PROFESSIONAL EMBODIMENT: WALKING, RE-ENGAGEMENT OF DESK INTERACTIONS, AND PROVISION OF INSTRUCTION DURING CLASSROOM ROUNDS

Unlike continuous whole-class (plenary) interaction, independent task work involves incipient teacher-student talk as the teacher typically ‘makes rounds’ to engage in brief desk interactions with students. This article draws on multimodal conversation analysis to investigate how teacher movement during tasks offers resources for re-engaging in desk interactions and offering task-related guidance. The focus is on teachers’ walking trajectories and ways of positioning the body, and students’ orientation to them, in i) (pre-)opening moments of a desk interaction, and ii) during a subsequent instructional turn that guides students with the on-going task. The analysis shows how the pedagogical actions of checking and assessing student progress as well as making oneself available to students become observable in ways of walking, and how students display bodily whether or not they need teacher help. Movement also offers resources for shifting from individualised to collective instruction during rounds. These findings suggest that ways of navigating the body in the classroom space can index pedagogical concerns, which the students can use to make sense of the teachers’ on-going and projected engagements.


INTRODUCTION
Nearly forty years of conversation analytic (CA) research into classroom interaction have shed considerable light on how instruction takes place through talk and embodied action in educational institutions.Much of this literature revolves around the topics of turn-taking and turn-allocation, the sequential organisation of talk, and the management of authority and rule transgressions in classrooms (for an overview, see Lindwall, Lymer and Greiffenhagen 2015).These studies have shown the diversity of instructional practices and student participation across different kinds of classroom activities, such as whole-class (plenary) instruction (McHoul 1978), independent or small-group tasks (Lerner 1995;Szymanski 1999), and practical lab experiments (Ford 1999).
Recent work emphasises the constitutive role of embodiment in classroom interaction, exploring gesture, gaze, touch and the handling of material artefacts (e.g.Cekaite 2015;Jakonen 2015;Kääntä, Kasper and Piirainen-Marsh 2016;Majlesi 2015).Despite such an 'embodied turn' (Nevile 2015), the existing literature still paints a fairly sedentary picture of classroom interaction.There is a lack of systematic attention on how ways of moving and coordinating bodies in spacemobilityconfigure the interactional unfolding of instruction and learning, and offer embodied resources for constructing multimodal action Gestalts (Mondada 2014a) in classroom settings.This may reflect the fact that in many classrooms students do spend much of their time seated.However, the same hardly applies to teachers, who routinely walk as they guide students, hand out materials, present new contents at the blackboard, etc.This opens up the possibility that teacher's deployment of their entire body in the classroom environment can offer embodied resources for conducting pedagogical actionperhaps even constitute such action in itselfand exhibit professional sensemaking, judgment, and competence in situ.The present article investigates these issues by using video-recordings from content and language integrated (CLIL) classrooms to analyse teachers' movement trajectories and body positioning as they circulate the classroom, guiding individual or small-group work.Drawing on multimodal CA, the study examines teacher mobility 1) just before teacher-student talk is re-engaged at a student's desk and 2) during subsequent task-related assistance to an individual student, group, or the whole class.By showing how mobility can create opportunities for task-related guidance and assist in addressing such guidance to specific recipient(s), the article hopes to extend the current understanding of teacher embodiment beyond phenomena such as gaze and gesture to include the skilled coordination of talk with the entire mobile body.

CLASSROOM INTERACTION AND TEACHER MOBILITY
As opposed to movement that things like hand gestures, head nods, or gaze shifts involve, mobility generally refers to 'movement of people's whole bodies' (Haddington, Mondada & Nevile 2013a: 4, emphasis added) either in a direct or a 'mediated or supported' manner (e.g., when driving a car or riding a bike).Teachers' walking in the classroom is a familiar instance of direct mobility, which, however, is largely unexplored from the perspective of how it can serve the construction of professional action.To the best of my knowledge, the only interactional study with a declared focus on mobility in classroom settings is Veronesi's (2007) exploratory study, published in Italian, which analysed four university lecturer's 'space management' during monological and dialogical activities.Veronesi (2007) observed that typically teachers moved between distinct points at the front of the lecture room, and that there was a distinction between student and lecturer 'territories', sometimes demarcated with a row of empty chairs between the two parties (p.111).In the English-language literature on classroom interaction, analyses of movement have been far more implicit.A careful reading of published studies yields either incidental observations about teacher movement or investigations of talk in situations that likely involve movement, such as when teachers provide individual guidance by a student's desk.Early examples of passing observations include a study by Shultz and Florio (1979), which described how a kindergarten teacher positioned herself in relation to the 'circle area' while making announcements to children.
