1.Introduction

Ethnomethodology and Conversation Analysis (EMCA) are cognate fields of study that investigate participants’ culturally based methods for the production of identifiable, contextually-appropriate and recognizably ordered practical actions in highly diverse situations. Over the years, studies of members’ practices have been carried out by using data retrieved from many types of environments, from ordinary conversation to institutional settings (the ethnomethodological studies of work, e.g medical consultation, business meetings, laboratory work, interaction in courtrooms, media events, etc.). Among the latter, educational settings have also received some attention from researchers since the early years of development of EMCA (i.e. Anderson 1982Anderson, Digby 1982 “The Teacher as Classroom Researcher: A Modest Method for a New Opportunity.” In Doing Teaching: The practical Management of Classrooms, ed. by George C. G. Payne, and Edward C. Cuff, 130–147. London: Batsford.; Sharrock and Anderson 1982Sharrock, Wes, and Bob Anderson 1982 “Talking and Teaching: Reflective Comments on In-Classroom Activities.” In Doing Teaching: The Practical Management of Classrooms, ed. by George C. G. Payne, and Edward C. Cuff, 170–183. London: Batsford.; Cicourel 1974Cicourel, Aaron V. 1974 “Some Theoretical Issues in the Assessment of the Child’s Performance in Testing and Classroom Settings.” In Language Use and School Performance, ed. by Aaron. V. Cicourel, Kenneth. H. Jennings, and Sybillyn. H. M. Jennings, 300–349. New York: Academic Press. ; Cuff and Hustler 1982Cuff, Edward C., and Dave E. Hustler 1982 “Stories and Story-Time in an Infant Classroom: Some Features of Language in Social Interaction.” Semiotica 42 (2): 119–145.; Heap 1982Heap, James L. 1982 “The Social Organization of Reading Assessment: Reasons for Eclecticism.” In Doing Teaching: The practical Management of Classrooms, ed. by George C. G. Payne, and Edward C. Cuff, 39–59. London: Batsford.; Hustler and Payne 1983Hustler, Dave, and George C. F. Payne 1983 “Power in the Classroom.” Research in Education 28: 49–64. ; Hustler and Payne 1985Hustler, Dave, and George C. F. Payne 1985 “Ethnographic Conversational Analysis: An Approach to Classroom Talk.” In Strategies of Educational Research: Qualitative Methods, ed. by Robert G. Burgess, 265–291. London: Falmer.; McHoul 1978Mchoul, Alec W. 1978 “The Organization of Turns at Formal Talk in the Classroom.” Language in Society 7: 183–213. ; McHoul and Watson 1984Mchoul, Alec W., and Rod Watson 1984 “Two Axes for the Analysis of ‘Common Sense’ and ‘Formal’ Geographical Knowledge in Classroom Talk.” British Journal of Sociology of Education 5: 281–302. ; Mehan 1974Mehan, Hugh 1974 “Accomplishing Classroom Lessons.” In Language Use and School Performance, ed. by Aaron V. Cicourel, Kenneth H. Jennings, and Sybillyn H. M. Jennings, 76–142. New York: Academic Press. ; Payne 1976Payne, George C. F. 1976 “Making a Lesson Happen: An Ethnomethodological Analysis.” In The Process of Schooling: A Sociological Reader, ed. by Martin Hammersley, and Peter Woods, 33–40. London: Routledge and Kegan Paul.; Payne 1982Payne, George C. F. 1982 “Dealing with a Late-Comer”. In Doing Teaching: The practical Management of Classrooms, ed. by George C. G. Payne, and Edward C. Cuff, 90–103. London: Batsford.; Payne and Hustler 1980Payne, George C. F., and Dave Hustler 1980 “Teaching the Class: The Practical Management of a Cohort.” British Journal of Sociology of Education 1 (1): 49–66. ) and continue to receive attention in recent years (e.g Anderson 2011Anderson, Michael 2011 “Drama Education, Ethnomethodology, and ‘Industrious Chatter’.” In Methodological Choice and Design: Scholarship, Policy and Practice in Social and Educational Research, ed. by Lina Markauskaite, Peter Freebody, and Jude Irwin, 93–100. Dordrecht: Springer. ; Austin, Dwyer and Freebody 2003Austin, Helena, Bronwyn Dwyer, and Peter Freebody 2003 Schooling the Child: The Making of Students in Classrooms. London: Routledge.; Freebody and Freiberg 2000Freebody, Peter, and Jill Freiberg 2000 “Public and Pedagogic Morality: The Local Orders of Instructional and Regulatory Talk in Classrooms. In Local Education Order: Ethnomethodological Studies of Knowledge in Action, ed. by Stephen Hester, and David Francis, 141–162. Amsterdam, London: John Benjamins. , 2011Freebody, Peter, and Jill Freiberg 2011 “Ethnomethodological Research in Education and the Social Sciences: Studying the ‘Business, Identities, and Cultures’ of Classrooms.” In Methodological Choice and Design: Scholarship, Policy and Practice in Social and Educational Research, ed. by Lina Markauskaite, Peter Freebody, and Jude Irwin, 79–92. Dordrecht: Springer. ; Eglin 2009Eglin, Peter 2009 “What do we do Wednesday? On Beginning the Class as University Specific Work.” Canadian Review of Sociology 46 (1): 39–57. ; Gardiner and Anderson 2017Gardiner, Paul, and Michael Anderson 2017 “Structured Creative Processes in Learning Playwriting: Invoking Imaginative Pedagogies.” Cambridge Journal of Education, Online first. ; Hester and Francis 2000Hester, Stephen, and David Francis 2000 “Ethnomethodology and the Local Educational Order.” In Local Educational Order, ed. by Stephen Hester and David Francis, 1–19. Amsterdam: John Benjamins Publishing Company. ; Francis and Hester 2004Francis, David, and Stephen Hester 2004 An invitation to EM/CA: Language, society and interaction. London: Sage.; Macbeth 2000Macbeth, Douglas 2000 “Classrooms as Installations: Direct Instruction in the Early Ages.” In Local Educational Order, ed. by Stephen Hester and David Francis, 21–72. Amsterdam: John Benjamins Publishing Company. ; Rendle-Short 2006Rendle-Short, Johanna 2006 The Academic Presentation: Situated Talk in Action. Aldershot: Ashgate.; Tyagunova and Greiffenhagen 2017Tyagunova, Tanya, and Christian Greiffenhagen 2017 “Closing Seminars and Lectures: The Work that Lecturers and Students Do”. Discourse Studies 19 (3): 314–340. ).

However, second language (L2) teaching environments, especially where non-English languages are taught and where teacher and students do not share the same mother tongue (L1), still comprise a significant void in the field of praxiological studies (for some exceptions, see Kasper 2004Kasper, Gabriele 2004 “Participant Orientations in German Conversation-for-Learning.” The Modern Language Journal 88 (4): 551–567. ; Ohta 2011Ohta, Amy S. 2011 Second Language Acquisition Process in the Classroom. Mahwah, NJ: Erlbaum.). Moreover, and considering the potential variety of cultural methods people may use to interact in such settings, aspects of the sequential organization of utterances, and the relationships to which members orient their actions for and in the ongoing production of that organization, should be conjointly explored.

