Numa Markee

Department of Linguistics

University of Illinois at Urbana-Champaign

4080 Foreign Languages Building

707, S. Mathews

Urbana, IL 61801

Tel: 217 419 2145

Email: nppm@illinois.edu

1

DRAFT

On competence and intersubjective agency:

A post-cognitive perspective

ABSTRACT

Inspired by the stance of ethnomethodological indifference to a priori theories and constructs (Garfinkel & Sacks, 1970; Lynch, 1999), I argue that a significant part of the etic (i.e., researcher-relevant) theorizing that has gone into the evolutionary development of the construct of interactional competence over the years is not relevant to work in ethnomethodology (EM) and conversation analysis (CA) that seeks to understand the processes of second language classroom interaction and learning. More specifically, in the pages that follow, I begin by briefly reviewing how competence has been understood in different literatures, focusing in particular on how this notion has morphed over time into the related constructs of communicative and interactional competence. I then argue that many of the theoretical accretions that have been borrowed from various etic, a priori theories of language and language learning should be bracketed by ethnomethodological CA practitioners. In this context, I also argue that a great deal of so-called ethnomethodological CA work in applied linguistics is in fact nothing of the kind and is much better described as hybrid form of CA. Thus, if we are serious about doing ethnomethodological CA, I suggest that we need to revert to a theoretically sparer, emic (i.e., participant-relevant), post-cognitive version of competence, which emphasizes the idea of intersubjective agency. In the empirical section of the paper that follows the literature review, I then illustrate how the construct of intersubjective agency may help us understand the micro-processes of classroom interaction and language learning behavior. To this end, I use video fragments, associated cultural artifacts, and transcripts of embodied talk-in-interaction to develop a post-cognitive, multimodal conversation analysis of classroom interaction produced by different levels of learners in different iterations of an ESL class that targeted members of the local community in a Mid-Western town in the United States. These analyses show how two complementary curricular levels of planning (see Suchman, 2007) — that is, tasks-as-work-plans (i.e., written classroom materials) and tasks-as activity (i.e., the actual, methodological implementation of these work plans; see Coughlan & Duff, 1994) — put students in the position of having to deploy intersubjective agency, which potentially allows them to learn new language as a by-product of talk-in-interaction. Finally, I show how this innovative curricular approach may be used to challenge the bureaucratic notion of level. More specifically, I show how even false beginners can potentially outperform more advanced students who are exposed to the same pedagogical materials but who have to meet more stringent, pre-specified requirements of enacted interactional activity. I conclude the paper with a brief discussion of where the kind of research program outlined here might lead us.

2

Inspired by the stance of ethnomethodological indifference to a priori theories and constructs

(Garfinkel & Sacks, 1970; Lynch, 1999), I argue that a significant part of the etic (i.e.,

researcher-relevant) theorizing that has gone into the evolutionary development of the construct

of interactional competence over the years is not relevant to work in ethnomethodology (EM)

and conversation analysis (CA) that seeks to understand the processes of second language

classroom interaction and learning. More specifically, in the pages that follow, I begin by briefly

reviewing how competence has been understood in different literatures, focusing in particular on

how this notion has morphed over time into the related constructs of communicative and

interactional competence. I then argue that many of the theoretical accretions that have been

borrowed from various etic, a priori theories of language and language learning should be

bracketed by ethnomethodological CA practitioners. In this context, I also argue that a great deal

of so-called ethnomethodological CA work in applied linguistics is in fact nothing of the kind

and is much better described as hybrid form of CA. Thus, if we are serious about doing

ethnomethodological CA, I suggest that we need to revert to a theoretically sparer, emic (i.e.,

participant-relevant), post-cognitive version of competence, which emphasizes the idea of

intersubjective agency. In the empirical section of the paper that follows the literature review, I

then illustrate how the construct of intersubjective agency may help us understand the micro-

processes of classroom interaction and language learning behavior. To this end, I use video

fragments, associated cultural artifacts, and transcripts of embodied talk-in-interaction to develop

a post-cognitive, multimodal conversation analysis of classroom interaction produced by

different levels of learners in different iterations of an ESL class that targeted members of the

local community in a Mid-Western town in the United States. These analyses show how two

complementary curricular levels of planning (see Suchman, 2007) — that is, tasks-as-work-plans

3

(i.e., written classroom materials) and tasks-as activity (i.e., the actual, methodological

implementation of these work plans; see Coughlan & Duff, 1994) — put students in the position

of having to deploy intersubjective agency, which potentially allows them to learn new language

as a by-product of talk-in-interaction. Finally, I show how this innovative curricular approach

may be used to challenge the bureaucratic notion of level. More specifically, I show how even

false beginners can potentially outperform more advanced students who are exposed to the same

pedagogical materials but who have to meet more stringent, pre-specified requirements of

enacted interactional activity. I conclude the paper with a brief discussion of where the kind of

research program outlined here might lead us.

LITERATURE REVIEW

Since other contributors to this symposium have done an excellent job of tracing how the

theoretical notion of interactional competence has evolved over the years, I will not duplicate

these efforts in any detail here. Rather, I propose to motivate a discussion of why it is so

difficult for ethnomethodological conversation analysts to invoke this literature in a legitimate

fashion.

