Lacking bridges and apprehensive tensions: The impact of emotions and contextual factors on German L3 teachers’ perceptions and grammar teaching practices

1 Teacher emotions and their interplay with context

Emotions are part of a complex web of teachers’ practical knowing (Golombek, 1998) and have, as such, a natural place in research on teachers and teaching. They are an integral part of classrooms, teaching practices and of teachers’ lives and worlds (Benesch, 2018, 2019; Her & De Costa, 2022; Sutton & Wheatley, 2003; Zembylas, 2004). As a psychological subsystem, teacher emotions influence and interact with teachers’ cognition, behavior, motivation and formation of identity (Borg, 2019; Dörnyei, 2009; Frenzel, 2014; Golombek & Doran, 2014; Oxford, 2021; Reio, 2005; Sheppard & Levy, 2019; Song, 2016; Sutton & Wheatley, 2003). Consequently, teaching and emotions are ‘deeply interrelated in complex ways, both epistemologically and constitutively’, as Zembylas (2004, p. 198) points out. However, due to the complexity of the concept and a lack of theoretical agreement across disciplines, teacher emotions have been difficult to define in the research literature (Barcelos & Aragão, 2018; Chen & Cheng, 2022; Uitto et al., 2015).

As an educational research field, teacher emotions gained attention in the 1990s, whereas a specific interest for the affective and emotional aspects of language teachers’ professional lives developed in the mid 2010s (Barcelos & Aragão, 2018; Barcelos & Ruohotie-Lyhty, 2018; Chen & Cheng, 2022; Cowie, 2011; Golombek & Doran, 2014). This development is in line with the ‘affective turn’ (Pavlenko, 2013) away from a merely rationalist viewpoint on teaching that excludes emotional aspects of labor (see Noddings, 2016; Rose, 1983). The turn reflects an increased research interest in emotional and motivational factors in foreign language teaching and learning in the past decades. A recent overview of studies on teacher emotions shows that two thirds of the published research come from 10 countries, where the list is headed by the USA, Australia and Germany (Chen & Cheng, 2022). The development of the field across subject areas and geographical contexts has shed light on the impacts of local cultures and social environments on teachers’ emotions (Chen, 2016; Jiang et al., 2021; Sheppard & Levy, 2019; Yin & Lee, 2012), not the least in the domains of language teaching in various parts of the world (Bahia et al., 2013; Barcelos & Aragão, 2018; Cowie, 2011; Nazari et al., 2023; Song, 2016).

Central to research on teacher emotions is the viewpoint on classrooms as emotional spaces and teaching as an emotional practice that, due to its relational nature, produces affective and emotional responses in teachers (Bress, 2006; Denzin, 1984; Hargreaves, 1998, 2000; Her & De Costa, 2022; Nias, 1996; Sheppard & Levy, 2019; Winograd, 2003; Yin & Lee, 2012; Zembylas, 2004) and requires the involvement and investment of the teacher’s full persona (Nias, 1999; Pérez Cavana, 2019). These responses are a result of the teachers’ interactions with students, colleagues, other stakeholders in and around the context, as well as with impersonal phenomena such as material conditions, policy and curriculum reforms and the context-specific instructional situations and demands posed upon them in the teaching of individual school subjects. Hence, research on teacher emotions report findings related to teachers’ well-being at work and (dis)satisfaction with the working conditions (e.g. Chen, 2016; Winograd, 2003), emotional strategies in dealing with curricular change (e.g. Her & De Costa, 2022), emotions in relation to the teaching of specific subjects, such as foreign languages (e.g. Pérez Cavana, 2019), emotions towards colleagues (e.g. Chen, 2016; Nias, 1999), and towards students and their performance (e.g. Cowie, 2011; Hargreaves, 2000; Nias, 1996; Zembylas, 2004). The latter is a salient feature found in Cowie’s (2011) results, where examples are provided of teachers expressing satisfaction in seeing their students make progress and becoming more confident and fluent in the language (p. 239).

