all figures indicating frequency in the tables below [tabs 4-5] have been normalised by 10,000 words; comparison with Björkman's findings is therefore allowed. Table 4 displays the frequency of use of each pragmatic strategy by each lecturer, and a comparison with figures from the Swedish setting. 14 Table 5 displays the frequency of each non-standard form by each lecturer. As can be seen, pre-and post-dislocation -included in Table 4 as well, given their pragmatic contribution to lecturing effectiveness -are also present, being themselves non-standard syntactic features. 15 Table 6 shows a comparison with average per hour figures obtained in the Swedish setting. 16 14 Figures from the Swedish setting are those reported in Björkman (2013, 124). The only exception to this is the figure regarding questions, which was reported elsewhere (Björkman 2010, 80). Instances of repetition because of disfluency are not regarded as pragmatic strategies: they are just identified and then removed from instances of repetition. Therefore, they are not contemplated in the summation. Another note on Björkman's figures: although the summation was obtained adding up figures normalized by 10,000 words, it refers to 4 lectures only, whereas the present study has considered 7. To make figures at least roughly comparable, a proportion was computed and Björkman's figures were multiplied by the ratio 7/4 (figures in parentheses are the original ones, referring to 4 lectures).    Results reported [tab. 4] reveal that the seven lecturers make use of a variety of pragmatic strategies to enhance their communicative effectiveness. On the whole, questions and rhetorical questions are by far lecturers' most favourite pragmatic tool to involve the audience and receive feedback, being used in the attempt to improve the chances that the message delivered be understood by students. In a similar vein, the second and third most frequent strategies are involving the audience (you) and commenting on common ground (we), a finding that seems to confirm the importance attached to referring to the audience as a discourse participant and co-constructor of meaning. A comparison with the Swedish results reveals interesting findings. To begin with, the same strategies investigated by Björkman were indeed observed in the Italian setting as well, thus allowing us to respond negatively to the scholar's question and confirm her speculation: A question that may emerge from the discussion in the present study is whether the pragmatic strategies discussed here are in EL.LE e-ISSN 2280-6792 9(3), 2020, 345-366 357 any way sui generis [Author's emphasis]. There is little reason to think that the pragmatic strategies observed in the present corpus should be absent in other types of high-stakes interaction in other academic settings where speakers strive to get the message across. (Björkman 2011b, 961) In general, however, the frequency of pragmatic strategies found in our corpus is much higher [tab. 4]: in one case only (i.e. commenting on details and on contents of task) the Swedish result is slightly higher than ours; in a couple of cases (i.e. labelling a speech act and commenting on common ground) there is only a tiny difference between the two settings; in most cases, notably in questions and rhetorical questions, self-repair (both language and content), and commenting on terms and concepts, our figures are strikingly higher than those obtained by Björkman.

Are non-standard morphosyntactic features common in lecturers' spoken production? What are the most frequent ones? Do these findings differ from those reported by the Swedish scholar?
Our findings [tab. 5] show that the non-standardness characterizing the seven lecturers' oral delivery is varied and frequent. The most numerous deviant cases are found at noun phrase level, namely in article usage and not marking plural of noun. Second, non-standardness at verb phrase level follows, with tense and aspect issues and dropping of subject pronoun. Finally, non-standard question formulation is also quite frequent, and this finding goes parallel to what was found for pragmatic strategies (cf. supra): since questions are by far the most frequent pragmatic tool used by lecturers, there is ample reason to think that a considerable degree of non-standardness could be found in questions themselves.
Our figures, compared to the Swedish ones, are substantially higher [tab. 6]. The only non-standard morphosyntactic features for which our results are slightly lower are double-comparatives/superlatives and plural forms/countability, whereas all the others are much more frequent: article usage is eight times more frequent, not marking plural of noun almost ten times, and non-standard question formulation is twenty-one times more frequent than in the Swedish setting. As far as this last result is concerned, a quick revision of Björkman's findings is in order: in her analysis of communicative effectiveness in academic settings, it was found that non-standard syntax in question formulation can be a critical factor, leading to overt disturbance and communication breakdown (Björkman 2008(Björkman , 2009(Björkman , 2013. This seems to sound warning bells for the context investigated, since numerous instances of non-standard questions were found in the lectures analysed here. EL.LE e-ISSN 2280-6792 9(3), 2020, 345-366

Is there any correspondence between lecturers' adoption of pragmatic strategies and deviation from standard morphosyntax when lecturing?
To answer this research question, a comparison between the seven lecturers is needed. Results are summarised here [graph 1].