Let me begin by noting that the topic of interactional competence is one of the most

important issues in the CA literature on language learning within applied linguistics (see, for

example, Brouwer & Wagner, 2004; Firth, 2009; Hall, Hellermann & Pekarek Doehler, 2011;

Hellermann, 2006, 2007, 2008; Hellermann & Lee, 2014; Kasper & Wagner, 2014; Lee &

Hellermann, 2014; Markee, 2000, 2008; Pekarek Doehler, 2010, 2013; Pekarek Doehler &

Pochon-Berger, 2015, among others). But where does this concept come from? Briefly, the twin

notions of competence and performance were first coined by Chomsky (1965), who argued that

all normal human beings were hard-wired from birth with a highly abstract knowledge of

4

Language, and it was this individual, psycholinguistic, biological endowment — which he

named competence — that enables all of us to learn our native language with 100% success

within the first four or five years of our lives. In this model, performance is a residual category

that is of little interest to researchers working on generative syntax. These ideas were

enormously influential in the language sciences and beyond, but it was not long before

Chomsky’s ideas came under fire. The first major challenge to Chomsky’s position was

mounted by Hymes (1972), who argued that competence was as much a sociolinguistic concept

as it was a cognitive one. More specifically, Hymes argued that Chomsky had drawn the lines

between competence and performance in the wrong place, and suggested that the notion of

communicative competence provided a better account of how language learning and use were

intertwined.

These proposals were foundational for the theoretical development of communicative

language teaching and testing (see, in the first instance, Canale & Swain, 1981, and subsequent

work by, among others, Bachman & Palmer, 1982; Canale, 1983; Celce-Murcia, and Dörnyei &

Thurrell, 1995). A common characteristic of all these publications is that the notion of

communicative competence subsumes a number of presumed sub-competences (for example,

linguistic/grammatical competence, discourse/sociolinguistic competence, pragmatic

competence, actional competence, sociocultural competence, and strategic competence). While

such categories are undoubtedly useful for curriculum designers and testers as a check list of

what domains of language should be covered by communicative materials designers and testers,

the problem with this literature as I see it is that it is driven by a variety of etic, theoretical

perspectives — a point to which I return shortly. Furthermore, in terms of the internal

consistency of these various models of communicative competence, it is not always clear how

5

these different sub-competences are inter-related, nor how well they deal with intersubjectivity.

For these reasons, Kramsch (1986) called for research on interactional competence as a potential

solution to these kinds of problems.

This suggestion has received massive uptake in the applied linguistics literature. For

example, Hall (1993, 1995, 1999) adopted a socio-historical/cultural perspective to better

understand this construct. Young (whose work still constitutes the most developed non-CA

treatment of interactional competence) followed up on this work by adopting a practice theory

standpoint to further theorize this notion (see Young, 2000, 2009; 2011; Young & He, 1999;

Young & Miller, 2004). More specifically, over time, Young has incorporated all the major

insights from CA about turn taking, repair and sequence organization into his model.

Furthermore, his observations about the multi-semiotic nature of interaction are compatible with

the independent development of multimodal CA (see Goodwin, 2013). But since Young has

never self-identified either as an ethnomethodologist or as a CA researcher, he has also felt free

to incorporate a broad mix of insights from a number of etic theories of language and language

learning, including discourse analysis, sociocultural theory, systemic functional grammar, and

critical theory (especially the work of Bourdieu and other luminaries in this tradition). While

some may regard this as a desirable openness to potentially complementary theoretical

perspectives that is judiciously deployed in the service of new knowledge construction, I would

argue against such theoretical eclecticism.

More specifically, to my mind, the incorporation of these various influences into Young’s

writings is problematic for anyone who wishes to engage with the topic of interactional

competence from a specifically ethnomethodological CA perspective because: 1) Bourdieu’s

ideas raise difficult methodological issues regarding the question of how to deal with context (for

6

example, is context to be understood in broad cultural and political terms, as Bourdieu advocates,

or in the narrow sequential sense favored by conversation analysts? See Markee, 2015 for

further discussion); 2) although Young is careful to say that interactional competence is an

interactional accomplishment, not a matter of individual ownership of linguistic knowledge by

individuals (see, in particular, Young, 2011), he nonetheless sometimes contradicts himself, as

when he states that interactional competence “includes both knowledge and the employment of

that knowledge in different contexts of use” (Young, 2011: 438). Finally, and this is by far the

most important objection that I have to Young’s model, his theoretical eclecticism runs afoul of

the key principle of ethnomethodological indifference to a priori theories and constructs (see

Garfinkel & Sacks, 1970; Lynch, 1999). More specifically, for anyone who wishes to remain

true to CA’s radical ethnomethodological, emic origins, practicing such indifference is a non-

negotiable matter, because it is precisely CA’s agnostic stance toward exogenous theory, coupled

with its relative disinterest in theory (see Markee, 2008; Schegloff, 1984) that makes it such a

distinctive approach to analyzing classroom interaction and learning data.

For these inter-related reasons, I prefer to use the theoretically leaner term competence,

which is the term that is still used in the CA literature in sociology. Although this term was

borrowed from Chomsky by the founders of CA, this concept is entirely shorn of any

developmental connotations in CA and refers exclusively to participants’ observable orientation

to the underlying normative practices that organize turn-taking, repair and sequence

organization. Furthermore, in a standard move used by ethnomethodologists and conversation

analysts who wish to develop novel emic insights into how classroom processes of interaction

and learning are socially organized, I propose to bracket all the etic, pre-conversation analytic

literature alluded to above.1

7

I also claim that much of the excellent work carried out by, among others, John

Hellermann and his associates on the longitudinal development of interactional competence is an

example of hybrid, not ethnomethodological CA, in that this socio-cognitive research program is

heavily informed by the a priori, etic theoretical concerns of sociocultural theory (see, for

example, Thorne & Hellermann, 2015). Of course, I recognize how controversial this claim is.

Following Pekarek Doehler (2010), Hellermann’s work may represent the most systematic

program of CA-inspired research in this area and has without question yielded many useful

empirical insights into language learning over time. But I am entirely unpersuaded that CA

needs to appeal to exogenous theory to make useful contributions to social SLA, and I therefore

propose to also bracket the necessity for any such theoretical underpinnings in CA work on

classroom interaction and language learning.