Similarly, Mehan (1989: 10) mentioned teachers' 'regular travels around the classroom' in a study on computerised classrooms; Lerner (1995: 127) noted how a teacher waited 'nearby' a group only to later instruct the same group further; and Markee (2005: 200) observed how the position of a teacher in the room related to students' off-task talk.Yet, (multimodal) interaction studies that would systematically analyse teacher movement as part of some specific classroom activity seem non-existing.
Despite such a lack of systematic attention, some evidence suggests that mobility offers resources for transforming, suspending, resuming, and terminating activities.For example, Macbeth's (1992) study demonstrates how movement makes the teacher's engagement with interruptions (reproaching and receiving a message) to an on-going activity visible so that students can anticipate precisely when the main activity will resume.Similarly, Kääntä et al. (2016) observed how, in the connection of a whole-class definition activity, the teacher's movement away from the blackboard toward the students transforms the activity from a visual presentation of a concept to its conversational elaboration with students.Macbeth (1992) also showed how 'walking to be seen as beginning' the class (p.143) after the bell rings allows the teacher to assemble a chattering cohort.Lastly, walking away from somebody can function as a sequence-closing action: This is the case when a student marks the teacher's one-on-one guidance as sufficient by leaving the teacher's desk or, vice versa, when the teacher leaves a student's desk and at least temporarily disengages from talk after the student's understanding claim (Koole 2010: 193-195).These and other movement trajectories within the room involve a sense of continuation of the social encounter (i.e., lesson), unlike when an emotional argument escalates to the point that a student leaves the room (see Evaldsson and Melander 2016).
Mobility is also a prerequisite to giving individualised guidance when students work on tasks alone or in groups, and the teacher circulates in the classroom, which is the focal activity in this study.Such activities are a recognisable part of classroom life and involve what has been termed as 'desk interactions' (see e.g., Tanner 2014) during which teachers typically 'make rounds' (Greiffenhagen 2012) in the room, akin to a waiter serving customers.As Tanner (2014) has shown, teacher-student desk interactions have a repetitive nature, not only because teachers routinely help different students with the same problem or question, but also because they might interact with each student recurrently during a single task.Even though Tanner's (2014) focus was not on the multimodal analysis of teacher mobility, her findings suggest that circulating in the class allows teachers both to form routines through assessing student work and to differentiate instruction based on individual students' needs.Greiffenhagen's (2012) study focused on how the teacher verbally guides the collaboration of student pairs at desktop computers during rounds.He observed that concerns of classroom management and pedagogy intertwine in such interactions, so that pupils both display 'an appearance of working' (p.36) and 'expect a reaction' (p.37) from the approaching teacher.This indicates that students may monitor and orient to teacher movement in the room, in the sense that mere approaching can project the beginning of talk.
Part of the institutional role of teachers includes the right to interact with individual students at any time, and as a student sees the teacher walking towards herself, she may anticipate that the teacher will say something with a practical purpose for the current activity.
Beyond the sequential analysis of classroom interaction, the general educational literature is familiar with the Japanese term kikan-shido, which translates as 'between desks instruction'.A study by O'Keefe, Xu and Clarke ( 2006) used a large international corpus of video-recorded mathematics lessons and post-lesson interviews of teachers and students to identify four principal functions for teachers' walking around the classroom.These include monitoring, guiding, and organising student activity, as well as engaging in social talk, all of which contain more specific sub-functions. 1 Although O' Keefe et al. (2006) did not analyse classroom interaction from a sequential perspective, their interviews showed teachers reporting using movement and body posture/positioning strategically, for example, by kneeling down in order to avoid intimidating students or appearing to 'come down' on them (p.99).