The extracts on which I shall focus were drawn from Portuguese language lessons (for beginner and advanced students). Nonetheless, it is central to mention that, in this study, I am not proposing an approach that conforms with the classical view of language acquisition process (i.e.: how students develop specific language abilities during a lesson or over a certain period of time). Here I am adopting an ethnomethodological indifference (Garfinkel 1967Garfinkel, Harold 1967 Studies in Ethnomethodology. Cambridge: Polity Press.; Garfinkel and Sacks 1970Garfinkel, Harold, and Harvey Sacks 1970 “On Formal Structures of Practical Action.” In Theoretical Sociology: Perspectives and Developments, ed. by John C. McKinney, and Edward A. Tiryakian, 338–366. New York: Appleton-Century-Crofts.) to focus specifically on members’ methods that show how these members are engaged in the practical work of making a language lesson happen.

For this reason, this is not a study that uses CA to focus necessarily on the efficacy of pedagogical approaches or the development of specific language abilities to inform theories and research methods in the field of Second Language Acquisition (for references to studies that have successfully demonstrated the usefulness of CA for this end,11.It is not my intention here to criticize any field of inquiry, but only to mention that my concern in this paper is to describe people’s methods used to make an L2 lesson happen, which are "practically primordial" in my analysis and free (in members’ perspectives) of any theoretical representation. Members’ methods can only be empirically found out and (as Garfinkel (1967)Garfinkel, Harold 1967 Studies in Ethnomethodology. Cambridge: Polity Press. would put) have little standing as topics of generic theorizing. see: Brouwer and Wagner 2004Brouwer, Catherine, and Johannes Wagner 2004 “Developmental Issues in Second Language Conversation”. Journal of Applied Linguistics 1 (1): 29–47. ; Hellerman and Lee 2014Hellermann, John, and Yo-An Lee 2014 “Members and their Competencies: Contributions of Ethnomethodological Conversation Analysis to a Multilingual Turn in Second Language Acquisition.” System 44 (1): 54–65. ; Kasper 2004Kasper, Gabriele 2004 “Participant Orientations in German Conversation-for-Learning.” The Modern Language Journal 88 (4): 551–567. ; Markee 2008Markee, Numa 2008 “Toward a Learning Behavior Tracking Methodology for CA-for-SLA.” Applied Linguistics 29 (3): 404–427. ; Mondada and Pekarek Doehler 2004Mondada, Lorenza, and Simona Pekarek Doehler 2004 “Second Language Acquisition as Situated Practice: Task Accomplished in the French Second Language Classroom.” Modern Language Journal 88 (4): 501–535. ; Ohta 2011Ohta, Amy S. 2011 Second Language Acquisition Process in the Classroom. Mahwah, NJ: Erlbaum.; Seedhouse 2004Seedhouse, P. 2004 The Interactional Architecture of the Language Classroom: A Conversation Analysis Perspective. Malden, MA: Wiley-Blackwell.; Sert 2015Sert, O. 2015 Social Interaction and L2 Classroom Discourse. Edinburgh: Edinburgh University Press.; Young and Miller 2004Young, Richard F., and Elizabeth R. Miller 2004 “Learning as Changing Participation: Discourse Roles in ESL Writing Conferences.” The Modern Language Journal 88 (4): 519–535. ) but a study that explores and describes features of the classroom interactional order as topics of analysis in their own right.

In other words, I will be looking at how classroom participants make commonplace scenes visible, how they perform their ‘taken-for-granted’ actions through the sequential organization and the categorial (or relational) orientation of their utterances in a way that any participant in that environment can identify those actions as part of an L2 class. Pedagogical efficacy and L2 competence can be elements visible in those commonplace scenes, but are neither the only nor the main features of interest. What I propose here is a study of the concrete characteristics of an interaction, which, according to Speier (1970)Speier, Matthew 1970 “The Everyday World of the Child.” In Understanding Everyday Life: Towards a Reconstruction of Sociological Knowledge, ed. by David Sudnow, 188–217. Chicago: Aldine Publishing., will be a study of the phenomena that people normally and routinely do in their everyday activities. This is, I believe, a good indication on how such phenomena are central to study language as a typifying medium and an “organized storehouse of commonsense knowledge of the society” (Watson 2016Watson, Rod 2016 “Harold Garfinkel and Pragmatics.” In Handbook of Pragmatics 20, ed. by Jan-Ola Östman, and Jef Verschueren, 1–64. John Benjamins. ), rendering the discussions I will present useful to the field of pragmatics.

2.Classroom events as phenomena of order*

In this paper my concern is to show and discuss L2 classroom events as phenomena of order* (Garfinkel 2002Garfinkel, Harold 2002 Ethnomethodology’s Program: Working out Durkheim’s Aphorism, ed. by Anne W. Rawls. Lanham, MD: Rowman & Littlefield Publishers.), pointing to the fact that the analysis of sequence (turn taking, adjacency pairs, repair, extended sequence and topic structuring) and the membership categorization work carried out by the participants inform each other (Hester and Eglin 1997Hester, Stephen, and Peter Eglin 1997 “Membership Categorization Analysis: An Introduction.” In Culture in Action: Studies in Membership Categorization Analysis, ed. by S. Hester, and P. Eglin, 1–23. Washington, D.C.: University Press of America.). The term ‘order’ here is marked with an asterisk (*) because, following Garfinkel (2002Garfinkel, Harold 2002 Ethnomethodology’s Program: Working out Durkheim’s Aphorism, ed. by Anne W. Rawls. Lanham, MD: Rowman & Littlefield Publishers., 118), order* in an ethnomethdological sense is a collector and a “proxy for any and every topic of reason, logic, meaning, proof, uniformity, generalization, universal, comparability, clarity, consistency, objectivity, objective knowledge, observation, detail, structure, and the rest” (italics in the original). Without order* humans cannot achieve practical understanding, that is, they cannot repair the natural indexicality of language to the extent that is ‘just enough’ to permit of a shared base of interaction among them. It is true that neither ordinary members nor professional analysts can fully repair indexicality on any particular occasion and that understanding can never be totally, perfectly ‘objective’. However, if we do not assume that there is some order* in communication, then everything will seem chaotic and the interaction will be ‘fuzzy’, out of alignment. There will be no sufficient interchangeability of standpoints or ‘reciprocity of perspectives’ in a social encounter and we will not be able to observe any phenomena or accounts that members rationally produce to be discussed analytically.

In the L2 classroom space, for example, even assuming that teacher and students do not normally share the same level of competence in the target language, order* is immanent, it is always there. Order* is realized through a number of local contingencies that will become visible only if we take a close look at the actions that members are performing to make a lesson event happen. Take as an example a hypothetical situation, adapted from Garfinkel (2002)Garfinkel, Harold 2002 Ethnomethodology’s Program: Working out Durkheim’s Aphorism, ed. by Anne W. Rawls. Lanham, MD: Rowman & Littlefield Publishers. in his chapter about university lectures, in which a substitute teacher on his/her first day says “you will remember from the last class that I…”. If that happened (even if this utterance were spelled out in student’s L1 – to make sure that they would understand), the students would at least look very surprised and confused. Which “I”? They would probably recall what they did, but not what “I” (the substitute teacher) did, since “I” was not present in the last class. In other words, the teacher is not ‘the regular teacher of that class’, so he could not articulate the topic of the last class this way. Order* is locally established at the moment that they are interacting to recall what happened last class, and this order* is dependent on a stock of social shared knowledge that the participants have of the last class. Consequently, the category ‘teacher’ is not a freestanding, ‘given’ category, present only because the participants are inside an educational space. It is a category associated to locally established predicates (actions) that are valid and recognizable only at that specific moment.