For researchers who do not affiliate with EM or CA, this may seem an unimportant storm

in a small teacup, which is of little interest to the broader SLA community. However, if CA

researchers in applied linguistics revert back to a more purist CA conception of competence —

which is what I am advocating — that means that what CA is contributing to social approaches

to SLA is a post-cognitive, not merely a socio-cognitive perspective on language learning. But

what does adopting a post-cognitive perspective on interaction and language learning entail? A

preliminary statement of the issues may be found in a number of seminal publications,

specifically, Goodwin (2006); Schegloff (2006); Potter (2000); Te Molder & Potter (Eds.),

(2005) and the special issue of Discourse Studies on this topic curated by Van Dijk (2006). This

body of work — in which matters of cognition and learning are not an inherent part of CA

competence but are treated as topics of talk that are observably achieved in and through the rich

surface of interaction (Edwards, 2006) — is one of the most exciting developments in

8

ethnomethodology (EM), discursive psychology (DP) and CA in the last few years. Furthermore,

these ideas have already prompted a small post-cognitive literature in the applied linguistics

literature on learning processes from a socially-contexted language learning perspective (see, for

example, Kasper, 2009; Markee, 2008, 2011; Markee & Kunitz, 2013; Markee & Seo, 2009;

Mori & Hasagawa, 2009).

In the context of this discussion, I also suggest that we use the term intersubjective

agency as the preferred post-cognitive term to refer to how participants routinely enact

observable, embodied language learning and other contingently relevant pedagogical behaviors

that occur in the moment and over time during the course of classroom interaction. This new

term is designed to emphasize that ethnomethodological conversation analyses of classroom

interaction and learning are prototypically concerned with the social accomplishment of actions

that may result in language learning. Thus, using the Oxford English Dictionary as my source

for defining what this phrase means, intersubjectivity is to be understood as understanding that

“exists between conscious minds”, while agency is concerned with “action, capacity to act or

exert power; active working or operation; action, activity,” respectively.

Whatever learner-participants actually learn as a by-product of engaging in such activity

is, of course, an important matter, but proficiency is not part of competence as this term is

understood in CA. Furthermore, what students learn (or do not learn) at different times in their

lives becomes a matter that is worthy of investigation by researchers only when participants

themselves first talk or enact this issue into emic relevance. So, for example, the work done by

the participants studied in Markee & Kunitz (2013) on whether the Italian word ristorante is a

masculine or feminine noun clearly involves the students themselves talking about what they

think they know about the grammar of this word. As such, it becomes a relevant issue for

9

researchers to investigate also. This perspective is clearly quite different from that of

mainstream hybrid CA work on interactional competence, in which what students know in the

abstract about their second language(s) at different moments in their lives is an a priori, etic

concern.2

In the empirical analyses that now follow, I use video fragments, associated cultural

artifacts, and transcripts of embodied talk-in-interaction to develop a curricular, multimodal

conversation analysis of classroom interaction produced by students enrolled in an ESL class that

targeted members of the local community in a Mid-Western town in the United States. These

analyses show how two complementary, curricular levels of planning (see Suchman, 2007) —

more specifically, classroom materials (written tasks-as-work-plans) and classroom management

activities (the implementation of tasks-as-activity; see Coughlan & Duff, 1994) — worked hand

in glove to enable the students to deploy intersubjective agency as observable behavior in real

time. In the process, I show how this notion may be used to contest the bureaucratic notion of

level and how, when properly challenged by innovative task-based materials and assessment

specifications, even false beginners are potentially able to outperform more advanced students

who are exposed to the same pedagogical materials but who have to meet more stringent, pre-

specified requirements of enacted interactional activity. I conclude the paper with a brief

discussion of where the kind of research program outlined here might lead us.

DATA AND PARTICIPANTS

The participants in this paper are ESL students who range in levels from false beginners to native

speakers/advanced users of English who participated in different iterations of a demonstration

class on task based language teaching which I taught at least twice a year over a period of 27

years. The demonstration class was part of a course on task based language teaching for

10

graduate students enrolled in a MATESL program at a US university. The ESL students whose

talk is analyzed here include false beginners, low intermediate and native speakers/advanced

users of English, but given the symposium’s interest in lower level learners, I concentrate mostly

on the talk of false beginners and intermediate ESL learners; I am the teacher in all the fragments

analyzed here. I also developed the materials that were used in these lessons, which are

exhibited in Figures 1 and 2.

Figure 1 shows the picture that is used to set up a classic one-way information gap task.

In this task, four small groups of story tellers who all have access to this picture have to explain

under timed conditions to a student standing at the blackboard (the story drawer, who does not

have access to the picture) what happened to the person lying in the hospital bed. The story

drawer’s job is then to reproduce this picture as accurately as possible. As part of the design of

these materials, a number of difficult lexical items (at least for lower level students) have been

inserted in the knowledge that they are unlikely to know these words and that, consequently, they

will be forced to deploy their collective intersubjective agency in real time in order to solve this

problem. Thus, these materials do not teach students how to become competent in a particular

area of interaction (say opening or closing sequences). Rather, they force students to deploy

their intersubjective agency and to use whatever linguistic resources they currently have at their

disposal to solve the information gap at hand as best they can.

11

Figure 1: One-way picture description task (Markee, 1997: 36)

In Figure 2, we see that Task 1 specifies whether the principal focus is on fluency or

accuracy, what the story tellers and story drawer have to do, respectively, how they have to do it,

and the time they have available to accomplish this task. Importantly, this task also pre-specifies

the expected standard of graphic accuracy that the drawer has to be able to achieve in order for

the task to be completed successfully. So, for false beginners, success is defined as the students

being able to produce a very approximate version of the picture in 40 minutes. But for

NSs/advanced users of English, the amount of time allocated for task completion is much

12

shorter, typically 15 minutes, and the number of errors that are allowed, typically 10 errors of

detail, is much more challenging. And finally, Task 1 also specifies the kind of language that has

been been typically generated by this task over 27 years of empirically observed implementation.