Altogether, the few interactional studies concerned with mobility in the classroom show that teacher movement can project trajectories, activities, and engagements.They suggest that places within the classroom are not only part of a fixed material environment, but also socially organised, having a 'coalescent and changable character' (Macbeth 1992: 143).What is currently missing is systematic multimodal attention to how instructional practices are configured by and fitted to the material learning environment through the skilled management of the body in space.For example, Greiffenhagen (2012) set specifically to analyse what the teacher does 'when he or she bends down to a child […] and talks to the child' (p.12), and a similar focus on talk and embodied conduct from that point onwards is visible in other studies of desk interactions (e.g.Koole 2010;Tanner 2014).While the analysis of talk is important in understanding the interactional organisation of desk interactions, it leaves unanswered the question how such instructional episodes are made possible and given some specific local sense through micro-practices (Mondada 2016) of teacher walk, such as 'lingering' nearby a student, bending over a student's desk, or walking slowly across the room.

INSTRUCTION AND INSTRUCTED ACTION
In the educational context, 'instruction' typically refers to the overarching activity that takes place in classrooms.This is different from how the term is used in ethnomethodology and conversation analysis (EM/CA), in which it is often seen as a specific directive utterance or an embodied action that forms the first part of a paired action sequence.In Garfinkel's (2002: 197-218) terms, a difference can be made between 'instructions' and 'instructed actions' (i.e., responses to instructions).While instructions and instructed actions are found in a broad range of interactional contexts, 'lessons' are one typical habitat for them, both in classrooms (e.g., St. John and Cromdal 2016) and in other, perhaps inherently more mobile settings such as dancing or driving lessons (Keevallik 2010;de Stefani and Gazin 2014;Levin et al. 2017).
In particular, this recent CA literature on instructing-on-the-move has shown the close relevance of time and space for the accomplishment of instructional turns, and how in these settings taking into account the unfolding environment is 'analytically inescapable' (Levin et al. 2017: 12).
Along these lines, in this study instruction refers to teacher turns that provide assistance, guidance or feedback to one or more students regarding the current learning task.I use the term occasioned instruction to highlight that the content of such instructions is routinely sensitive to student problems which the teacher observes during a round, rather than a pre-planned matter.Instruction thus not only emerges in and through episodes of desk interactions during a round, but it is also one main function of rounds.
In sum, the present study explores the instructional work done by teachers as they circulate in the classroom during individual or small-group tasks.The specific focus is on teacher movement and body positioning 1) in 'pre-openings' of desk interactions between the teacher and an individual student or a small group, and 2) when the teacher provides an instructional turn, directing it to an individual student, group, or the whole class.The study thus chiefly contributes to prior interactional literature on individualised instructional activities in the classroom (e.g.Greiffenhagen 2012;Tanner 2014;Majlesi 2015), and expands it with a mobility perspective.In this sense, it also differs from Veronesi's (2007) study, which analysed lecturers' movement in the lecture room during on-going whole-class talk.In contrast, re-engaging a desk interaction and providing occasioned instruction require a reorganisation of participation framework, sometimes more than once within the course of an interactional episode.As will be shown, these reorganisations have much to do with how teachers move their body in professional ways in the classroom environment.

DATA AND METHOD
This study draws on a corpus of 22 English-medium content lessons that were video-recorded using two or three cameras in secondary-level classrooms in Finland in 2003-2011.The data include classes in Physics, Biology (both Year 7) and History (Year 8), taught as optional courses to mainly native-Finnish students in small-scale content and language integrated (CLIL) programmes in two lower secondary schools.The class sizes ranged between 6 (Physics) and 19 (History) students.
The study is based on a collection of 50+ desk interactions during individual and smallgroup tasks.While episodes from all subjects were carefully analysed and showed comparable practices of teacher movement, all extracts in this article come from the History classroom.However, it should be pointed out that the material organisation of the classrooms differed so that in the History class students sat in groups of 4-5 (see Figure 1.1 for an approximate class layout), unlike in the other classes where they were seated in pairs forming rows and columns. 2   The analysis of selected interactions builds on multimodal CA studies of mobility and interaction (e.g.studies in Haddington, Mondada, and Nevile 2013b;Mortensen and Hazel 2014;Mondada 2014a) and extends them into a classroom setting, which is characterised by highly asymmetrical movement (of only one party).Transcription of embodied conduct follows Mondada's (2014b) conventions, and turns that contain students' mother tongue (Finnish) are translated into English.All names are pseudonyms.