As it is possible to observe in our daily routines, participants do assume some stock of social shared knowledge to interact, although can only do so for ‘all practical purposes’ (Garfinkel 2002Garfinkel, Harold 2002 Ethnomethodology’s Program: Working out Durkheim’s Aphorism, ed. by Anne W. Rawls. Lanham, MD: Rowman & Littlefield Publishers.). So we here assume that there is rationality in every process of mutual and practical understanding, a concerted achievement of a specificity of sense as to permit joint actions. But this rationality and specificity of sense are not treated as ‘given’ as they are in linguistics or social sciences (for example, in theoretical conceptions of discourse formats or social theory). They are not treated as theories in EMCA, but as phenomena, which are observable by the methods people use in the local and natural process of sense making.

These phenomena of order* in EMCA studies are considered topics of inquiry in their own right, rather than analytical formats that modulate and constrain subsequent studies on the data. For this reason, every time I recall the term order*, it should be read as a collector of terms that represent many ‘topics’ of order, which will be explored in this study as locally produced, naturally accountable ‘phenomena’ of order* (Garfinkel 2002Garfinkel, Harold 2002 Ethnomethodology’s Program: Working out Durkheim’s Aphorism, ed. by Anne W. Rawls. Lanham, MD: Rowman & Littlefield Publishers.).

Still following Garfinkel (2002Garfinkel, Harold 2002 Ethnomethodology’s Program: Working out Durkheim’s Aphorism, ed. by Anne W. Rawls. Lanham, MD: Rowman & Littlefield Publishers. and also 1967Garfinkel, Harold 1967 Studies in Ethnomethodology. Cambridge: Polity Press.), the phenomena of order* that will be explored in our data are instances of local (in situ) phenomena that are jointly produced and recognized by participants as such. This means that when a phenomenon of order* is jointly established there will be a socially recognized L2 event taking place in the course of the interaction. So, for example, when a specific relational pair of membership categories is invoked, say, teacher-student, participants are also producing phenomena socially recognized as predicates of that relational pair. An example of this could be “[filling in the blanks]”. The square brackets (“[ ]”) here means two things: (i) that this L2 event is a recognized phenomenon of order* in the classroom environment; and (ii) that this phenomenon of order* will be turned into a topic of ethnomethodological inquiry to be examined in its own right.

So, for example, while students are [filling in the blanks] the teacher might be seated on his/her chair waiting for students to finish, or wandering around the class identifying students with difficulty to do the task (Mehan 1989Mehan, Hugh 1989 “Microcomputers in Classrooms: Educational Technology or Social Practice?” Anthropology and Education Quarterly 20 (1): 4–22. ). The sequence that will follow will then depend on the local contingencies of that class, which in turn depends on what the students do and the categorization that the teacher invokes for them (such as ‘students with no problem to do the task’ or ‘students with difficulties to do the task’). How the activity [filling in the blanks] will develop is something we cannot say beforehand until we look at the participants conjointly doing it.

When examining a phenomenon of order*, we will be discussing and explicating how the participants show enough competence to orient to a specific relational pair and sequential aspects of the interaction to produce that phenomenon. We will then be ‘bracketing off’ this phenomenon by examining the methods participants use to unfold the interaction and make it happen for that specific situation, as “another first time” (Garfinkel 1967Garfinkel, Harold 1967 Studies in Ethnomethodology. Cambridge: Polity Press., 9). This is not to say, however, that interactional equivalences cannot be identified between two L2 events of ‘the same nature’, but only to assume that the methods used by the participants are dynamic as that they can be constantly updated and modified, and that a close and unmotivated ‘look’ at the data is necessary to analyze and explicate these methods as ‘radical’ phenomena of ethnomethodological inquiry.

3.Members and membership

What amounts to my analysis is the description of the methods members use to make sense of each other’s actions and make a specific social encounter (in this case, a language lesson) happen. Drawing from the foundational works of Sacks (1992)Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell. and Garfinkel (1967)Garfinkel, Harold 1967 Studies in Ethnomethodology. Cambridge: Polity Press. the concept of ‘member’ or ‘participant’ for an ethnomethodologist has a particular meaning. ‘Members’ or ‘participants’ are people with a shared stock of common-sense knowledge about the world they live in and also with a shared competence to act out that knowledge in real settings. They are people whose methods are observable and analyzable (not only by the researcher, but also by the members themselves) to bring off a social event. We can say that through their actions and methods they identify themselves as parties to the transacting of a specific event. A language class, for example, does not constitute a particular social event only because it has a legal, political or institutionalized concept, but mainly because there are ‘competent course of actions’ (Garfinkel and Sacks 1970Garfinkel, Harold, and Harvey Sacks 1970 “On Formal Structures of Practical Action.” In Theoretical Sociology: Perspectives and Developments, ed. by John C. McKinney, and Edward A. Tiryakian, 338–366. New York: Appleton-Century-Crofts.) being produced by participants in that specific encounter. When a classroom participant produces a competent course of action in a given situation and the other participants recognize this action as a predicate of the ‘teacher’ or a ‘student’ of that class, then that indicates he or she is a ‘member’. Thus ‘membership’ is expressed in actions (Watson 2016Watson, Rod 2016 “Harold Garfinkel and Pragmatics.” In Handbook of Pragmatics 20, ed. by Jan-Ola Östman, and Jef Verschueren, 1–64. John Benjamins. ) and, as language is at the core of our social relations, the ‘competent mastery of the natural language’ is at the core of membership (Sacks 1992Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell.).

For this reason, it is important to clarify that by membership we do not mean ‘organized groups’ or ‘organizations’, as Sacks (1992Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell., 41) wisely observes. When talking about membership we are referring to categories that members share in common and apply to the other members in a specific situation (Payne 1976Payne, George C. F. 1976 “Making a Lesson Happen: An Ethnomethodological Analysis.” In The Process of Schooling: A Sociological Reader, ed. by Martin Hammersley, and Peter Woods, 33–40. London: Routledge and Kegan Paul.). ‘Teacher’, for example, is a category usually applied to a member during a classroom event. This category, as for any other category, is applicable because: (i) that participant (the teacher) is competent enough to produce utterances/actions that conform to the fact that he/she is conducting a class; and (ii) because the other members (students) are competent enough to react to the utterances as if these utterances were predicates of a person who is conducting a class. Membership categorization is then an interactively produced phenomenon and an analytical element to understand how talk unfolds.

Be that as it may, regarding this last point, it is necessary to make a more explicit comment. Since the outset of EMCA, scholars (such as Sacks 1972Sacks, Harvey 1972 “An Initial Investigation of the Usability of Conversational Data for Doing Sociology.” In Studies in Social Interaction, ed. by David Sudnow, 31–74. New York: Free Press.; Sacks 1992Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell.; Speier 1970Speier, Matthew 1970 “The Everyday World of the Child.” In Understanding Everyday Life: Towards a Reconstruction of Sociological Knowledge, ed. by David Sudnow, 188–217. Chicago: Aldine Publishing.; Watson 1978Watson, Rod 1978 “Categorization, Authorization and Blame-Negotiation in Conversation.” Sociology 12: 105–113. ) have shown how talk-in-interaction exhibits both a sequential order (usually observed in the reflexive sequence of turns produced by different members) and a referential orientation along the course of many types of social encounters. Consequently, there is no a priori reason to emphasize the sequential or categorial elements in the analysis of the interactions under scrutiny in this paper. My argument here is then to show that we should not cast our analysis on L2 classroom events according to one ‘take’ or the other, but contextualize the inextricable relationship of both aspects for analysis within an ethnomethodological frame of reference.