Task 2 specifies how the picture drawn on the blackboard is to be evaluated, while Tasks

3-5 (not shown here for reasons of space) show how teachers might extend the two original tasks

for an entire unit’s worth of work lasting about a week of instruction. In sum, this unit

exemplifies what might be called a relatively weak implementation of task based language

teaching (se Markee, 1997, for further discussion). That is, what makes this unit communicative

is that the order of Presentation, Practice and Produce found in allegedly more traditional forms

of language teaching is reversed. Thus, in Task 1, students begin by communicating

naturalistically as best they can in light of the communicative resources that are available to

them, and then, in Task 2, in conjunction with the teacher, they return to examine the accuracy of

the language they have produced in the first phase of activity.

1. Picture description task 1 (focus on fluency; small group work). Time: About 40 minutes

Story tellers: Get into groups of four. Plan for about 5 minutes how you are going to describe the picture to the story drawer at the blackboard. Your task is to describe the picture to the story drawer, so that s/he can draw it on the blackboard.

Story drawer: When your classmates describe the picture to you, try to reproduce the picture they are describing as accurately as you can. Be sure to ask as many questions as you need in order to clarify the instructions you receive from your classmates (VERY IMPORTANT!)

Standard of graphic accuracy (= how accurate does the drawing have to be?):

o Beginners : Getting the very rough gist of the story. A very approximate picture is OK. Time limit for completing the task: Approximately 35-40 minutes.

o Intermediate: Getting the gist of the story. An approximate picture is OK. Time limit for completing the task: Approximately 25-30 minutes.

13

o Native speakers/Very advanced : No more than 10 errors of detail. Time limit for completing the task: 15 minutes.

Language this task will generate: Imperatives; the language of spatial relationships; the language of description (present tense, present continuous); figurative language; repair (comprehension checks, clarification requests, circumlocution).

2. Debriefing task (focus on accuracy; teacher fronted). Time: 20-30 minutes

With the help of the teacher, compare the original picture with the one on the blackboard and evaluate whether you have met the standard of graphic accuracy specified in Task 1.

. . .

Figure 2: Instructions for Tasks 1 and 2

ANALYSIS

For present purposes, we are specifically interested in what happens interactionally during the

implementation of Tasks 1 and 2.

First, let me show some data from a class that was composed of NSs/advanced users of

English which will illustrate what language learning behavior looks like. It so happens that

Fragment 1 shows how participants deal with a pronunciation problem, but the practices which

the participants orient to here are equally relevant to addressing other levels of language, and

they are endemic in all of the data that I have, irrespective of the students’ level of English.

M is the story drawer. She is an advanced user of English from Honduras who has

learned American English. J is an advanced user of English from Hong Kong who has learned

British English. K is a native speaker of American English, while S and B are, like M and J,

advanced users of English (all of the ethnographic information about whether the participants are

NSs or not is recoverable from the video tape). The talk exhibited in Fragment 1 barely lasts two

or three seconds.

INSERT FRAGMENT 1 ABOUT HERE

14

As we can see from M’s use of the open class repair initiator what in line 24, M does not

understand the word vase because in line 22, J uses the British rather than the American

pronunciation for this word. In line 25, J is unable to figure out what is problematic about her

previous turn and so she repeats the word vase, using the same British pronunciation that she had

used previously. Given that the opportunity for self-completion in next turn in line 25 of the

other initiated repair in line 24 has come and gone, this repair sequence is finally achieved as,

first, an other-completed repair by K in line 26, and then in line 27 as a collaborative turn done in

chorus by J and two of her group mates, S and B, during which everybody settles on the

American pronunciation of this word. In lines 28 and 29, M first repeats the word, also using the

American pronunciation for this word, and then thanks the group for clarifying this matter. M’s

turn closes this sequence, and another unidentified speaker in the group then continues to

describe what is in the vase in line 30.

In terms of how these participants deploy their collective intersubjective agency in this

talk, note that there is a great deal of repetition in this short fragment. However, while the

repetition in line 25 displays an inability on J’s part to understand what M finds problematic in

J’s turn in line 22, K’s repetition of the word in line 26 not only other-completes M’s other-

initiated repair on behalf of J but simultaneously provides new information abut how this word is

pronounced in American rather than British English. Furthermore, the choral response by S, J

and B in line 27 functions as an acceptance of K’s repair, and also gives these three participants

the opportunity to practice how this word is pronounced in American English. And as already

noted, M’s repetition of this word in line 28 is a precursor to closing the sequence down. Thus,

we can see that that, even in this banal, utterly mundane fragment of language learning behavior,

whatever learning actually occurs (and there is some evidence from later on in this lesson that J

15

has indeed taken this information onboard, at least in the short term), it is a by-product of some

rather sophisticated epistemic work that is rooted in the observably transient identities of the

participants as language experts or non-experts. Furthermore, this fragment also illustrates how

cognition and learning are progressively achieved as socially distributed phenomena, whose

results have the potential to benefit multiple participants who position themselves on the

sidelines of interaction, not just the individual whose talk is the formal object of explicit repair

work.

Let us now look at data which come from a class composed of low intermediate learners.