WALKING, THE RE-ENGAGEMENT OF TEACHER-STUDENT INTERACTION, AND OCCASIONED INSTRUCTION DURING CLASSROOM ROUNDS
Rounds are a recursive activity in that the teacher goes from one desk interaction (with a student or a small group) to another, sometimes only making a brief remark such as 'good' or 'mhm', and at other times engaging in longer interactions.In the present data, walking is not uniform conduct, but instead there are observable differences in how teachers approach desks, how quickly they move from one student/group to the other, etc.This section argues that different embodied designs of walking can index distinct pedagogical actions, conveying to students a sense of whether the teacher is, for example, doing 'being available' to student initiations or checking in detail how their task work proceeds.
While rounds serve a variety of pedagogical, surveillance-related, and organisational tasks including handing out materials, checking progress, and providing extra assignments, the guidance of task work sets teachers two very practical problems.First, they need to see students' work to be able to assess how they are progressing and/or display availability for possible student initiations in a manner that minimises interruption to task work (extracts 1-3).Second, if a student needs assistance, the teacher needs to make decisions regarding whether such assistance is relevant to others (extracts 4-5).The following analysis shows how teachers observably manage these tasks through ways of moving, and how students orient to teacher movement.

Movement as a resource for monitoring task work and displaying availability
Extracts 1-2 illustrate the recursive nature of rounds and show how the teacher's way of moving into a desk interaction can index whether the teacher is conducting a quick progress check or examining students' work in greater detail and orienting to a possibility of taskrelated problems.In the extracts, groups of students are planning quiz questions as the teacher leaves her desk and begins a round.The analysis in extracts 1-2 follows her approach to and engagement with two adjacent groups of students (see Figure 1.1).Importantly, the compliment does four things: it i) displays access to students' work, i.e., that the teacher can see that they have written something on their task sheet; ii) works as an account for having interrupted the group; iii) displays availability, however brief in temporal duration, for possible queries; and iv) instructs the students by assessing their progress.
However, the students maintain their orientation on the task-at-hand, and, apart from Tuuli's quick glance up (line 6), they neither shift their gaze to the teacher during her turn nor respond to the compliment verbally.The students' non-engagement in talk with the teacher works as a display that they have no task-related queries and can competently continue their work.
Extract 2, a direct continuation of the previous situation, shows how the teacher resumes the round and initiates interaction with the next group.This time she 'announces' her approach with a question and stops in a position where she can examine the group's task sheet in greater detail and for a considerably longer time, thereby alluding that the students may have problems with the task and providing a chance to report such problems.As the teacher is retreating from the first group of students (line 6), a student in the second group (Riku) briefly glances at her direction, showing that he monitors the teacher's movement (lines 7-8).The others maintain orientation to the task sheet in front of Mauri.As the teacher walks round the desk towards Mauri, she asks whether the students have been able to begin the task (line 10).As a turn that opens interaction, a yes/no question such as this works differently from the compliment of Extract 1 in that it projects a verbal answer.Asking a question on the move also allows the teacher to 'announce' her approach and gives her time to manoeuvre into a position, right next to Mauri, from where she can see the students' task sheet upright and in detail (see fig 2 .2). Insofar as a question invites an answer, such an entry device also projects a longer engagement with the group, at least for the duration of a question-answer sequence.This temporal expectation is also evident in her seeking a place where she can take a stable position to check the group's progress.
The question treats starting the task work as an achievement, something that the students might not 'manage', and provides a possibility to report any potential problems related to the task.However, apart from Juuso's gaze shift and (possibly) Mauri's faint affirming response ('mmh') at line 11, the group does not attend to the teacher in any way.
Mauri's response is highly ambiguous as regards whether the students have indeed 'managed' to start the task, and as such an insufficient response to the question.The teacher's 'mm↑hmm' (line 12), uttered with rising intonation while walking, can thus be seen as not so much a follow-up to Mauri's response as a display of a first recognition that students have been able to do something.Such a cursory look does not yet allow for an assessment of the quality of the group's performance, and, given the ambiguity of the students' response, this is what the ensuing examination of the task sheet during the 8.3-second silence (line 13) seeks.