Traditional explanatory frames of emotions would typically include a listing of discrete emotions with verbal category labels, often paired as dichotomies (Frenzel, 2014). More recent viewpoints instead regard emotions as dynamic processes occurring as responses to events in the social context and leading up to action, rather than as passive states within an individual (Barcelos & Ruohotie-Lyhty, 2018; Benesch, 2018, 2019; Frenzel, 2014; Jiang et al., 2021; Scherer & Moors, 2019; Sheppard & Levy, 2019; Sutton & Wheatley, 2003). From this perspective, emotions are embedded in sociocultural power-related phenomena that follow certain emotional rules and unspoken norms of the social context (Ahmed, 2014; Her & De Costa, 2022; Song, 2016; Winograd, 2003; Yin & Lee, 2012; Zembylas, 2004).

Furthermore, emotions could be analysed as operating on different levels within one and the same context, such as e.g. the micro, meso or macro ecologies of education (e.g. Bahia et al., 2013; Her & De Costa, 2022; Nazari et al., 2023). It is, for example, suggested that positive teacher emotions are mainly experienced in classrooms (i.e. the micro level), such as enjoyment with the teaching content and relations with the students, whereas negative teacher emotions typically stem from factors operating at the meso and macro levels, i.e. at the local school institutions or in the broader socio-political context of the educational system in question (Bahia et al., 2013; Cowie, 2011; Chen, 2016), such as heavy workload, perceptions of a low status of the subject and institutional and curricular demands. Other contextual factors affecting teachers’ work are their interactions with colleagues, school management and parents, to which teachers might have emotional responses that at the same type shape their professional and pedagogical perceptions and mediate their beliefs, values and discourses (see Chang, 2020; Hochschild, 1979; Zembylas, 2004). Consequently, this highlights the impact contextual factors from the local institutional context have on the shaping of teachers’ cognition, beliefs and instructional practices (Basturkmen, 2012; Borg, 1998, 2006, 2019; Johnson, 1994; Phipps & Borg, 2009; Richards et al., 2001; Sanchez & Borg, 2014).

2 Emotions and teacher cognition

Since previous research has clearly shown that ‘emotions hover like a shadow over the lives and work of teachers’ (Chen & Cheng, 2022, p. 418) due to the emotional demands (Chen, 2016; Yin & Lee, 2012) placed upon teachers from various stakeholders surrounding the school context, it is suggested that teacher emotions can exert a constraining effect on teachers’ pedagogical thinking, behavior and instructional decision-making (Sutton & Wheatley, 2003; Zembylas, 2004). It is, for example, shown that emotions have an influence on teachers’ instructional effectiveness and classroom management (Barcelos & Aragão, 2018; Frenzel, 2014), as well as on several cognitive functions, such as categorizing, problem-solving, goal setting, choice of teaching strategies, efficacy beliefs, and motivation in general (Sutton & Wheatley, 2003, pp. 337–340). This emphasizes the nature of classrooms as emotional spaces, in which teachers make instructional decisions guided not only by their cognition, but also by emotional considerations in their pursuit of professional fulfillment and a sense of satisfaction (Cowie, 2011; Sheppard & Levy, 2019).

The acknowledgement of the emotional aspects of teachers’ work and pedagogy, as well as the interplay of emotions and instructional actions (Barcelos & Aragão, 2018; Jiang et al., 2021; Zembylas, 2004), has also resulted in an updated view on the conceptualization of language teacher cognition as a research field and umbrella term for what language teachers ‘think, know and believe’ (Borg, 2006, p. 1), which in turn developed from the theoretical viewpoint of language teachers’ beliefs as a ‘unique filter through which second language teachers make instructional decisions, choose instructional materials, and select certain instructional practices over others’ (Johnson, 1994, p. 440). The greater role that affective components have gained in the field recently can be seen in the following passage, where one of the most influential names in the field now adopts a more holistic description of the research interest of language teacher cognition, namely:

Inquiry which seeks, with reference to their personal, professional, social, cultural and historical contexts, to understand teachers’ minds and emotions [emphasis added] and the role these play in the process of becoming, being and developing as a teacher. (Borg, 2019, p. 1167)

This updated view and acknowledgement of the role of teacher emotions follows the call for an affective turn in the broader SLA field (Pavlenko, 2013), and is a central theoretical point of departure in the present article.