Fragment 2 (which, for analytic convenience, is subdivided into eight smaller fragments) also

focuses on interaction that occurred during Task 1, and shows how these particular participants

deployed their intersubjective agency to understand and use the word crutches (one of the

predictably difficult words deliberately introduced into the picture to force students to engage in

circumlocution) which was known by at least one of the story tellers, B, but which was not

known by the story drawer, S. The transcripts in Fragments 2.1-2.8 also include frame grabs to

show how absolutely crucial transcribing the embodied aspects of interaction are to

understanding classroom talk, since this is an environment in which all kinds of cultural artifacts

(here, mostly the black board) are routinely talked into relevance by participants.

INSERT FRAGMENT 2.1 ABOUT HERE

As we can see, in lines 1-7, the word crutch is introduced into the on-going activity by W,

who knows the English word for the objects lying under the patient’s bed. It transpires in line 8

that S does not know what this word means, and she therefore asks for a clarification. However,

notice that Frame Grab 1 (FG1) provides earlier evidence that S does not know the word crutch,

as her first embodied guess enacts somebody in a wheel chair pushing herself along with her

16

hands. In line 11, an W? tries, unsuccessfully, to explain what crutch means, and the rest of the

turn is taken up by laughter prompted by whatever the word uhntoo means.

In Fragment 2.2, S further pursues a clarification of what crutch means. More

specifically, during the 0.3 second pause in line 17, she “asks a question” by making a circular

motion around her waist (see FG2). In terms of the preference organization of Question-Answer

adjacency pairs, this first pair part (FPP) sets up the expectation that an answer will follow as the

preferred second pair part (SPP) in next turn. However, as shown by the 0.6 pause in line 18

following this embodied question, it is not altogether clear what this mime means.

Consequently, W continues her attempt to explain the meaning of this word by providing

information about where in the drawing the crutches are located. In other words, she begins in

lines 19-20 to wrestle with what is described as the language of spatial relationships in the

pedagogical materials exhibited in Figure 2.

INSERT FRAGMENT 2.2 ABOUT HERE

More specifically, when W says “his o(it)o e:h in- to- un- unde:r the: bed,” she

directs S to look at a particular part of the drawing-so-far.3 Lines 21-22 are unintelligible, but as

we can see from FG3, it seems that S has correctly interpreted what part of the picture she needs

to work on. Note that this understanding is achieved as an embodied action that is deployed

simultaneously with her unintelligible turn in line 22. In summary, S now knows where to look,

but she still does not know what crutch means.

In Fragment 2.3, therefore, the participants engage in more clarification work on this

lexical item. As in Fragment 2.2, however, the way B tries to clarify this word is embedded in

more interactional work that employs the language of spatial relationships as a resource for

further explanation.

17

INSERT FRAGMENT 2.3 ABOUT HERE

More specifically, in lines 27-31, B begins her turn by referring to they, not it, as a

referent for the noun crutches. Her turn is heavily marked by a number of perturbations (pauses,

cut-offs, vowel stretches) —which are all behaviors that are associated with the production of an

ongoing repair — which finally yield the grammatical information that crutches normally come

in pairs. In lines 33-35, W continues with her previous practice of using the language of spatial

relationships to provide information about where the crutches are to be found, specifically

“beside the bed” (as opposed to her previous formulation “under the bed” in line 19 of Fragment

2.2). After a brief initial pause of 0.6 seconds in line 37 (which may be an artifact of the overlap

in lines lines 35 and 36), B collaboratively completes W’s unfinished turn in line 35. However,

as shown by the inconclusive lack of uptake in lines 39-42, this work by B and W does not

provide the clarification that S is looking for.

In Fragment 2.4, we can see that W continues to deploy the language of spatial

relationships as the principal means of getting at the meaning of the word “crutch.”

INSERT FRAGMENT 2.4 ABOUT HERE

More specifically, in line 44, W again tells S that the crutch is beside the bed, and then in

line 46, adds the information that it is on the ground. In lines 47-48, nine seconds of an

unintelligible buzz of talk by multiple speakers ensue. For some reason, when the camera pans

back to S, FG 4 shows that she is standing with her back to the blackboard looking toward the

story tellers. And for the next 12 seconds, she continues to stand in front of the blackboard not

drawing. More specifically, in FGs 5-7, we can see that she first looks to her left, then to her

right and finally back toward the story tellers.

18

As we can now see in Fragment 2.5 the teacher treats this inactivity as a problematic lack

of deployment of intersubjective agency by S, which leads him to make a rare intervention in this

phase of the lesson plan.4

INSERT FRAGMENT 2.5 ABOUT HERE

More specifically, after a further four seconds of silent inactivity by S in line 49, the

teacher looks at his watch (see FG8), but even though S is looking at him as he enacts this

gesture, nothing happens for another four seconds (see line 50). Consequently, in FG9, the

teacher gestures toward the blackboard and tells S in lines 51-52 “you should draw what

they have told you,” with emphatic stress on the word draw. This instruction leads S to turn

back toward the blackboard and to start drawing for the next six seconds (see line 53).

In Fragment 2.6, S again indicates that she has not understood W and B’s previous

attempts to explain what crutch means.

INSERT FRAGMENT 2.6 ABOUT HERE

More specifically, in FG11, S turns round to face the story tellers and, as she does this,

says in lines 54-56: “I can’t (understand) ((unintelligible) you said that (she:::h.)”

In FGs 12 and 13, she then receives embodied feedback in the form of nods by Y, who also

possibly says the word bed with low volume in line 57. In line 58, S may be taking up this lead

when she possibly repeats this word, which is preceded by two perturbations (“uh” and the 0.3

second pause) indicating uncertainty on S’ part. Notice that this is the first time that S

cannibalizes the word beside from W and B’s immediately preceding talk and recycles it into her

own talk. In line 59, somebody says “yeah” and in lines 60 and 61, W and B again pursue the

question of where the crutch(es) are located in relation to the bed. These turns show how S is

now back on track in terms of the importance of maintaining intersubjective agency with her co-

19

participants, although it is also clear that the problem of what crutch means remains unresolved.