This is the main methodological claim of this study, which will bring together some materials drawn from language classroom settings.

4.Classroom talk and membership categories

Ethnomethodological studies of classroom interaction (for example, Heap 1982Heap, James L. 1982 “The Social Organization of Reading Assessment: Reasons for Eclecticism.” In Doing Teaching: The practical Management of Classrooms, ed. by George C. G. Payne, and Edward C. Cuff, 39–59. London: Batsford.; Hester 2000Hester, Stephen 2000 “The Local Order of Deviance in School: Membership Categorization, Motives and Morality in Referral Talk.” In Local Education Order: Ethnomethodological Studies of Knowledge in Action, ed. by Stephen Hester, and David Francis, 197–222. Amsterdam, London: John Benjamins. ) show us that membership categories ‘teacher’ and ‘student’ are almost always taken for granted during classroom encounters. Moreover, it is also observed that these categories are not independently deployed, but are ‘relationally paired’ (Sacks 1972Sacks, Harvey 1972 “An Initial Investigation of the Usability of Conversational Data for Doing Sociology.” In Studies in Social Interaction, ed. by David Sudnow, 31–74. New York: Free Press.) since they are a “result of an interaction between particular categories of person” (Hester 2000Hester, Stephen 2000 “The Local Order of Deviance in School: Membership Categorization, Motives and Morality in Referral Talk.” In Local Education Order: Ethnomethodological Studies of Knowledge in Action, ed. by Stephen Hester, and David Francis, 197–222. Amsterdam, London: John Benjamins. , 205). As a result, we will never have the category ‘teacher’ being invoked in isolation during a classroom event, since the category ‘student’ must also be present so that activities socially recognized as ‘classroom activities’ can make sense. For this reason, the interactional outcome of the membership categories ‘teacher-student’ forms a particular class of membership categories, called a standardized relational (or categorial) pair. This pair better depicts members’ orientations during a class instead of the simple and independent representation of separated identity descriptions of ‘teacher’ and ‘student’.

In addition, as shown by our data, relational pairs such as teacher-student, although usually taken for granted by members, may also become less salient when another relational pair emerges. In second language lessons, for example, (qualified) categorial pairs such as ‘native speaker-non-native speaker’ may have greater salience while the pair ‘teacher-student’ is momentarily moved to the background. In this case, the pair ‘native speaker-non-native speaker’ works as a new and concurrent pair that may attract members’ orientations.

Although this dynamic layering may be present as an inspectable phenomenon in many observed cases, there is no doubt that the general relational pair ‘teacher-student’ is an always-relevant pair during classroom encounters. We call it an ‘omnirelevant categorial pair’ (Fitzgerald and Housley 2015Fitzgerald, Richard, and William Housley 2015 “Introduction to Membership Categorization Analysis.” In Advances in Membership Categorization Analysis, ed. by Richard Fitzgerald, and William Housley, 1–22. London: SAGE. ; Sacks 1992Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell.). When the omni-relevant ‘teacher-student’ categorial pair is in use during a class, i.e. is an “oriented-to” category pair (Zimmerman 1998Zimmerman, Don H. 1998 “Identity, Context and Interaction” In Identities in Talk, ed. by Charles Antaki and Susan Widdicombe, 87–106. London: Sage., cited by Richards 2006Richards, Keith 2006 “Being the Teacher: Identity and Classroom Conversation.” Applied Linguistics 27 (1): 51–77. ), we will be able to see activities in that space socially recognized (by members and analysts) as ‘classroom activities’. These activities are regarded as ‘classroom activities’ not because they are carried out inside an institutional space, but because they are done by members, who show orientation to certain aspects of talk in a sequentially tied and categorial-relevant way.

This point is central to study language as an interactional medium, since analyzing classroom events as both sequentially organized and categorially inspectable phenomena is to bring members’ actions to the fore, to be analyzed as the primary and most fundamental element of enquiry. This claim will become clearer in the next section, by considering some data.

5.Categorizing membership and organizing the sequence of utterances in L2 classrooms

We will observe and discuss snippets of talk drawn from two classes (namely C1 and C2) of different courses. One refers to a beginner level course of Portuguese with 15 students (C1) and another to an advanced level course of the same language (C2) with 13 students. Only one class from each course was video-recorded. Both classes were taught in a Chinese university and taught by the same teacher. The participants gave their signed consent to have their interactions video-recorded and observed according to the general guidelines for research ethics established by the university where this study was conducted. The declaration form signed by the participants of this study is available from the author of this paper.

During the recorded sections, a camera was placed in one corner of the room (next to the door). As it was not a large class, no other recording devices (such as microphones distributed across the classroom) were used.

In the two excerpts presented, the teacher’s utterances are indexed as ‘T’, while students’ joint utterances are indexed as ‘Ss’. This may be problematic to some extent, especially because the categories ‘teacher’ and student’ inform our reading of the utterance. It is important to mention, however, that these categories are always potentially present in any class and, for this reason, they have a double reflexivity (Watson 1997Watson, Rod 1997 “Some General Reflections on ‘Categorization’ and ‘Sequence’ in the Analysis of Conversation.” In Culture in Action: Studies in Membership Categorization Analysis, ed. by Stephen Hester and Peter Eglin, 49–75. Washington, D.C.: University Press of America.), since they are not only observed by the analyst, but also invoked by the utterances the participants produce during the interaction. This means that although we are assigning category designations of teacher and students in our transcription, these categories are never pre-given to the members of the interaction, but produced locally based on participant’s orientations toward their own actions and methods.

5.1Categorizing natives and non-natives

The first extract is drawn from C1. As participants from this classroom do not share the same L1, C1 has English as the language of instruction. However, all the classroom materials and specific instructions that require minimum competence in the L2 are often designed and delivered in Portuguese. Cantonese (and, less frequently, Mandarin) is often used in parallel conversations among the students, since it is the L1 of the majority of them. Nonetheless, a moment depicted in Extract 1 below shows the use of Cantonese by the teacher, who has Portuguese as his L1 and, apparently, minimum competence in the local language. At the moment displayed in Extract 1, the students were engaged in an activity in which they had to listen to an audio recording in the target language and complete the missing parts of that recording into their textbooks.

A free translation is provided beneath each turn uttered in other language than English. As we are not doing any formal analysis of grammar, we decided not to provide a gloss for every word so that the reading of each excerpt can become more fluent. The transcription conventions are available in the Appendix.

(1)

1  T     just a moment (0.5) let me adjust the volume ((adjusts volume of device))
           (12.0)
2    T     fine ((teacher plays audio file in Portuguese; students required to fill in 
           blanks in worksheets))
3    Audio o Pedro descreve a sua nova namorada   ao seu amigo  Rui
           Pedro   describes his  new  girlfriend to his friend Rui
4    Rui   quantos anos tem ela?
           how     old  is  she?
5    Pedro vinte e dois
           twenty two
6    Rui   como é ela?
           what does she look like?
7    Pedro é    alta e magra
           she’s tall and slim
8    Rui   e   o cabelo?
           and the hair?
9    Pedro é    preto e   comprido
           it’s black and long
10   Rui   oh Pedro, que   bonita.    e   ela     não tem uma irmã?
           oh Pedro, she’s beautiful. and doesn’t she have a sister?
11   Ss    ((after audio file played, students look silently at teacher))
12         (3.5)
13   T     hóu faai oh= ((speaks in Cantonese))
           too fast
14   Ss    =HA HA HA HA HA
15   T     =hóu faai ((turns to board))
           too fast
16   Ss    =HA HA HA HA HA
17   Harry tài kuài((speaks in Mandarin))
           too fast
18         (.)
19   T     muito rápido (4.0) ((writes on board)) muito rápido (0.5) hóu faai muito
           too   fast                             too   fast         too fast too
20         rápido (0.5) rápido(.) fast (2.0) muito rápido
           fast         fast                 too   fast
21         (1.0)

22         a:h I’ll play it again, okay?