3 Emotions in L3 (grammar) teaching

Language teaching can be seen as having an ‘emotionally and situationally sensitive nature’ (Barcelos & Aragão, 2018, p. 507), where the teachers are in a particularly vulnerable position. This viewpoint suggests that the specific context in which language teaching is situated opens up for emotional responses in teachers that could be studied from a subject-matter perspective. One example of language teachers’ vulnerable position is the fact that they have to invest their whole persona to a larger extent than teachers of other school subjects, for example by feeling defensive when students question particular features of the target language or culture (Pérez Cavana, 2019). Another example of FL teacher-specific emotions is anxiety and doubt about non-native teachers’ own language skills when facing students with a high proficiency level of the target language obtained outside of school (Song, 2016). In the Scandinavian context, although no previous studies explicitly target teacher emotions, literature on L3 teaching and teachers contains emotional keywords which can help research uncover affective aspects of L3 teachers’ professional lives. For example, two studies by Erickson et al. (2018, 2022) on L3 teachers in Sweden mentioned ‘passion’ and ‘the joy of working with youths and passing on language’ (2018, p. 11; author’s translation), as well as ‘joy from “opening doors” and seeing people grow’ (2022, p. 180). Furthermore, an inspection report by the Swedish Schools Inspectorate targeting the teaching of modern languages echoed teachers who explicitly voiced feelings of relief when certain students dropped out of their FL classes (Skolinspektionen, 2010, p. 7).

Teacher emotions as an analytical lens can be applied to grammar teaching, where Phipps and Borg (2009) noted tensions between language teachers’ stated beliefs and actual practices (see Basturkmen, 2012). In a thesis on secondary level Spanish L3 teachers’ beliefs on grammar, Nylén (2014) implicitly touched upon teacher emotions as a factor shaping FL teachers’ perceptions of grammar teaching. In a similarly implicit vein, Haukås (2016) reported how L3 teachers in Norway wished that their students would know more grammar from previous language learning. In both studies, L3 teachers expressed uneasiness about having to start from scratch with new student groups, since these typically lack the basic elementary grammar knowledge that the teachers would expect them to have. Contextual factors like these could be seen as ‘causal antecedents of emotions’ (Jiang et al., 2021, p. 3) within the instructional frames of grammar teaching in FL and L3 contexts. In response to the constraints of factors such as group sizes, curricular demands and students’ aptitude, motivational levels and expectations, teachers might resort to certain instructional approaches to grammar teaching primarily as means of maintaining control and exercising classroom management (Borg, 1998, 1999, 2006; Sanchez & Borg, 2014; Uysal & Bardakci, 2014; Watzke, 2007).

Furthermore, it has been suggested by Borg (2006, p. 124) that teachers engage in grammar teaching ‘not because they feel it enhances language acquisition but because they feel it is something their students expect’. Previous research conducted on university courses reported discrepancies between students’ and teachers’ perceptions of grammar, suggesting that teachers tend to underestimate how much their students enjoy grammar work in class (Brown, 2009; Schulz, 1996). Literature in the field also provides accounts of teachers’ enjoyment of working with grammar instruction (Johnston & Goettsch, 2000), viewpoints that grammar work enables opportunities for more creativity in teaching (Liviero, 2017) and examples where teachers had ‘positive feelings about the role grammatical metalanguage played in L2 [second language] learning’ (Borg, 1998, p. 19). However, reluctant stances towards grammar as a teaching context are also found in the field (e.g. Watson, 2015). On this note, Askland’s (2018) study with teachers of L3 Spanish reported teachers calling for caution with grammar teaching, since one could ‘kill students with grammar’ (p. 69). This type of emotion related to grammar teaching in foreign languages can be regarded as pedagogical emotion (Sheppard & Levy, 2019), where subject matter is an ‘important context for considering teachers’ emotions’ (Sutton & Wheatley, 2003, p. 343). Just like teacher cognition in general, these emotional stances towards grammar are often retraceable to the teachers’ own previous experience as language learners (see Borg, 1999).

As seen in the literature overview, emotional and affective components can be identified in practically all work that reports teachers’ own voices about their professional life. However, few studies in the field of foreign language education explicitly set out to address teachers’ pedagogical emotions as a construct in its own right. This holds particularly true for research on grammar teaching as a teaching content and, more specifically, the teaching in L3 contexts (as opposed to the teaching of English as an L2). Hence, this constitutes a gap in the field which will be addressed in the present study.