Notice also how the nods and mutual eye gaze in FGs 12 and 13 provide us with important

participant-relevant information about how this renewal of intersubjective agency is achieved by

a combination of talk that is exquisitely choreographed with embodied behavior.

In Fragment 2.7, S, B and W do further interactional work that focuses on the language of

spatial relationships.

INSERT FRAGMENT 2.7 ABOUT HERE

More specifically, in FG14, S continues to look in the direction of Group 2 as she says

“beside?” in line 62 with high rising intonation. Notice also that the horizontal, roughly chest

high gesture that is shown in this frame grab may be an iconic representation of this word.

Halfway through S’ turn in line 62, B overlaps S by saying “beside on the-” in line 63. This

turn is cut off abruptly by some simultaneous unintelligible talk in line 64. As shown in FG15, B

makes a low iconic gesture as she utters her talk in line 63. Notice however, that this gesture is

different from the one used by S in FG14. After a 0.3 second pause in line 65, B reproduces

what she said in line 63 almost exactly when she says “and yeah beside on the- (0.3).”

As B utters this turn, W simultaneously mimes the concept of beside (see FG16), although again,

this gesture is somewhat different from the one used by B in FG15.

In Fragment 2.8, S again pursues the question of where the crutch(es) are located,

although this time she chooses to recycle the phrase “on the floor.” which, as we have already

seen, was first introduced into the conversation by W in line 46 of Fragment 2.4.

INSERT FRAGMENT 2.8 ABOUT HERE

More specifically, as she utters this phrase in line 67, she simultaneously makes a

sweeping horizontal gesture at roughly chest height (see FG 17) that closely resembles the

20

gesture that she previously used in FG14 in Fragment 2.7. The phrase seems to generate some

positive feedback by Y, who quietly repeats “on the floor” in line 69. After a 0.4 second

pause, S turns back to the blackboard (see FG18) and begins to draw for the next five seconds.

Although we cannot at this stage see what S is drawing, Y and others (who have a direct line of

sight onto this activity) start to laugh approvingly in lines 71-75. Lastly, in line 76, S finishes off

her drawing work at the blackboard.

Finally, in Fragment 2.9, the rest of the class joins in the laughter as S steps aside from

her drawing, thus allowing the story tellers to evaluate how successful S has been in placing the

crutches in relation to the bed.

INSERT FRAGMENT 2.9 ABOUT HERE

More specifically, in line 77, W produces a laughter token that is achieved as an in-

breath, which, to my ear at least, does amused surprise. In lines 78-79, the rest of the class joins

in, and in FG19 and the simultaneously choreographed talk in lines 80-81, the teacher chooses

this as an opportune moment to gesture at Group 3 and to pass the next turn at talk to this group.

The picture that S has drawn is reproduced in Figure 3 (this picture is taken from the subsequent

debriefing phase of Task 2, when the teacher is commenting on the crutches-related work the

students have done), which confirms that, through their embodied interactional work shown in

Fragment 2, S has somehow successfully managed to work out what the word crutches means

and that these objects are located close to the patient’s bed.

Figure 3: Image showing the image and location of the crutches in S’s drawing

21

Let us now briefly look at what happens during the teacher’s debriefing in Fragment 3,

which occurs during the implementation of Task 2. More specifically in Fragment 3.1, the

teacher is providing contingently relevant feedback on various aspects of the language learning

behavior that he observed during the students’ production of the talk exhibited in Fragment 2.

His talk is delivered slowly, with a lot of emphasis given to important words in unfolding phrasal

and sentential TCUs, and these TCUs are set off from the rest of the talk by pauses which

bookend each unfolding unit (see, for example, “(0.8) where the crutches (0.2)5 are.

(0.6)”in lines 1-11).

INSERT FRAGMENT 3.1 ABOUT HERE

The same deliberate kind of delivery is observable in Fragment 3.2 (see for example, the

way the phrases “on the ground,” beside the bed” and “on the floor” are given prominence by the

same practices of prosodical marking of important words and by bookending each clause with

pauses in lines 11-15. In addition, this fragment is interesting because it demonstrates that

deploying intersubjective agency is not an action that only students do. Teachers also routinely

do this in a variety of ways. So, in addition to packaging his feedback in Fragments 3.1 and 3.2

through the artful use of prosody and pausing that is designed to highlight unfolding grammatical

structure, the teacher also talks about the lexical item crutches in a distinctively incidental way,

and embeds his feedback on this word within much more detailed feedback on the the students’

use of the language of spatial relationships. In other words, this evaluative commentary is

observably designed to be contingently relevant to the specifics of the students’ talk in Task 1.

INSERT FRAGMENT 3.2 ABOUT HERE

More specifically, when the teacher mentions the word crutches in lines 12-13, he treats

this word as a communicative problem that has already been solved, and which therefore

22

deserves little further commentary. We can see this in the way the phrase “>ookay. these

things,o<” in lines 12-13 is marked as an aside by its quicker delivery. In addition, the teacher’s

pointing gesture, the smiling face and the intense eye gaze directed at the students (see FG20),

which all invite the students to recognize this as work that has already been done, all gloss this as

an incidental piece of information. Notice also that the two iconic gestures in FGs 21 and 22 are

a crucial part of how the talk in lines 14-15 — which recycles the language of spatial

relationships previously used by the students in Task 1 — unfolds on a moment-by-moment

basis. More specifically, in FG21, the teacher mimes the concept of on as he says this word in

line 14. When he repeats the phrase on the ground in line 15, he again choreographs his

production of of the word on with another more emphatic gesture which is achieved as an

audible deictic thump on the table. In other words, in order to understand that the talk in line 15

is actually a first position repair, we have to have access to the embodied data contained in FGs

21 and 22.