In this first sequence, we find something that strikes our attention at first sight. That is the teacher’s reaction (line 13) following students’ silence at line 12. After waiting 3.5 seconds for some kind of response from the students, the teacher utters something in Cantonese, the L1 of the majority of the students (“hóu faai oh”, meaning “too fast”), providing an account for students’ lack of response to the audio recording they have just heard. The audio as it is possible to observe (lines 3–10) provides a jocular end, but students remain in silence after listening to it (line 12). The teacher’s reaction (line 13) to students’ silence shows us an account for students’ lack of understanding, since regarding an audio as ‘too fast’ provides an account for the listening competence of the students as (which we have mentioned above) quite limited.

This account produced by the teacher takes us to the point that is central here: the use of students’ L1. Immediately after the teacher produces this account, students laugh in unison at line 14. The same thing happens at lines 15–16, which allows us to suggest that there is a jocular sequence here. Previous studies on CA (Glenn and Holt 2013Glenn, Philip and Elizabeth Holt 2013 Studies of laughter in interaction. London; New York: Bloomsbury Publishing. ; Jefferson 1979Jefferson, Gail 1979 “A Technique for Inviting Laughter and its Subsequent Acceptance Declination.” In Everyday Language: Studies in Ethnomethodology, ed. by George Psathas, 79–96. New York: John Wiley & Sons., Jefferson 2015Jefferson, Gail 2015 Talking about Troubles in Conversation. Ed. by Paul Drew, John Heritage, Gene Lerner, and Anita Pomerantz. New York: Oxford University Press.) have alerted us to the occurrence of laughter within conversation where laughter does not reflect humor or jocularity. In this case, T defuses the problematic aspect of this task – such as, students do not know where to start – and does this by formulating the students’ problem, in a jocular fashion.

Although the joke in the audio was not recognized by the students, here we can finally identify the sequence as a jocular sequence since the interactional work of the participants points to the presence of such a phenomenon. According to Sacks (1992Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell., 120), “one way that persons go about seeing activities is by reference to some procedures which they take it properly occur as the activities occur”. It is possible to observe that the teacher has produced a jocular utterance by virtue of the fact that students laughed right after the teacher uttered something in their L1 (as if the utterance at line 13 were working as a ‘punch line’). Then our question here should be: to which elements are students orienting to regard the teacher’s utterance (“hou faai oh”) as a jocular utterance? In other words, what is making them laugh?

To answer this question, we may first take into account the sequential elements of talk. As already mentioned, at lines 13–16 we can observe that participants react to the utterances produced by the teacher as a sequence of jokes. In “Sequencing: Utterances, Jokes, and Questions”, Sacks (1992)Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell. reminds us that when one uses jokes is something like ‘buying a drink among a bunch of people’ (p. 100). When one does that, drinks start to come in rounds. The teacher’s utterance (“hóu faai oh”) – as an invitation for laughter – appears to have an analogous feature.

After the first “hóu faai oh” (line 13) is uttered, the laughter in response (line 14) provided space for T to repeat this utterance (line 15) inviting more students to join in with laughter (line 16). Harry is oriented to this organization as he provides the same expression as T; however, he does so in Mandarin (line 17, see below). T’s repetition begets further laughter (Jefferson 1979Jefferson, Gail 1979 “A Technique for Inviting Laughter and its Subsequent Acceptance Declination.” In Everyday Language: Studies in Ethnomethodology, ed. by George Psathas, 79–96. New York: John Wiley & Sons.), building upon the class’s response to the first utterance produced by the teacher at line 13. These invitations to laughter/laughter response highlight the internal work of the sequence itself. As in a summons-answer sequence (Schegloff 1968Schegloff, Emanuel A. 1968 “Sequencing in Conversational Openings.” American Anthropologist, 70 (6): 1075–1095. ), when a member responds to the first pair-part in a second position of this adjacency pair, we can see the relevance of the first pair-part for the continuation of the sequence. When the teacher repeats the utterance “hou faai oh” (line 15), he is not only inviting more laughter, but also using this utterance as a ‘hint’ and his non-verbal move (turning to the board) to indicate that further talk is about to come. Nonetheless, the teacher’s utterance here is not enough to carry out this enterprise. Students’ attention (depicted by a refraining from laughter at line 18) is necessary to show that students are following the sequence. We can say that students’ laughter at line 16 is a reaction to the joke (the same uttered previously at line 13), but the lack of laughter at line 18 is a conditionally relevant move (Schegloff 1968Schegloff, Emanuel A. 1968 “Sequencing in Conversational Openings.” American Anthropologist, 70 (6): 1075–1095. ), which makes reference to a ‘nextness’ that will be located within the sequence. However, this ‘nextness’ seems not to be the phenomenon of order* to which Harry (line 17) orients himself. His actions indicate that the use of students’ L1 (regarded as a ‘joke’ by his fellow students) is still relevant.

As pointed by Sacks (in his lecture mentioned previously) after the first joke is told, every person will have a chance to talk (jokes start to come in rounds). As the teacher’s utterance in Cantonese (the L1 of the majority of the students) was regarded as a ‘joke’, this opens a space for other students, such as Harry (line 17) to utter something in his L1, Mandarin (“tài kuài”). However, as mentioned, in this situation, students do not produce any laughter after Harry’s utterance. The sequence of ‘jokes’ then seemed to have reached an end. This is probably associated with the teacher’s move produced at line 15, which, as already indicated, worked as a hint indicating that further talk would be produced. This is materialized in students’ reaction at line 18, in which, as already observed, a refraining from laughter becomes visible. This lack of laughter is an indication that they were waiting for further instruction, which came off at line 19, where the teacher switched back to Portuguese and started to write on the board explaining the meaning of the words uttered before in Cantonese. We can say that the lack of laughter and the move performed by the teacher are associated with the pursuit of a lesson topic as a counter to the continuation of a series of jokes. The teacher’s move was then a clear indication that a ‘next classroom event’ was selected, which was: [providing the equivalent of “hou faai” in the target language].

However, was the selection of the ‘next classroom event’ only due to the sequential organization of the utterances as discussed so far? The answer should be a clear “no”. There was also a categorial work furnishing the sequence of the interaction. When the teacher uttered some words in Cantonese (line 13), the category ‘teacher’ was momentarily put in the background giving place to the category ‘non-native’. This probably explicates the element to which students were orienting when they laughed. Once the category ‘non-native’ was invoked, the pair ‘native-non-native’ was established (with students acting as ‘natives’ and the teacher as ‘non-native’). However, when the teacher switched to English (the language of instruction of the class) and turned to board, he was starting to fit a ‘mentionable’ (Schegloff and Sacks 1973Schegloff, Emanuel A., and Harvey Sacks 1973 “Opening up Closings.” Semiotica 8 (4): 69–99. ) lesson topic into the course of the conversation. It is central to highlight that this is accomplished by employing some of the resources available (for example, the board) in the classroom to the local organization of his utterance. This move, combined with the students’ reaction (silence at line 15), indicated that a lesson topic was being introduced and that the category ‘teacher’ became more operative again. So the categorial pair ‘teacher-student’ had its full operation resumed, being re-invoked as a rational and jointly accomplished phenomenon of order*.