Graph 1 Comparison between lecturers
After a careful examination of results, three situations stand out: 1) the case of lecturer FL_05; 2) the case of lecturers FL_01 and FL_02; 3) the case of all the others.
As can be seen in the table, FL_05 ranks first in all dimensions, with a considerable margin with respect to those who rank second. 17 -as for pragmatic strategies, she is the lecturer who, on the one hand, adopts them much more often than all the others; on the other hand, instances of "note that" to signal importance (36% of instances) and use of rhetorical "ok?" and "right?" (69%) abound in her oral delivery, which hints at both limited variety in communicative repertoire and improper use of these formulas as routines, resulting in their reduced pragmatic effectiveness; -as for morphosyntax, she is the lecturer who produces the highest number of non-standard forms in her oral delivery, and numerous non-standard questions are among them.
17 As for pragmatic strategies, the difference between FL_05 (first) and FL_02 (second) is 151, whereas in all the other cases the difference is much smaller. As for nonstandard forms, the difference between FL_05 (first) and FL_01 (second) is 235, whereas, again, in all the other cases the difference is much smaller.

FL_01 and FL_02 identify opposite cases:
-FL_01 is the lecturer who adopts pragmatic strategies the least and, in addition to this, instances of "note that" to signal importance (64% of instances) and use of rhetorical "ok?" and "right?" (60%) abound in her oral delivery, again indicating lack of variety and routine use; FL_01 ranks second in terms of frequency of non-standard features, not marking plural of nouns being the most frequent one; -on the opposite, FL_02 is the lecturer who ranks second for pragmatic strategy use, even if rhetorical "ok?" and "right?" (83% of instances) are quite abundant in her oral delivery; FL_02 ranks last for non-standard morphosyntax, all features being infrequent.
The other four lecturers seem to align to the pattern identified for FL_05, although in a less extreme way: the more (or less) frequent the use of pragmatic strategies, the more (or less) frequent the presence of non-standard morphosyntactic features. More specifically, -ML_01 ranks third in all dimensions. As for pragmatic strategies, it is remarkable that, particularly in this case, "ok?" and "right?" are very frequent among rhetorical questions (94% of instances), again indicating lack of variety and routine use; -ML_02 and FL_04 share a mid-position in ranking as the former ranks fourth in pragmatic strategies and fifth in non-standard forms, whereas the latter ranks fifth in pragmatic strategies and fourth in non-standard forms; -FL_03 occupies the sixth position both for pragmatic strategies and non-standard morphosyntax. As far as the former are concerned, it is noteworthy that, unlike what was pointed out in previous cases, FL_03 makes the least use of rhetorical "ok?" and "right?" (18% of instances). However, she is the lecturer who speaks the fastest.
If ratios are computed between the number of pragmatic strategies used and the number of non-standard forms produced, figures show -not surprisingly, after having examined the three situations above -that FL_02 ranks first (9.81), as the discrepancy between the former and the latter is the widest. On the opposite, FL_05 ranks last (1.55), since such discrepancy is much narrower. For the same reason, FL_01 ranks second last (1.66). Our findings, obtained in a specific academic setting (i.e. a medium-sized public university in northern Italy), indicate that the seven lecturers -although showing varying degrees of non-standard morphosyntax in their spoken production, which, in turn, indicate varying levels of language competence -not only are able to use pragmatic strategies but also adopt them quite often, compared to what EL.LE e-ISSN 2280-6792 9(3), 2020, 345-366