Ending the silence, the teacher produces an assessment (lines 14-17) of one quiz question on the students' task sheet.The assessment is designed so that its valence (or the fact that it is an assessment in the first place) is not yet recognisable when Juuso shifts his gaze back to the teacher during the intra-turn silence at line 14.The turn-so-far could still continue as a negative assessment, and it is possible that the gaze shift orients to this.That the teacher is praising the students' work is understandable against her pre-task instruction, which she repeats here nearly verbatim: students will gain points in the quiz if their questions are suitably difficult so that the answers by other groups are neither all correct nor wrong.
Thus, Mauri's latched-on agreement with the assessment (line 16), which he produces as soon as it is clear that the teacher is praising them and before the teacher's turn has come to syntactic completion, also retrospectively claims that the students had no trouble with the task in the first place, and that guidance was not needed.Once the ambiguity concerning the group's competent task performance, which was at stake from the opening seconds of the desk interaction, has been resolved, the teacher begins to walk away from the group even before she completes the assessment (line 17).By leaving, she treats the group as able to continue their task work.

Initiating interaction with a moving teacher
As extracts 1-2 showed, students can routinely and unproblematically continue task work when the teacher approaches them, to the degree that their embodied conduct of 'huddling around' the task materials and the near absence of turns-at-talk directed to the teacher seem almost to resist the invitation to engage in desk interaction.Somewhat paradoxically, such embodied behaviour and the occasional glances towards the teacher may still orient to the teacher's presence in that they construct an interactional display of doing 'working very hard' in teacher vicinity, which Greiffenhagen (2012: 36) has termed as 'an appearance of working'.As was shown, such a display can be taken by the teacher as an enactment of competence that indicates that the students need no help, and that the teacher can therefore continue the round.
At the same time as rounds are a teacher practice for monitoring and guiding task work, they afford students the opportunity to seek instruction by initiating talk with the teacher in task-related matters.In the present data, student initiations tend to occur quite soon after the teacher moves into a classroom space where she can be seen to be available. 3Securing the teacher's attention during a round requires interactional competence (Hall, Hellermann & Pekarek-Doehler, 2011; also Cekaite, 2009)   While the teacher is on the move, Tilda begins a question (lines 2-8) that contains multiple instances of self-repair and that aims to clarify whether students' earlier notes should be used in the current task (see Extract 5 for a further discussion of the local meaning of the question).Notice that the question beginning is synchronised with the teacher's disengagement with the previous group so that line 2 begins just when the teacher is taking the big step forward right by Tilda's desk.The big step projects that the teacher could walk past Tilda's desk at the very least, if not the whole group, which means that Tilda is pressed for time to produce a recognisable action if she is to secure the teacher's attention.
The teacher aligns bodily with Tilda's initiation of talk as she gradually brings her walk to a halt and repositions in a posture where she is observably doing 'listening'.This takes place through a series of small side steps (figs 3.1, 3.6-3.7),which follow the turnconstructional unit at line 2 and allow her to face Tilda's group, standing hands behind her back and resting her balance firmly on both feet.The teacher remains in this position throughout the question and the beginning of her answer (shown later in Extract 5).
Taken together, extracts 1-3 show how teacher movement and body positioning in the classroom can open up opportunities for instruction.They demonstrate how the teacher's body is deployed to accomplish the pedagogical actions of searching for and displaying visual access to students' task work and of making oneself bodily available to the students, for example, by 'leaning in' or taking a listening position.Such embodied actions, which take place in the (pre-)opening moments of desk interactions, serve the tasks of diagnosing and addressing instructional needs; together with students' orientation (or lack thereof) to the teacher, they constitute a largely embodied negotiation on whether talk by the desk is projected.When the teacher does not stop to scrutinise a student's work in detail but only quickly glances at her progress and makes herself bodily available, the responsibility to report task-related problems is increasingly attributed to the student.Thereby it is possible for the teacher to interpret students' continuation of task work and lack of orientation to the teacher as doing 'being able to carry on' with the task.Hence, she can pass by a student or a small group without saying a word, such as at the beginning of extract 3.

Movement as a resource for collectivising and synchronising students' learning processes
As the previous extracts illustrate, a routine physical location for guidance during rounds is In a situation like this, a whole-class instructional turn is occasioned by the teacher's engagement with a just-prior group, and its sense as a brief reminder is configured by the way she manages the constraints of the classroom environment and the on-going activity (round).