The suspension of the pair ‘native-non-native’ was also observable during Harry’s turn. His utterance in his L1 (Mandarin) would fit ‘better’ the situation if it were produced by a non-L1 speaker (as happened in previous lines when the teacher uttered something in Cantonese). Since participants did not regard Harry’s utterance as ‘laughable’, they were also indicating that Harry’s utterance was categorized as ‘standard’, nothing special. ‘Native-non-native’ is a relational pair of membership categories that was more ‘in use’ at lines 13–14 (rather than the relational pair ‘teacher-student’), but not at line 17, when students were already more aligned with the pair ‘teacher-student’ again. The pair ‘teacher-student’ is the potentially ‘always-relevant’ (or omnirelevant) pair in a classroom encounter; it is an “oriented-to” device throughout the interaction (Zimmerman 1998Zimmerman, Don H. 1998 “Identity, Context and Interaction” In Identities in Talk, ed. by Charles Antaki and Susan Widdicombe, 87–106. London: Sage., in Richards 2006Richards, Keith 2006 “Being the Teacher: Identity and Classroom Conversation.” Applied Linguistics 27 (1): 51–77. ). When these (omnirelevant) relational pairs are in use, there is no way of excluding their operation (Sacks 1992Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell.). However, there is no doubt that the pair ‘teacher-student’, while omnirelevant, might not always be the only pair in operation as, often, other pairs may become more salient during the course of the interaction.

This has to do with the category-distribution order* that goes beyond the production format established in initial turns. The pair ‘teacher-student’ then furnishes the production format in the beginning, but other relevant pairs might emerge through a dynamic layering of ongoing relevancies depending on the interactional methods participants use or even some contingency to which they orient themselves. When that happens, there is a hitch in the progressivity of the interaction (Fitzgerald, Housley, and Butler 2009Fitzgerald, Richard, William Housley, and Carly W. Butler 2009 “Omnirelevance and Interactional Context.” Australian Journal of Communication 36 (3): 45–64., 62) as people momentarily “halt doing what they are doing” and an operation of an omnirelevant category has to be restored so that the previous activity can be resumed.

Participants produce and are very sensitive to categories and their utterances show “the intricate and inextricable interweaving of categorial and sequential aspects of talk” (Watson 1997Watson, Rod 1997 “Some General Reflections on ‘Categorization’ and ‘Sequence’ in the Analysis of Conversation.” In Culture in Action: Studies in Membership Categorization Analysis, ed. by Stephen Hester and Peter Eglin, 49–75. Washington, D.C.: University Press of America., 69). At lines 13–14, it seems that the categorial pair ‘teacher-student’ was less operative at the moment, since the pair ‘native-non-native’ was more evident due to the participants’ reactions to the expression uttered in Cantonese by the teacher. However, when the teacher turned to the board in, the categorial pair ‘teacher-student’ seemed to be more relevant again, and a new classroom event [providing the equivalent of “hou faai” in the target language] was established (lines 19–20). The relational pair ‘teacher-student’ does not need to be specified or formulated out loud all the time during a classroom event. According to Sacks (1976Sacks, H. (1976) On formulating context. Pragmatics Microfiche, 1 (7): pp. F5–G8., F10), there are ways of invoking orderliness of activities, “without having to formulate what the terms are of that orderliness, let’s say, i.e., without having to say who you are and who I am for now. Or what you are really doing now”.

Sacks meant that we do not need to specify the omnirelevant pair ‘teacher-student’ all the time. However, when participants start to orient to other pairs while in class (as it happens in the extract above), the pair ‘teacher-student’ needs to become operative so that the institutional character of the encounter can be resumed. This happens through a relevant sequence of moves: [turning to the board] (line 15) – which works as an attention-getting device – and [paying attention to T] (lines 17–18).

This points to the fact that categories are not relevant because they are potentially ‘embedded’ in the discourse as analytically objects, that is, structurally or sequentially-given, but because they are reflexive tied and mutually constitutive (Watson 1997Watson, Rod 1997 “Some General Reflections on ‘Categorization’ and ‘Sequence’ in the Analysis of Conversation.” In Culture in Action: Studies in Membership Categorization Analysis, ed. by Stephen Hester and Peter Eglin, 49–75. Washington, D.C.: University Press of America.). The production of an utterance can be categorial and sequential, as we can see looking back at the sequence depicted at lines 13–14. Participants’ conjoint orientations to the utterance “hou faai oh” inform the categorial relevance here. Students were analyzing the teacher’s utterance as something that transcends the activity in course, something not directly related to the classroom event [listening to the audio and filling in the blanks] they were producing. The laughing produced by the students was working as an indicator that another activity was taking place [making a joke]. But the category ‘teacher’ is usually less ‘laughable at’ since it is marked in interaction as superior in status (Speier 1976Speier, Matthew 1976 “The Child as Conversationalist: Some Culture Contact Features of Conversational Interactions between Adults and Children.” In The Process of Schooling: A Sociological Reader, ed. by Martyn Hammwersley, and Peter Woods, 98–103. London: Routledge & Kegan Paul.) to ‘student’. The fact the teacher was laughed at then points to a categorical context in which the relational pair ‘teacher-student’ was (semi-)suspended as the other pair ‘native-non-native’ became more relevant. This reciprocal elaborative aspect of utterances and other moves carried out by participants render the sequential analysis of their talk as category-bound (Watson 1997Watson, Rod 1997 “Some General Reflections on ‘Categorization’ and ‘Sequence’ in the Analysis of Conversation.” In Culture in Action: Studies in Membership Categorization Analysis, ed. by Stephen Hester and Peter Eglin, 49–75. Washington, D.C.: University Press of America.).

Although aspects of social structure are visible during a class, which help participants distinguish utterances as teacher’s talk and as students’ talk. Yet these aspects are only observable, kept, changed or replaced as phenomena of order* according to participants’ sensible production and monitoring of an utterance, not to an unexamined, aprioristic sense imposed by the analyst.

5.2Categorizing speakers of different language varieties

As noted in the description of our data, C2 was composed of advanced-level students with a very good understanding of Portuguese. In the class from which the second excerpt was drawn, the teacher was working with his students on textual coherence. They were reading a composition on social violence and exploring how different ideas were introduced in relation to the main topic of the text and developed through its paragraphs. However, the students became confused about a word in the text. The teacher offered an explanation, but this did not seem to solve the problem. Therefore, the teacher asked some of the students to repeat the word in their own language to repair the indexicality of the term and ensure that everybody understood it. The main point of analysis happens in the subsequent lines, when the participants find a synonym for that term in Portuguese.

The following excerpt comprises the same transcription conventions used in the first two excerpts, with one exception. As the language of instruction in C2 was Portuguese, an English translation is provided on the right-hand side of each line rather than beneath the line to better orient the reader.