360
Swedish lecturers do (Björkman 2013). On the whole, our findings seem to support House's position, according to which pragmatic fluency does not necessarily require conforming to standard norms, as speakers can be pragmatically fluent in their own ways without following standard patterns (House 1996).
The results obtained heretofore -especially those regarding situations 1 and 3 discussed above -seem to be in line with findings from the Swedish context and, more generally, with what the literature states (cf. § 2), i.e. the use of pragmatic strategies 'compensates' for the possible presence of non-standard morphosyntactic features in ensuring communicative effectiveness, "form seems to follow function" (Cogo 2008, 60).
However, if we take a closer look at FL_05 (i.e. many strategies, many non-standard forms) and FL_01's (i.e. few strategies, many non-standard forms) profiles, the following questions arise: is such a 'compensation relationship' between pragmatic strategies and nonstandard morphosyntactic features always in place? Even when nonstandard forms are many (cf. FL_01, FL_05)? Even when critical nonstandard question formulation is persistent (cf. FL_05)? Or does a 'tolerance level' exist, in terms of frequency of non-standard features, beyond which the effect exerted by pragmatic strategies is limited? To this end, we shall recall the results of some studies, carried out in the Italian (Clark 2017) and international (Suviniitty 2012;Gundermann 2014) contexts, which indicate that students generally hold strict views on their lecturers' language competence in English, and often assess them negatively based on the language competence they perceive: these findings might support the hypothesis that a 'tolerance level' actually exists, beyond which lecturers ought not to venture, as they would run the risk of reduced communicative effectiveness -perceived and, most likely, effective as well. Our line of reasoning seems to be supported by Macaro's hypothesis too, according to which "EMI teachers need to have stepped over a certain threshold of proficiency before they can teach successfully" (Macaro 2018, 81).
From an educational linguist's point of view, as for practical implications deriving from the first part of our wider research, it is noted that lecturers -who teach in non-language-teaching contexts, and whose L1 is not English -need support in becoming aware of the role of language as a teaching tool. This should be done through a systematic reflection on: a. usefulness of pragmatic strategies as tools to enhance communicative effectiveness (thanks to increased comprehensibility of speech, higher transparency in lecture structure, possibility to interact with students, receive feedback and check their understanding); b. critical effects produced by non-standard formulations, paying special attention to the ways in which questions are posed.

361
The research that has been presented in this contribution is limited in two main ways. First, results obtained (all from the lecturers' side) have not been matched with the students' point of view: since lecturers' communicative effectiveness, as perceived by their students, has not been investigated, we are not able to give definite answers to the questions on the relationship between pragmatic strategy use and non-standard morphosyntax that have been touched upon above. Furthermore, rich qualitative data from field notes (collected during lecture observation) and semi-structured interviews with lecturers have not been contemplated in the analyses presented here: such data offer detailed information on lecturers' teaching experience (through English but not only) and their perceptions on English-medium teaching related issues (e.g. role of language proficiency, relationship with students) (Bier 2020). Only after a triangulation of findings will it be possible to make more accurate claims on those issues, about which we have just begun to scratch the surface.

Conclusions
If lecturers did not use pragmatic strategies and non-standard morphosyntactic features abounded in their oral delivery, given the complexity of the content provided (difficult and new for students) and given the added complexity deriving from the language used (which is neither the lecturers' nor the students' L1), lectures may be potential minefields where students do not get the opportunity to negotiate meaning and check their understanding. ELF settings are by nature challenging settings for all speakers involved, and without opportunities to negotiate meaning, there is an increased risk of disturbance in communication. (Björkman 2011b, 961) The main reason for which pragmatic strategies are used, in ELF in general, not only in the academic context, is to prevent possible communication breakdowns and increase the explicitness and comprehensibility of one's speech (Mauranen 2006(Mauranen , 2007Mauranen, Ranta 2009). In academic non-language-teaching contexts, communicative failure may be caused by the language of instruction or by the content taught or, more realistically, by a blend of both. Language-related obstacles may hide in the language of the discipline -i.e. subject-specific terminology, genre(s) and academic discourse characterizing the discipline itself -but also in the language competence of both the lecturer, who uses English as the academic lingua franca (and whose oral morphosyntax may show a degree of devia-EL.LE e-ISSN 2280-6792 9(3), 2020, 345-366 362 tion from the standard), and the students, who have different L1s and use English as the language of and for learning. Discipline-related difficulties, instead, reside in the highly specific and complex content provided, the full understanding of which requires that students develop competence in the language of the discipline as well: we can thus understand that separating content from language, and vice versa, is not an easy task, either in theory or in practice.
Based on the results presented in this contribution, we can formulate a first, tentative confirmation of our preliminary definition of the language-teaching methodology interface, evidence of its activation being found in those situations where the lecturer strives to be clear, to make him/herself understood by students, by means of the adoption of those strategies whose main objective is neither the quantity of content provided nor adherence to native-speaker English but comprehensibility on the students' part: language is thus both a vehicle for information/facts/content and a tool to offer students the scaffolding of such content, the focus being on the instrumentality of language, on the role of language in mediating the construction of knowledge, rather than in the mere transmission of it (Halliday 1978).