Beginning a whole-class turn while continuing walking to a next engagement presents the practical task of coordinating these courses of action.Alternatively, she could stop walking and suspend the round for the duration of the whole-class instruction, but accomplishing it this way makes the instruction recognisable as what it is, i.e., a brief, unintrusive reminder of an already-instructed matter.Thus, the embodied design of the turn reinforces the sense that the teacher's verbal directive ('remember') also conveys.By way of these design features, the turn does not insist on having the undivided attention of the cohort but is instead designed to minimally interrupt students' on-going task work.A careful examination of the video recordings indicates that only one out of seventeen students in the class shifts their gaze towards the teacher at any time during her turn, which suggests that few students attend to the turn.
The final extract, a continuation of extract 3, shows a different (mobile) practice of shifting from desk to collective instruction.In the extract, the teacher begins to answer Tilda's question individually by her desk but suspends her turn and, relaying Tilda's question to the class, completes the answer in whole-class interaction.By managing the question with such an audience shift within an on-going sequence, the teacher treats one student's problem as relevant and urgent to everybody in the room.A key resource in bringing about the shift is the teacher's walking away from Tilda's desk to in front of the blackboard, behind the teacher's own desk.As figure 5.1 shows, in this location the teacher is not only able to face the entire class but she can also visibly be seen to have ended the round and be unrestricted by its temporal constraints, unlike in extract 4. The extract begins as Tilda's question 'When we have these notes, do we need to write the differences?' comes to conclusion and she looks at the teacher.-------------------------- As part of the question, 'these notes' refer to an earlier task in which the students have listed 'differences' between Tudor period and modern schools, to prepare for the current task of creative writing of describing the Tudor school from a student's viewpoint.Students have now been provided a sentence beginning, 'If I was a pupil in a Tudor school…', which they are to continue.In essence, then, Tilda's question implicates a misunderstanding regarding the perspective they should adopt for the current task: that a student during the Tudor era would know about 'differences'.An orientation to this is visible in how the teacher rejects the implication (lines 9-10) by using the provided sentence to remind Tilda why students should 'actually' not write about differences.The combination of the teacher's shoulder shrug and 'so' without continuation leaves it to Tilda to interpret the significance of this somewhat implicit advice.Her response both claims ('yeah') and demonstrates (beginning to write) understanding of the advice, treating the matter resolved for all practical purposes.
At line 12, the teacher begins what could be the beginning of a directivepossibly to disambiguate the advice that 'so' only implied.However, she cuts off talk and disengages from the desk interaction by withdrawing gaze and by starting to walk between the student desks towards the blackboard.While on the move, she attempts to secure the attention of the Over lines 16-18, the teacher instructs the class by responding to the reported question.
While talking, she first walks to her desk, stops there for a while, picks up a paper, and walks back across the board to take the paper to a bin.The guidance she gives is noticeably more elaborate and explicit than the one she offered to Tilda in the desk interaction: she now emphatically prohibits the students from writing the differences ('you don't') and instead directs them to 'imagine' being pupils in a Tudor time school.This is accompanied with a gesture of lifting both hands in front of her upper torso, palms up, which she begins during the intra-turn silence preceding 'imagine' and releases on the word's onset.In one continuous movement, she drops the gesture and picks up the paper from her desk.At line 18, the teacher provides what her 'so' at lines 17 (and earlier at line 10) projected: a straight answer to Tilda's question about whether students 'need' to write the differences.While the majority of students maintain focus on their task in extract 5, relatively more students (four out of fourteen) shift their gaze to the teacher during the instruction than in extract 4.
While extracts 4 and 5 both show instances of whole-class instruction occasioned by engagements during a round, they differ in terms of the temporal projections introduced by ways of moving during the instructional turn.In extract 4, the (short) time it takes to walk from one desk interaction to the next just about suffices to remind the class of an earlier pretask instruction.In contrast, suspending the round and going to the front of the classroom to address the class can be seen to make the instructional turn somewhat more salient and visible, which can be seen as a useful strategy for tackling a new problem that has emerged during the task work (extract 5).Yet, both teacher turns have a sense of instructing the class with a relatively brief task-related remark in passing, in between other engagements, instead of insisting students to stop task work for the duration of the instruction.In both cases, the end of the instructional turn 'trails off' as the activity transforms into another activity (of checking a next group or transporting a paper to the bin).Thus, they seem designed as if to minimise interruption to student work -and based on students' gaze behaviour during both extracts, minor interruption is indeed achieved.