1 Participants

Since research in authentic classrooms often implies necessary choices based on feasibility rather than on purely methodological grounds (Borg, 2006, 2013), the four participants, henceforth referred to as Teachers A–D (see Table 1), were recruited through a purposive non-random convenience sampling (Mackey & Gass, 2016). An invitation to take part and recommend colleagues teaching GL3 was sent out to members of a local network for in-service L3 teachers, and the first four participants whose school managers granted them permission to participate were selected.

OPEN IN VIEWER

Table 1.

Overview of the participants.

Teacher	Gender	Subjects taught	Teaching experience	Mother language	Total lesson time spent on grammar teaching in GL3 (self-reported) (percentage)	Size of the GL3 group in year 7
Teacher A	Female	German, Swedish	21 years	Swedish	20	~ 20 students
Teacher B	Female	German (both L3 and L1), English	22 years (in Sweden and 5 in home country)	German	15–20	~ 10 studentsa
Teacher C	Female	German, English	27 years	Swedish	25	~ 20 students
Teacher D	Male	German, English (and qualified for Swedish)	27 years	Swedish	35	~ 30 students

Notes. L1 = first language. L3 = third language. ^aIt should be noted that Teacher B’s school stood out as a considerably smaller unit than the other three, hence, the smaller size of her GL3 group in year 7 compared to the groups of the other participants.

The only inclusion criteria used were an official national teaching certificate and at least five years’ experience of teaching German at the lower secondary level, since it is considered that teaching experience shapes teaching practices to a greater extent than teacher education alone does (see Borg & Burns, 2008), and that subject teachers’ pedagogical knowledge develops with years of experience (Gudmundsdottir & Shulman, 1987). Furthermore, previous research has shown that experienced language teachers, in comparison to novices, are more familiar with the professional discourse and can more easily articulate their pedagogical beliefs, which also better correspond with their pedagogical actions (e.g. Basturkmen, 2012; Basturkmen et al., 2004; Feryok, 2005).

As can be observed from Table 1, the participants share many similarities when it comes to the most elementary background variables. What stands out and could have a possible impact on the results is that Teacher B is a native speaker of German and has some experience of teaching in a German-speaking country before moving to Sweden. All four participants are experienced GL3 teachers having in total more than 20 years in the profession. Also, none of the teachers has changed workplace in the last couple of years and are, consequently, well established in their current context. All four teachers also use the same GL3 textbook, namely Alles Deutsch (Karnland et al., 2014).

2 Instruments and procedures for data collection

A first approach to finding out what constitutes teachers’ perceptions of their practice is to simply ask them what they think. For research on teacher emotions, Barcelos and Aragão (2018) also suggest adding ethnographic elements (i.e. observations), ‘since emotions are contextual and embodied’ (p. 517). Therefore, the present study uses a mix of three data types (see Table 2) that triangulate the themes found in the analysis.

OPEN IN VIEWER

Table 2.

Overview of data collection.

Data type	n	Time	Procedures and instruments
Individual interviewsa	4	2–4 weeks before the lessons (36–52 minutes).	Following interview guide (see Appendix A).b Recorded and transcribed by the researcher.
Classroom observations	12 (3 × 4 teachers)	During lessons (40–65 minutes each). Cycles of three consecutive lessons with each teacher.	Field notes document with bullet points. Document later elaborated into narrative by the researcher.
Teachers’ oral post-observational reflections	12 (3 × 4 teachers)	Immediately after each lesson (5–10 minutes).	Notes added to the field notes by the researcher during the conversation.

Notes. ^aIndividual interviews were chosen in order to let all participants elaborate their personal narratives without inhibitions. A delayed focus group interview, to which all four participants were invited, was carried out at a later stage within the research project. These data remain to be reported elsewhere. ^bThe interview guide also contained other questions about grammar teaching practices relevant to the overarching research project but not within the scope of the present article.