The same kind of detailed multimodal work continues in Fragment 3.3.

INSERT FRAGMENT 3.3 ABOUT HERE

More specifically, we see the same kind of tightly knit deployment of talk and gesture in:

line 16 and FGs 23 and 24; line 17 and FGs 25 and 26, and line 19 and FG27. This talk not only

recycles the adverbial clauses originally used by the students in Task 1 (see lines 19-20 in

Fragment 2.2 for the source of the clause under the bed; lines 34 and 37 in Fragment 2.3 for the

source of the clause beside his/the bed; and lines 45-46 in Fragment 2.4 for the source of the

clause on the ground in Fragment 2.4), it also explicitly draws their attention to these clauses as

learning objects that are important resources for doing the language of spatial relationships.

Finally, in Fragment 3.4, the teacher goes beyond the language that was used by the students and

23

models the meaning of the new clause between the friend and the patient by deploying exactly

the same kind of interactional and embodied practices that he had used in his review of the

familiar language discussed previously.

INSERT FRAGMENT 3.4 ABOUT HERE

Finally, let me end this empirical section by showing how well each level of students did

in reproducing the picture shown in Figure 1. Figure 4 shows the final drawing produced by M

in the class composed of NSs/very advanced speakers of English. This drawing was judged by

the students themselves (who had to complete this picture within 15 minutes, with no more than

10 errors of detail) not to have met the criterion of graphic accuracy pre-specified in Task 1 of

the materials.

Figure 4: Final drawing by NSs/very advanced speakers of EnglishFigure 5 shows the final drawing produced by S in the low intermediate class. Although

this drawing does not include the speech bubble that connects the patient and the three pictures

that constitute the story of what happened to him, the students did meet the less demanding

criterion of getting the gist of the story within 25-30 minutes. For this reason, I would argue that

these students did meet the criterion of graphic accuracy pre-specified in Task 1.

24

Figure 5: Final drawing by early intermediate/late beginners

Finally, Figure 6 shows how C puts the finishing touches on her drawing — that is,

drawing the speech bubble that connects the patient to the story— just before the teacher called

time for Task 1.

INSERT FIGURE 6 ABOUT HERE

Figure 6: Final drawing by false beginners

More specifically, in this two camera recording set up, we can see the story drawer —

who is called C — drawing the speech bubble in the upper video around the material which

reproduces the three boxes at the top of the drawing in the original picture. Below this video is

the video of the story tellers, and we can see how the talk and embodied actions by I (the smiling

man at the extreme right of the lower video) enables C to complete Task 1. I has very little

English and has participated very little up to this moment in the interaction. However, in lines 2-

7, he tells C to draw a speech bubble, but he does not know how to say this in English. In lines 4

and 7, he therefore artfully deploys his intersubjective agency by making two speech bubble

gestures in the air in front of him, which C correctly interprets in line 15 as an instruction to draw

a speech bubble. Thus, at the very last possible moment, one of the linguistically weakest

students in this class is able to make a crucial contribution to the successful completion of Task 1

by this group of false beginners. Notice that, in this respect, the false beginner group is able to

outperform the low intermediate/late beginners group, who never included a speech bubble in

25

their drawing. So this observation, coupled with the fact that the students were able to complete

Task 1 within the time limit specified for this level of learners, warrants the conclusion that this

group also successfully completed Task 1, even though their language skills were quite low.

Finally, these observations have interesting implications for classroom assessment. First,

I would argue that, if students whose proficiency varies from native speakers/advanced users of

English to false beginners are all observably able to use the “same” materials as catalysts for

doing successful language learning behavior, at least in the short term, then we need to rethink

how useful the notion of levels is, particularly when this construct includes, as it must, an

interactional dimension. In this context, I believe that conversation analysts and language

assessment specialists have an opportunity to collaborate in meaningful ways by developing

testing specifications that are more precisely stated than the standards of graphic accuracy used

in conjunction with Task 1. And second, the data displayed in Figure 6 (specifically, the

participants’ smiling faces in the lower video, plus the laughter tokens in line 16) suggest that the

question of how laughter is contingently used as as a resource for doing self-assessment (whether

positive, as in these data, or negative, as in other data analyzed by Markee & Kunitz, 2015) may

be worth further empirical investigation.

26

CONCLUSIONS

In this paper, I have argued that the concept of interactional competence is difficult to reconcile

with a post-cognitive, ethnomethodological CA perspective, in that a great deal of the (quite

interesting) research that has been done on this construct is theoretically and methodologically

eclectic, and therefore violates CA’s stance of ethnomethodological indifference to a priori, etic

theory. For these reasons, I have suggested that we bracket the many theoretical accretions that

characterize our current understanding of interactional competence and return to the theoretically

sparer notion of competence, as this term has been understood by Schegloff and Sacks. Such a

move emphasizes the importance of action — and, therefore of intersubjective agency — and

divorces the traditionally developmental aspects of competence from its social, interactional

aspects.

In the empirical parts of the paper, I have provided a detailed analysis of how two levels

of curricular planning — materials design and methodological implementation — worked hand

in glove in the context of a demonstration lesson on task-based language teaching that was taught

to different groups of learners over a period of 27 years. More specifically, I have worked up a

post-cognitive account of classroom interaction produced by learners who span the beginner-

native speaker spectrum of proficiency in English. This analysis makes no appeal to any

cognitive notions of language or language learning. This account shows how these participants

were routinely able to solve complicated communication problems that were at times objectively

beyond their current levels of linguistic proficiency because they artfully deployed their

collective intersubjective agency to successfully do language learning behavior, at least in the

short term. These analyses have also consistently shown how important the process of

cannibalizing and recycling material from previous talk (whether by self or others) is to the

27

design of current and incipient turns, and provides further empirical support to the claim made by

Markee and Kunitz (2103: 657) that a great deal of classroom learning is “just an institutional

form of talk (emphasis in the original).