(2)

1    T       não? demônio é o mau, o mau,         no? demon is the evil, the evil,
2            aquela entidade que faz o mau,       that entity that does harm to people,
3            espíritos, espíritos                 the spirits, spirits
4            (2.5)
5    T       como se fala em chinês?              how do you say this in Chinese?
6            (0.6)
7    Ss      mò gwái ((fala ‘demônio’ cantonês))  mò gwái ((“devil” in Cantonese))
8            (1.0)
9    Cecilia demon                                demon
10           (0.6)
11   T       demons? aqui é demons (.)            demon? here it says demons (.)
12           porque são demônios ((indicando      because they are devils ((stresses
             que a palavra está no plural))       that word is in plural form))
13           (1.2)
14   Cecilia diabo=                               devil=
15   T       =é o diabo, exatamente=              =yes, the devil, exactly
16   Cecilia =diabo ((reproduzindo o sotaque do   =devil ((reproduces teacher’s
             professor))                          accent))
17           (.)
18   Ss      hhha ha [ha ha                       hhha ha [ha ha
19   T and   [ha ha ha ha                                 [ha ha ha ha ha
     Ss      

In this sequence of an L2 advanced level class, we can say that the relational pair of categories ‘native’ and ‘non-native’ is also present, but a sub-class of this pair starts to emerge. Students here distinguish between norms of the L2 and assume one of these varieties as their norm. This can be unpacked as follows.

Cecilia repeats the word “diabo” (line 16) reproducing the teacher’s accent. When she does that, students laugh regarding her utterance as ‘funny’. It is interesting, though, to observe that only Cecilia’s utterance is regarded as funny, not the teacher’s one, who had uttered the same word (with similar accent) in the previous turn. Here we can see how participants categorize identities. The teacher came from a region of Brazil in which the affricate [dʒ] rather than the phone [d] is used when the letter “d” is combined with the vowels “i” and “e” in pre and post-stressed syllables. Therefore, “diabo” is pronounced as [dʒi’abu], not [di’abu] as in standard European Portuguese. The teacher’s pronunciation of the word “diabo” as [dʒi’abu] is an utterance (or an account) that can be seen as a predicate of the category ‘speaker of Brazilian Portuguese’. This is in accordance with one of the viewer’s maxims identified by Sacks (1992Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell., 252), which is: “if a member sees a category-bound activity being done, then, if one can see it being done by a member of a category to which the activity is bound, see it that way”. When one can see what happened (pronouncing [dʒi’abu] instead of [di’abu]) then one can also see why that happened since, according to Hester (2000Hester, Stephen 2000 “The Local Order of Deviance in School: Membership Categorization, Motives and Morality in Referral Talk.” In Local Education Order: Ethnomethodological Studies of Knowledge in Action, ed. by Stephen Hester, and David Francis, 197–222. Amsterdam, London: John Benjamins. , 201), “both activity and motive can be seen as predicated of the category”, in this case, ‘Brazilian speaker of Portuguese’.

There is no doubt here that Cecilia ‘saw’ what happened, since she also reacted with laughter at line 19. We can call this a ‘category bound activity’ since this activity or utterance ([dʒi’abu]) is done by a member of a category (‘speaker of Brazilian Portuguese’) to which the activity is bound. The reaction produced by Cecilia’s fellow students at line 18 (laughter) after a micropause (line 17), however, shows that her utterance struck her colleagues’ attention and was regarded as ‘funny’. As the students (including Cecilia) spoke the European norm of Portuguese, Cecilia’s utterance is seen as something that does not hold for the category ‘speaker of European Portuguese’. In other words, Cecilia is not seen as a member of the category to which the utterance she produces is a predicate. There is an ‘anomalous pairing’ that generates a lack of fit between the activity (Cecilia’s utterance at line 16) and the membership category (speaker of European Portuguese) to which she belongs. We can then see a clear relational pair of categories at work here: ‘speaker of European-Brazilian Portuguese’.

The teacher’s same accent was regarded as something ‘natural’, ‘ordinary’, just like Harry’s utterance in the previous excerpt. Nonetheless, Cecilia’s accent (at line 16) produced immediately after the teacher’s utterance is puzzling and regarded as laughable by her colleagues (line 18). Here we can see the complicated and the mutually inextricable relation between relational categories and sequencing. Although the excerpt above is an instance drawn from a ‘formal’ setting (classroom), the relational pair of categories ‘teacher-student’ (at least as an unqualified, or general pair) does not remain static throughout the whole encounter. There are moments of ‘disruption’, since ‘informal’ categories such as ‘acquaintances’, ‘strangers’, ‘natives’ and ‘non-natives’ start to be combined with (or momentarily replace) the categories ‘teacher’ and ‘student’ and inform the production and monitoring of utterances. There is an order* being established. Utterances are then interactional realizations of a categorial order* (Watson 1994Watson, Rod 1994 “Catégories, séquentialité et ordre social [Categories, sequentiality and social order].” Raisons Pratiques 5: 151–185.), and not simple products of a discourse format.

I am not saying, though, that actions are not context-sensitive phenomena. We all know that in the classroom teachers and students have a routinely recognized set of rights, obligations and expectations. However, we cannot treat categories as analytically ‘given’, since such premature claims impose “potentially distorting constraints on subsequent analysis” (Richards 2006Richards, Keith 2006 “Being the Teacher: Identity and Classroom Conversation.” Applied Linguistics 27 (1): 51–77. , 59). Instead categories should be recognized through participants’ talk, which is not shaped by an interactional format, but by teacher and students themselves, who are displaying features of their talk when accomplishing their practical activities inside that environment. When doing that, they use certain membership categories (sometimes organized as paired categories) to which they orient their actions. This is part of the practical accomplishment and locally situated character of classroom activities (Hester and Francis 2000Hester, Stephen, and David Francis 2000 “Ethnomethodology and the Local Educational Order.” In Local Educational Order, ed. by Stephen Hester and David Francis, 1–19. Amsterdam: John Benjamins Publishing Company. ).

Looking back at the excerpt above, we can see how the teacher and students are orienting to the same relation pair of categories ‘speaker of European-Brazilian Portuguese’. This becomes evident, as already mentioned, when Cecilia produces a similar accent to her teacher’s and the rest of the class react with laughter. After hearing the students’ laughter in response to Cecilia’s utterance, the teacher (as well as Cecilia) joined them (line 19), which shows that the teacher was also orienting to the same relational pair. A good point that supports this conclusion is that by producing subsequent laughter, the teacher is indicating that laughter in this case is appropriate (Jefferson 1979Jefferson, Gail 1979 “A Technique for Inviting Laughter and its Subsequent Acceptance Declination.” In Everyday Language: Studies in Ethnomethodology, ed. by George Psathas, 79–96. New York: John Wiley & Sons.).

However, before both parties (teacher and students) get aligned with the same relational pair, we can observe that the participants (lines 1–12) were orienting to the relational pair ‘teacher-student’. This is also so since, categorical context-wise, it is usually more difficult to laugh at someone categorized as ‘teacher’. Students were not laughing at their teacher when he pronounced (more than one time) the word ‘diabo’ with his accent (I am indebted with Rod Watson for this brilliant observation). The other element that points to the fact that the relational pair ‘student-teacher’ was more operative at the moment is the sequential organization of members’ utterances. As we can see (lines 1–13), they were engaged in the task of [defining a meaning for the word ‘demônio’ (demon) in Portuguese], with the teacher [providing definitions] (line 2 – “that entity that does harm to people”) and [providing examples of other words in the target language] (line 3 – “the spirits, spirits”), or even [asking for an equivalent word in students’ L1] (line 5 – “How do you say this in Chinese?”).