The first data type collected was individual semi-structured interviews. These took place 2–4 weeks before the classroom observations and were based on an interview guide with questions about the participants’ perceptions and experiences of various aspects of GL3 teaching, with a special emphasis on grammar (for a list of questions where the answers were found relevant for the findings of the present article, see Appendix A). Some of the questions in the interview guide were designed to elicit information about the instructional context and content, thereby implicitly targeting utterances that would also contain the participants’ emotional and affective responses to these. This was done using examples of emotional keywords contained in previous research (see Section II.3) that can be interpreted as implicit voicing of pedagogical emotions in mind. The interviews lasted 36–52 minutes and were recorded and transcribed verbatim, with pauses and hesitations included. Quoted passages reported in this article were translated from Swedish and are simplified for the sake of readability, which means that the initially indicated pauses and hesitations have been omitted in the examples given.

The second data type was classroom observations in a cycle of three consecutive lessons ⁵ with each participant targeting a specific grammatical item (verb-second placement). The teachers were all given the same textbook material as a point of departure and were asked to design their lessons in any preferred fashion and invited to add any other materials they believed would suit the pedagogical purpose of addressing the topic. Since the participants shared a short preliminary draft of their plan for the lesson cycle, the researcher had an idea of what would be covered in class. However, the teachers made on-the-spot modifications of these plans, which is normal and typical for authentic classroom settings (see Ulichny, 1996). From a position at the very back of the classroom, the researcher observed the lessons and simultaneously and systematically documented the unfolding events as ethnographic field notes (see Walford, 2009) in written form (for practical reasons, mainly as bullet-points). The main focus was to observe how the teachers explained the content to the students, especially the grammar concerned. Given the research interest and the conceptualization of students’ attitudes towards the subject as a contextual factor that might give rise to emotional responses (see Section II), the observer tried, to the best of their ability, to pay attention to instances of audible students’ and teachers’ reactions and interactions when grammar was discussed and dealt with during the lessons.

A third data set consists of oral post-observational reflections between the participant and researcher immediately after each lesson. In these short conversations (5–10 minutes), the participants were first asked to give their overall view on the lesson. After some of the observed lessons, depending on the time available the researcher also brought up some questions about observed actions and could ask the participant, for example, whether the action was planned or occurred spontaneously, whether the action is typical for the instructed group, and how the teacher perceived their students’ reactions to the teaching content (i.e. in case they had time to take notice of that), and whether they wish to adjust particular aspects of the cycle before the next lesson with the group. Some interesting answers were given, where several implicit references were made to the Hawthorne effect (see Brown, 1992), given that the two weeks with the research lessons constituted a break from the ordinary course and textbook, which might have influenced both teachers’ and students’ behavior in the observed situation. The answers were added to the field notes document, mainly as bullet points during the conversation while the participant was still present, and then elaborated on and clarified in running text by the researcher (without the teacher present) just before leaving the school facility for the day.

3 Procedures for data analysis

A qualitative approach with narrative features was used in the analysis, since teachers’ accounts of their work-related emotions can be regarded as life story materials (Lieblich et al., 1998) and ‘quality is represented through narratives of personal experience’ (Elliott, 2007, p. 230). Specifically, thematic analysis (Braun & Clarke, 2012) was applied as an accessible and flexible method for ‘systematically identifying, organizing, and offering insight into patterns of meaning’ (p. 57). All three data sets were analysed in the quest for both manifest and latent themes and content (Graneheim & Lundman, 2004) related to the research question.

The first phase of the analysis consisted of transcribing the interview recordings and the restructuring of the field notes from the observations and post-observational reflections into coherent narratives. In this process, some major categories appeared particularly meaningful, two of which are reported in this article: (1) contextual factors affecting the teaching, and (2) emotional and affective responses to the instructional context. The interview transcripts then underwent a meaning categorization by being broken down into smaller coding units, which were sorted into appropriate sub-categories. These were continuously re-evaluated during the coding process and ultimately rendered the analytical frame presented below.

In the final phase of the analytical process, themes on an interpretative level were determined. Following Graneheim and Lundman (2004), a theme is here understood as ‘a thread of an underlying meaning’ and ‘as an expression of the latent content of the text’ (p. 107). Hence, it appeared fruitful to depart from the codes stemming from the interview transcripts, define tentative themes and then look for content in the other two data sets in order to triangulate the suggested themes and provide them richer descriptions. Once the themes were defined, the data material was read once again in order to identify meaning units that might have been overlooked in the earlier phases of the analysis.