CONCLUDING DISCUSSION
This article has investigated walking as a resource for pedagogical action in the context of 'making rounds' (Greiffenhagen, 2012)  More specifically, the analysis has illustrated subtle variation in approaching students in (pre-)opening moments, prior to and during first verbal turns.For example, teachers might work their way into a place from which a student's/group's text is visible upright to examine in detail whether task work is proceeding as it should, even if that place is not the closest or the most obvious for talking with the student/group (extract 2).In contrast to such close monitoringwhich carries the risk of being intrusiveavailability can be bodily displayed merely by walking slowly in the vicinity of a student desk as well as by 'leaning in' over a student for a quick look at her progress (extract 1, beginning of extract 3).These and other micro-practices of walking (Mondada 2016) are not only ways to accommodate to the affordances and constraints of the material environment such as bags lying on the floor and the empty spaces between desks or working stations, but they can also convey expectations about the possibility, length, and the sense of an incipient instructional exchange.
If distinct pedagogical actions or engagements are visible in how the teacher moves or where she stands, it is possible for students to discern them by monitoring the teacher, as they do in the extracts of this article (see also Markee 2005 for how students may orient to the teacher's position in the room during off-task talk).From the students' perspective, seeing, hearing or sensing the teacher becoming available by moving into a favourable space can be seen to provide 'a range of undifferentiated opportunity in which to initiate action' (Heath 1986: 33).Curiously enough, in the extracts of this article, students largely continue task work as the teacher approaches them: they keep their gaze on task materials bar the occasional glance towards the teacher and/or continue group talk instead of aligning with teacher-student interaction.While such an 'appearance of working' (Greiffenhagen 2012: 36) seems to resist engagement in an incipient teacher-student interaction, it also conveys that the student needs no help, thereby constituting part of the embodied negotiation of pre-opening moments.Non-alignment with the teacher via silence and gaze aversion has potentially at least two benefits for the organisation of interaction during rounds: if the teacher does not need to ask every student how their task work is going, she can both minimise her interruption to student work and allocate more of her limited time to those who do need help.
Such an orientation involves attributing part of the responsibility of identifying task-related problems to the students and is in play when the teacher passes a student/group without saying a word (e.g. at the beginning of extract 3).As a local practice, it therefore offers the teacher an analytical means for differentiating instruction (see also Tanner 2014).
Besides working as a resource for opening desk interactions, movement and body positioning can also assist in the task of addressing an instructional turn to a specific audience.The analysis has shown that, during rounds, instructions are not only provided to individual students while standing by their desk, but that they can also be addressed to the cohort, in differing length and salience.In the light of extracts 4-5, whole-class remarks during task work are clearly vulnerable to student inattention in the form of lack of gaze orientation to the teacher.This suggests that students' focus on the task, their sometimes nearly impenetrable task work 'huddle'which on the one hand assists the teacher in diagnosing who does not need help while she circulates in the classroomcan also come at a cost when making whole-class remarks during the otherwise individualised guidance of rounds.In designing effective whole-class announcements during a round, teachers may thus need to attend to two opposite concerns: They need to construct a turn which is noticeable enough to overcome the 'inertia' in students' shift of attention from task to the teacher, but one which nevertheless does not interrupt task work disproportionately compared to the 'seriousness' of the instruction.
When teachers address the whole class to collectivise an observation that the on-going round has occasioned, they manage relations between an individual student and the cohort by coordinating and synchronising students' learning processes.This involves the interpretative activities of perceiving what kinds of student queries and problems might be relevant to others, and of deciding how to relay them as occasioned instructions to the class.These relations become observable, for example, when teachers preface whole-class instructions by verbally locating the source of the instruction to a specific student (as in extract 5) or, sometimes, when they single out individual students by glancing at them while addressing the class (see St. John & Cromdal 2014: 255-259).Greiffenhagen (2012: 32-34)  1 Greiffenhagen (2012), too, lists functions for 'rounds': ratifying student performance, reminding what students have not yet done, maintaining classroom control, making whole-class announcements, and linking the current activity to future activities such as exams.