It’s only after Cecilia’s utterance at line 16 that the participants achieve the specifically identifiable relatedness between the relational pairs ‘student-teacher’ and ‘speaker of European-Brazilian Portuguese’, with the latter more operative than the former. This explicates how one utterance containing one membership categorization device (line 15) and another utterance containing a different device (line 16) comprise one tying procedure for establishing the specific relatedness (Watson 1997Watson, Rod 1997 “Some General Reflections on ‘Categorization’ and ‘Sequence’ in the Analysis of Conversation.” In Culture in Action: Studies in Membership Categorization Analysis, ed. by Stephen Hester and Peter Eglin, 49–75. Washington, D.C.: University Press of America.) of a sequence, in this case, one evaluation (line 15 – “é o diabo, exatamente”) and a comment on the evaluation (line 16 – “diabo”, when Cecilia reproduces teacher’s accent). In this particular sequence, the evaluation (line 15) is performed by the teacher under a ‘teacher-student’ relational pair, but the comment on the evaluation is not carried by the category ‘student’, since only another categorial pair ‘speaker of European-Brazilian Portuguese’ can seem to fit to account for the laughs produced later. The categorization device then changes from ‘classroom participants’ to ‘speakers of Portuguese’. Furthermore, it also sheds light on how utterances are managed and accomplished locally, with reference to the ‘here-and-now’ elements of the interaction.

As we can see, observations and analysis of the sequential and categorial aspects of interaction should not come here at the expense of each other (Hester and Eglin 1997Hester, Stephen, and Peter Eglin 1997 “Membership Categorization Analysis: An Introduction.” In Culture in Action: Studies in Membership Categorization Analysis, ed. by S. Hester, and P. Eglin, 1–23. Washington, D.C.: University Press of America.). Both membership categorization activities and sequential organization are inextricably intertwined. Therefore, one type of analysis will inform the other. The way participants categorize things in talk provides examples of their common (socially shared) knowledge of social structures, which furnishes the sequence of actions, which in turn explicate the production of the categorization work carried out by participants. In the example depicted above, we can see that the adjacency pairs (lines 15–16) work as a resource for the mutual and situated orientation of the relational pair ‘speaker of European-Brazilian Portuguese’ since the second pair part performed by Cecilia at line 16 is produced and hearably designed for just that first pair part (performed by the teacher in 15) at that single moment.

Cecilia (line 16) shared with her colleagues and the teacher the same knowledge about the aspects of the categorial pair ‘speaker of European-Brazilian Portuguese’ regarding the pronunciation of the word “diabo”, since the laughs following her utterance (sequentiality) show the in situ recognition of the categorial pair initiated by Cecilia.

We can then say, according to what Fitzgerald, Housley and Butler (2009Fitzgerald, Richard, William Housley, and Carly W. Butler 2009 “Omnirelevance and Interactional Context.” Australian Journal of Communication 36 (3): 45–64., 54) suggest, that the locally managed “use of socially shared knowledge can reveal a level of context within the interaction that has relevance beyond the local task”, since the relation pair ‘speaker of European-Brazilian Portuguese’ is invoked and ‘layered over’ the pair ‘teacher-student’, which became less relevant at that moment as a phenomenon of order*.

6.Concluding remarks

As claimed throughout this paper, sequential devices and the membership categories are both methods used by members that render their talk oriented to the ‘here-and-now’ of the interaction. These methods are related to the distinctly identifying features of these particular social actions in a particular setting (or, as Garfinkel (2002)Garfinkel, Harold 2002 Ethnomethodology’s Program: Working out Durkheim’s Aphorism, ed. by Anne W. Rawls. Lanham, MD: Rowman & Littlefield Publishers. would put it, ‘haecceities’). As such, the discussion of these methods should never come at the expense of the other. They are not simply complementary, but inextricably related as parts of a same whole.

All this implies a notion of membership categories as context-oriented phenomena. The concept of ‘gestalt-contexture’ (Garfinkel 1967Garfinkel, Harold 1967 Studies in Ethnomethodology. Cambridge: Polity Press.; Gurwitsch 1964Gurwitsch, Aron 1964 The field of consciousness. Pittsburgh: Duquesne University Press.) that includes categorial pairs is of relevance here, since, as reported by our data, a categorial pair invoked is not something in and for itself, but is informed by a complex contextualized interactional work accountably produced and recognized by the participants. By ‘gestalt’ I mean that a context is produced by joint practical actions, and that those actions are social. They are visible to the participants as whole, rather than as separate phenomena that are worked out by a strictly rational synthesis of meanings (Liberman 2013Liberman, Kenneth 2013 More studies in Ethnomethodology. New York: State University of New York Press.). For example, Gurwitsch (1964)Gurwitsch, Aron 1964 The field of consciousness. Pittsburgh: Duquesne University Press. argues that a melody may not be considered merely the sum of the notes of which it consists. This means that the organizational structure of social actions is not exhibited by single cases, but by a more complex combinatory configuration of its elements. Our understanding of the world is much more dependent on the relation of its elements than those elements in isolation. This, according to Gurwitsch (1964)Gurwitsch, Aron 1964 The field of consciousness. Pittsburgh: Duquesne University Press., is a logical necessity.

However, this does not mean that there are no saliences in this process. Subsequent actions can be responses to a salience of a gestalt-contexture. As shown by our data, depending on the interactional work, the omnirelevant relational pair ‘teacher-student’ is sometimes in the ‘taken-for-granted’ background, while some other times it is more foregrounded, more ‘explicitly apparent’ in the ongoing production and organization of the interaction. For this reason, Membership Categorization Analysis focuses on the “incarnate intelligibility” of conduct in particular contexts (D’hondt 2013D’hondt, Sigurd 2013 “Analyzing Equivalences in Discourse: Are Discourse Theory and Membership Categorization Analysis Compatible?” Pragmatics 23 (3): 421–445. ), which are made concrete by the complex interactional work carried out by the participants. Through the data discussed in this paper, it was possible to observe that participants were securing joint orientation and mutual understanding of the categorial pairs being invoked. And they were not doing that only by making categorizations, but also by orienting to the sequential features (and the local organization) of the utterances.

The observations and discussions brought here are useful to advocate the advantages of analyzing locally produced phenomena without separating closely intertwined and mutually constitutive elements such as sequential and categorial devices. By doing this we are treating the data as real source of phenomena; and the participants as real subjects, as members and producers of the phenomena, who do and analyze practices in situ, rendering them observable to the analyst. If we decide to develop a truly praxiological study, then we have to concentrate our discussions on the phenomena of order* ‘being reported’ by the data, on the features provided by the internal elements of the interaction, instead of deliberately excluding or moving some of them to the background only for the sake of some theoretical assumptions or pre-motivated methodological orientations. Categorization practices are relevant only to a particular occasion. This does not exclude the possibility to find similarly categorially organized elements in other situations (Carlin 2003Carlin, Andrew P. 2003 “Pro Forma Arrangements: The Visual Availability of Textual Artefacts”. Visual Studies 18 (1): 6–20. ) but such similarities need to be established empirically rather than alluded to and, if found, such similar elements should be seen as phenomena made to happen “for ‘another first time’” (Garfinkel 1967Garfinkel, Harold 1967 Studies in Ethnomethodology. Cambridge: Polity Press., 9).

This is, I think, a very important point for pragmatic studies. Since pragmatics investigates natural language uses, it is central to discuss these uses as they become available to participants in a conversation. If we, as analysts, try to fashion our discussions to fit theoretical frameworks instead of letting “the materials fall as they may” (Sacks 1992Sacks, Harvey 1992 Lectures on Conversation [Volumes I and II], ed. by Gail Jefferson with an Introduction by Emanuel, A. Schegloff. Oxford: Blackwell., 11), then we are not interested in language uses, but only in our own a priori beliefs about these uses.