The analysis and interpretations were scrutinized and clarified in a deliberation with a colleague (who is both a researcher and a practicing teacher), to whom the analytical frame yielded was presented. Where there were potential inconsistencies and questions, the examples from the raw data were double-checked and some adjustments in the coding scheme were made. Furthermore, a member-check was performed, in which an early draft of this manuscript was sent out to the four participants. This drew the author’s attention to some misinterpretations and helped clarify the situated meaning of the interpreted utterances.

1 The influence of contextual factors on pedagogical perceptions and practices

The contextual factors mentioned by the teachers in this study have been categorized as either instructional settings or related to the students. As for the first sub-category, the factors brought up refer mainly to group sizes, other teachers, time and scheduling, and GL3 textbooks. In contrast to the somber picture about the conditions for many L3 teachers in Sweden (see Section I), the individual participants in this case study report an overall satisfaction with the contextual factors framing the instructional contexts at their current workplaces (although many potential fields of tension and areas for improvement are clearly pointed out).

The very small group sizes of the GL3 groups (see Table 1) are an important contextual factor that the participants relate to their feelings of satisfaction with work. All but one teacher have groups fewer than 20 students, which is a considerably smaller number than the number of students found in groups of Spanish L3 and many typical Swedish school classes in general (where 30 students are very common). Teacher A, for example, mentions the fact that both the group sizes (none of her GL3 groups has more than 20 students) and the drop-out rates are smaller in GL3 than in the other L3s at her school, which she states as the single best thing about her current position as a teacher of German. A similar reference to group sizes and their relation to the sense of enjoyment in a GL3 teacher is found in Teacher B, whose school unit stands out a bit in the sample due to its atypically small size:

So my smallest group counts only 6 students, and the biggest one has 16. And since I have worked here several years and I am the only German teacher, I can really work with the students all the way up. And it facilitates very much that I know, okay, this we have already covered, but this not so much, that I will have to review. So it is, very easy, easy yeah, that I know how much they [the students] know. Also, I have many highly motivated students, so, yeah. (Teacher B, interview) ⁶

As the teacher in the quoted passage above mentioned, presence of colleagues (or the lack of them) is also a central contextual factor important for GL3 teachers’ perceptions of their teaching conditions. This is found as a recurring theme in the analysed data. Since GL3 is a relatively small subject in Sweden (Skolverket, 2023) there is usually only one German teacher at a school. Being the single German teacher can, however, be seen as an advantageous factor that gives the teacher a sense of freedom to decide what to teach and how to organize and sequence lesson content without having to make compromises with subject colleagues when it comes to, for example, decisions regarding the introduction of a new textbook.

The impact of other subject teacher colleagues’ teaching as a contextual factor framing GL3 teaching becomes particularly evident when grammar as teaching content is considered. The interview data indicate overall that the GL3 teachers introduce syntax and the concept of the clause elements before their colleagues do it in the Swedish lessons. Teacher A also calls it a pity that relevant linguistic terminology is not introduced anymore by the Swedish first language (L1) teachers, who allegedly tend to simplify grammatical terminology when teaching. According to Teacher D, teachers of Swedish nowadays seem to teach less grammar than in the past and, as an L3 teacher, one has to invest time in ‘getting that job done’.

The fact that GL3 teachers generally teach more grammar and introduce it earlier than their colleagues of other language subjects is, however, generally not seen as a problem, the present data suggest. Teacher B, for example, reports adapting her grammar content and pacing in GL3 to the timing of the grammar taught by her Swedish L1 colleagues, framing it in positive words as a pedagogical strategy rather than a problematic limitation:

Then I think, of course, that I have a great help from our competent teachers of Swedish. And I await them, until they have discussed the perfect tense, or the clause elements, and then, then I come in. (Teacher B, interview)

Furthermore, another observed theme on this topic is the participants’ highlighting of the position of the GL3 teacher as an exclusive provider of grammatical knowledge when other language teachers focus less on grammar. Teacher C, for example, expresses a sense of pride in the fact that her students of German can come to Swedish class with a decent amount of prior knowledge when it comes to the clause elements, since they have already been acquainted to that in the GL3 course. On a similar note, Teacher D states that he does not mind that the Swedish L1 teachers cover less grammar nowadays compared to some years. Since he enjoys teaching grammar, he has nothing against doing it himself.