2 While a systematic analysis of possible differences in mobility patterns across different material settings is beyond the scope of this study, the way desks are set up in the classroom may affect how easy it is to move between them.It is also possible that certain set-ups are more suggestive of the order in which a round progresses.For example, so called horseshoe formations may afford a more linear walking trajectory from one end to the other, whereas if the working stations are scattered around the class (such as in Figure 1.1), the order of the round is potentially less predetermined.Such material differences may affect how easily students can anticipate when the teacher approaches them and when it is 'their turn'.Yet other potential contextual differences might relate to opening interaction with an individual student vs. a group.Aligning a group in teacher-student talk is potentially a more complex matter as it may be engaged in talk or sustain otherwise a shared focus when the teacher approaches it.However, group members are not all the time aligned with the group activity, but they can disengage from it and direct passing glances at the approaching teacher (see e.g.extracts 1-2).
3 Students can, naturally, initiate interaction with the teacher from different positions in the classroom space and during different kinds of teacher's movement trajectories.It is possible that the configuration of bodies in space plays a role, together with on-going engagements, in differentiating whether a student initiation is seen as having been provided 'in their turn' or by 'shouting across the room'.

Figure 1
Figure 1.1.The teacher's movement trajectories and stationary positions in the classroom space during extracts 1-2 and 4.

Extract 2 .
Fig 2.1 (Teacher off-camera, left) Fig 2.2 14 TEA (hey) that might be, (0.6)^(0.2) juu ^looks at teacher--> 15 easy enough for somebody to know but difficult enough= in that students have to formulate a recognisable first action (in L2 in the present data) and time it appropriately by taking into account the teacher's position in the room, projected walking trajectory, and engagement with other students.The next extract shows how a student can seize the opportunity to ask a question from the teacher without verbally summoning her as soon as she moves into a 'favourable' space during individual task work, and how a desk interaction is set up as the teacher positions her body in a listening position to receive the student's turn-in-progress (figure 3.1).

Figure 5
Figure 5.1.The teacher's walk to the front of the room in extract 5.
in the classroom during individual or small-group tasks.A multimodal analysis has shown how teacher mobility serves the practical tasks of gauging students' task progress, displaying availability for possible help requests, and accomplishing shifts between individualised and collectivised instruction.The central claim of the article is that teachers' ways of moving around the classroom enable and configure the provision of instruction during task work, and that such walking trajectories can therefore be analysed as in situ displays of professional embodiment, orderly exhibits of the skilled work of a professional.
noted that in his study, the teacher individualised positive examples of student work and generalised negative ones, which prevented other students from determining 'which (if any) individuals are the occasion for it' (p.33).It may be possible that even when no such verbal identification is provided together with a whole-class announcement (such as extract 4 in this article), the teacher's movement trajectory and location in the room can still enable students to hear her whole-class turn as occasioned by a prior desk interaction with someone.A second way in which individual-cohort relations become a part of occasional instructions is through the pedagogical work that goes into transforming the nature of a desk interaction into a wholeclass instruction.Thus, it is part of doing 'being a teacher' to set examples by transforming things like an individualised praise for a good task answer into a reminder for how to make such answers (extract 4) or to treat a student's misunderstanding of the task as a 'good point' (extract 5), which can be addressed collectively without embarrassing the student.This article is an initial contribution to investigate the role of mobility in the professional work of teachers and in the organisation of classroom interaction.The goal has been to show how, in such large gatherings as classrooms, teachers use movement as a resource for accomplishing the complex multimodal Gestalts(Mondada 2014a) that instruction during task work involves.The most obvious finding to emerge relates to the close interconnections of mobility and social action in the classroom: teachers' ways of moving and deploying the body construct and embody pedagogical sensemaking and competence, tailored to the specific material learning environment on a moment-by-moment basis.Movement makes the teacher available to students and student groups, who have to analyse configurations of bodies in space to make sense of what on-going and projected engagements they index.A detailed investigation of teacher mobility and student orientation to it can offer insights into subtle and complex embodied actions that are relevant to participants in classrooms, such as doing 'being available' without imposing students.What is therefore of interest to future studies of mobility in the classroom is not only how walking and talking may intersect and project each other, but also how the analysis of movement can shed additional light into a range of so-far undocumented actions in familiar classroom practices.