In terms of future research directions suggested by this paper, I believe that the curricular,

planning-oriented approach developed here may usefully be situated within the

ethnomethodological “studies of work” tradition outlined by Hester & Francis (2000) (see also

Lynch, 1999), in which the analyst’s task is to develop an account of how local educational order

is accomplished by participants in and through talk on a moment-by-moment basis. As I

envision it, future applied linguistic studies of classroom interaction that are grounded in this

ethnomethodological tradition will be concerned with showing how classroom interaction is

systematically organized as a series of overlapping, more or less institutional, speech exchange

systems that are designed to promote not just language learning behavior, but a host of other

contingently relevant social actions — for example, the need to maintain face (see Markee,

2004), identity work (see Kasper, 2004), or — that may actually get in the way of successful

language learning. From this perspective, instances of unsuccessful language learning behavior

may prove to be as interesting as examples of successful learning. Finally, this twin concern

with understanding both successful and unsuccessful language learning behavior will feed into

on-going attempts to use ethnomethodological CA as a source of input that may inform locally

grounded pedagogical interventions and, more generally, bottom up approaches to curricular

design and innovation (see Lindwall & Lymer, 2005; Markee, 1997).

NOTES

28

REFERENCES

Brouwer & Wagner, 2004; Coughlan & Duff, 1994)Duranti (2010).Edwards, (2006).Firth, 2009; Garfinkel & Sacks, 1970). Goodwin 2006). Hall, 1993, Hall 1995, Hall 1999Hall, Hellermann & Pekarek Doehler, [Eds.], (2011).Hall & Pekarek Doehler, 2011; Hauser, 2011). Hauser, 2016)Hellermann, (2006).Hellermann, (2007).Hellermann & Lee, 2014; Kasper, 2009; Lee, 2006). Lee & Hellermann, 2014; Lynch, 1999).Markee (1997).Markee, 2000). Markee, 2008, Markee (2011).Markee, 2015). Markee & Kunitz, 2013; Markee & Seo, 2009; Mori & Hasagawa, 2009Pekarek Doehler (2010)Pekarek Doehler, 2013Pekarek Doehler & Pochon-Berger, 2015Potter (2000).Psathas (Ed.) (1990)Schegloff, 2006).Suchman, 2007)Te Molder & Potter (Eds.), (2005).Thorne & Hellermann, 2015).Van Dijk, 2006)Young 2000).Young, (2009).Young (2010).Young (2011).Young & Miller, 2004)

29

1 For readers who are not familiar with the EM and CA literatures, bracketing is an example of methodological relativism. No judgment (whether good or bad) is made about the quality of preceding etic research that has been carried out on a particular phenomenon (in this case, the notion of competence). Rather, the aim of bracketing is to enable the development of an emic account of members’ own understandings of how they accomplish competence in real time. 2 In this context, Nguyen (2011) points out is that there are two different ways of doing longitudinal research in the CA literature on classroom interaction and learning. The dominant approach, represented by, I would claim, the work of John Hellermann and others, she calls a studying learning approach, while the kind of approach I am advocating here is a doing learning approach (which, according to Shintani & Ellis, 2014, has achieved at least some success in documenting successful language learning over time). So far so good. She further notes (p. 178) that “while I agree with Markee [2008] that longitudinal studies using CA to understand learning should not compromise CA’s methodological principles in any way, I find it methodologically problematic to draw a line between when a participant is orienting to an interactional resource as a learning object and when she/he is not” (emphasis in the original). She then goes on to point out that there are many instances when longitudinal learning occurs without a member’s orientation to learning (through, for example, an observable repair sequence) being present. I have no quarrel with this last observation. However, I would counter that distinguishing between when participants do or do not orient to some aspect of talk as being an instance of language learning behavior is exactly what makes the doing learning approach an ethnomethodological project. More specifically, what I find methodologically problematic from a purist, post-cognitive CA perspective about the studying learning approach is that this latter approach has, in addition to importing exogenous theories of SLA to frame its research program, also unproblematically imported longitudinal research methods from applied linguistics which divorce language learning from its relevant sequential context(s). For example, the studying learning approach typically looks at different developmental levels of learning that are observable during different, arbitrarily chosen moments in time. But this approach is unable to show how learner-participants get from Time A, through Time B, to Time C. On the other hand, while a doing learning approach is able to do this very well in the relatively short to medium term, I acknowledge that it is less well equipped than a studying learning approach to show what different levels of linguistic attainment look like over the long term.3 If W is the speaker in line 12 of Fragment 2.1, the TCU “under the bed” in line 19 of Fragment 2.2 may be a self-initiated, self-completed repair by W of her previous attempt to say this phrase, which came out as “(uhntoo (0.3) uhntoo l/reg.)” 4 The teacher’s role in these materials is designedly non-interventionist. The teacher’s principal job in Task 1 is to make sure that the students keep to the time limits that have been set for their level, and to keep notes of the problems the students have with grammar, vocabulary, pronunciation, etc., as a resource for subsequent intervention in Task 2. Note that the teacher does not pre-teach vocabulary or grammar, nor does s/he normally intervene in Task 1 in any way, especially if the students ask for his/her help (for example, with grammar and vocabulary), as this would wreck the information gap which is foundational to this task’s design. This non-interventionist way of managing Task 1 forces participants to complete the task as best as they can with their current linguistic resources.5 This 0.2 second pause does not delimit a TCU; rather, it occurs at point of maximal grammatical control. By doing this, the teacher is able to give maximum emphasis to the word crutches, to whose understanding, as we have seen, the students allocated a significant amount of time and effort.