Another important contextual factor mentioned by the participants in the interviews is the time factor and the way the school management decides to distribute the L3 classes in the weekly schedule. Two of the teachers explicitly mentioned that the schedule policy is not driven by pedagogical but rather financial concerns, resulting in solutions that the L3 teachers are not always happy about. A common concern here is whether the L3 lessons should be distributed over the week as three shorter sessions or two longer ones. In the interviews, one of the teachers reminisces about a previous school where the management placed GL3 students from three different school years into one and the same teaching group in order to cut down on teaching staff expenses. Comparing it to this experience, the participant stresses the word ‘happy’ when explaining her satisfaction with the conditions at the current workplace.

The time factor is also related to the participants’ experiences and accounts of other contextual factors, such as textbooks, and to their perceptions of a desired way to teach GL3. For example, one of the participants finds the German textbooks available on the Swedish market too grammar-focused and the time constraints limit her intentions to select and prepare teaching content that she would rather want to use more frequently:

And then you don’t have time to develop the teaching, well then you just go. And if you have a group that is problematic, then you also get stricter and you regress. (Teacher C, interview)

Although it might demand a personal investment of time, thoughts and effort in coming up with and producing activities and materials, several of the participants refer to the freedom embedded in language teaching and the space for creativity and playfulness that the profession allows for. A clear example is found in the following quote:

I’m grateful to be a language teacher, I believe, because one can do many different things that more have the character of play. It’s the creative part of our profession, which I actually like very much. (Teacher D, interview)

As the data suggest, the teachers are very concerned with providing teaching that will find its best way to the students. In terms of word choice, they for example refer to this idea as ‘a mission’ and a need to ‘stand up for the craft’ when designing lessons ‘so that students will find it meaningful’. This suggests that the awareness of the students’ individual learner characteristics, attitudes and backgrounds are an important contextual factor that shape the frames for GL3 teaching. For example, the lack of precision in the students’ L3 production makes it tiresome for the teachers to carry out some creative activities in class, since these would require a great deal of instant feedback and corrections from their side. Consequently, several teachers hesitate about planning for some activities that they hold strong positive beliefs about. They also report prioritizing working with students’ motivation, even if this sometimes requires the teachers to carry on with activities they are personally not very fond of and that might require an additional workload.

Both Teacher B and Teacher C describe how they find it hard to help the students develop and produce more elaborated and complex sentences in German before they have developed enough for this. The teachers also reflect the belief that metalinguistic awareness (see Bolitho et al., 2003; Falk et al., 2015; Nielsen, 2021) in students is a prerequisite for successful learning of foreign language grammar and they lament when students are lacking it:

I think, there are still students who are a bit, a bit insecure [and] their metalanguage is not activated. They don’t have a feeling of where the verb is. What is the subject in a sentence? I mean, that you just must grasp. And not every student does. (Teacher C, interview)

Students’ inabilities to grasp the introduced concepts also mediate (and sometimes slow down) the intended grammar curricula of the GL3 teachers. In this respect, Teacher B reports on changes in her instructional practice compared to her first years of teaching in Sweden:

Well. Uhm. In the beginning I was not so meticulous with the clause elements. Because, I did not fully understand how tricky that is [for the students]. Because as a German speaker, you just know all that [laughs]. (Teacher B, interview)

Similarly, she nowadays introduces the German accusative and dative prepositions much earlier than she did in her first teacher years in Sweden, because she has noticed how tricky these concepts are for Swedish students at the lower secondary level. The students’ lack of basic grammatical terminology is also the reason why Teacher C mentions spending a larger share of her total lesson time on grammar teaching today compared to 10–15 years ago.

2 The influence of emotions on the participants’ perceptions and teaching practices

As shown above, several contextual factors inside and outside the classroom have impacts on the teaching practices of the four teachers in this study. Another powerful source of influence on teachers’ perceptions and practices are the emotions they experience in relation to events that occur in the teaching context or to a specific teaching content. As a result of the thematic analysis in this study, three narratives about teacher emotions concerning grammar teaching in GL3 will be laid out in the following.