Conversation Analysis and its Relevance to Language education

A Webinar Series
May 7, 14, 21, 28, 2022
(Saturdays, 9:00am -12:00pm)
Center for Research on Bilingual Education
National Taipei University of Education
Dr. Donald Carroll, Shikoku Gakuin University, Japan
abufletcher2@gmail.com
Conversation Analysis &
its Relevance to
Language Education

An introduction to
Conversation Analysis

What CA is…
…a fundamentally different way of looking at and explicating
human interaction.
…a robust (and rigorous) research methodology that looks more
deeply at the fine-grained details of life than any other
perspective.
…an extensive body of empirical research, now well into its sixth
decade, into the observable micro-practices of talk-in-interaction.
…a window out onto an entirely different view of the nature of
language and, therefore, how language might be taught.
…the relevant “body of knowledge” on which the teaching of
spoken interaction should be based.

What CA isn't…
…It isn’t a set of “rules for having good conversation” or “rules
for how to speak politely.” It's not rules at all.
…It isn’t just another name for “discourse analysis” though the
two research fields/paradigms share many common goals.
…It isn’t just recording some talk, then writing it down, which is
to say, it’s not reducible to a simple series of mechanical or
technological steps.
…It isn’t a language teaching methodology, akin to, say, the
“communicative approach.”

Analysis of
Conversational Data
(ACD)
vs.
Ethnomethodological CA
(SSJ CA)
Sacks, Schegloff, and Jefferson (1974)

Language as
social interaction

Conversation Analysis and its Relevance to Language education

Domains of interactional organization
 Turn-taking
How do participants take turns in conversation?
 Turn-construction
How are turns-at-talk structured, produced, and perceived?
 Sequence organization
How are single turns-at-talk organized into sequences of turns?
 Preference organization
How do participants in talk-in-interaction manage social challenges?
 Practices of repair
How are threats to intersubjectivity handled in interaction?
 Action formation
How do conversational actions come to be recognized?

CA has examined real-
world, situated interaction
at a resolution previously
unimaginable.
…from close looking at the
world we can find things that
we could not, by imagination,
assert were there.
(Sacks, 1984:25)

"…from close looking at
the world we can find
things that we could not,
by imagination, assert
were there."
(Sacks, 1984:25)

Sample of CA transcription
…from close looking at the
world we can find things that
we could not, by imagination,
assert were there.
(Sacks, 1984:25)

Turn-taking
in talk-in-interaction

In any conversation, people
take turns talking. Otherwise,
it wouldn’t be a conversation.
How is turn-taking organized?

Pre-allocated turn-taking
In bowling everyone agrees before the game begins who will
bowl first, second, third, and so on. This sort of turn-taking
is called "pre-allocated."
For some types of social events the speaking order is
decided in advance or follows a strict set of established
rules. Examples would be a graduation ceremony, a formal
debate, and courtroom proceedings.
In some cultures, in some (formal) settings, there might also
be societal rules for pre-allocation of turns-at-talk, for
example, based on seniority or social rank.

Turn-taking systems
There is also the sort of turn-taking common to classrooms,
in which the teacher manages who speaks, either by calling
on a particular student or allowing students to "nominate"
themselves by raising a hand, after which the teacher
selects one of them.
However, most talk does not follow a system of pre-allocated
turn-taking. This includes most spoken interaction (i.e. "talk-
in-interaction") ranging from casual conversation to the
myriad varieties of talk-at-work (and even in classrooms).
On these occasions, we say that the turn-taking is "locally
managed" turn by turn. It is not decided in advance.
But how is this accomplished?

The signal model
In some types of communication a signal is used to mark the
end of a speaking turn. This is the case with walkie-talkies
and CB-type radios. Both operate on a push-to-talk basis.
A: The operation begins at 0600. Over. ((releases button))
B: Affirmative. Over. ((releases button))
A: Good luck. Over and out. ((releases button, shuts off))
One proposal for a system of conversational turn-taking is
the so-called "signals" model, in which the ends of speaking
turns are "marked" via various linguistic features, including
silence. According to this model, other participants wait for
these signals, which indicate that the current speaking is
going to "yield the floor" to another participant (Duncan,
1974; Duncan and Fiske, 1977).

CA and the SSJ system
A different description of conversational turn-taking, which
more closely matches observations of recorded natural-
occurring interaction, was laid out by Harvey Sacks,
Emanuel Schegloff and Gail Jefferson (1974) in their now-
classic CA paper: “A simplest systematics for the
organization of turn-taking for conversation.”
The SSJ turn-taking system better accounts for 1) the
empirical fact that next speakers are able to begin speaking
with little or no-gap between turns, 2) the common
occurrence around points of possible transition of natural
overlap, 3) not just of how transition happens but also how a
given next speaker comes to have rights to the turn.

Some resources for TRP projection*
• Syntax
• Prosody
• Pragmatics
• Gaze and other embodiment
• Habitual phrasing (?)**
A: Well if you knew my argument why did you bother=
=to a:[sk
B: [because I'd like to defend my argument.
*The subtle, but important, distinction is that in the SSJ system, possible
next speakers are projecting possible NOT ACTUAL completion.
**See Jefferson 1986 on "recognitional onset."

CA and the SSJ system
1a. Current speaker may, but need not, select next speaker.
1b. If current speaker does not select next speaker, a next
speaker may, but need not, self-select by speaking first.*
1c. If no other participant self-selects, current speaker may,
but need not, speak again. This can also be worded as
Current speaker may, but need not, continue speaking.
2. Rules 1a, 1b, and 1c continue to apply in this order at
every possible transition relevance place until speaker
transfer occurs.
*Research suggests that participants may also make claims on next speakership
via various forms of embodiment.

Let's try out Rule 1b:
Self-selection
1) THE COUNTING GAME
2) I LIKE…
3) I WANNA GO TO…

A new TCU vs. an increment
1c Current speaker may speak again.
J: I went shopping yesterday
(.)
I was looking for some hiking boots.
(.)
Some friends are going to Nagano next month.
1c Current speaker my continue speaking. (Adding
increments)
J: I went shopping yesterday.
(.)
to look for some hiking boots.
(.)
Since some friends are going to Nagano next month.

Adjacency pairs
One mundanely observable feature of talk-in-interaction is that
prior turns often place limitations on the sorts of turns that might
expectably occur next. This relationship is broadly referred to as
conditional relevance.
Harvey Sacks, one of the founders of Conversation Analysis,
coined the term adjacency pairs (AP) in the late 60's for pairs of
turns, which are regularly found (in recorded talk) next to each
other. The two halves of an AP are referred to as the first-pair
part (1PP) and the second-pair part (2PP). Sacks et al.
(1974:716-8) summed up the operation of an AP as follows:
"A basic rule of adjacency pair operation is: given the recognizable
production of a first pair part, on its first possible completion its
speaker should stop and a next speaker should start and produce
a second pair part from the pair type of which the first is
recognizably a member."

Adjacency pair as normative expectation
There are no hard and fast rules in conversation that participants MUST
follow…only normative expectations. Following a greeting one can
expect a greeting in return. This is such a strong expectation that, if no
greeting is forthcoming, participants, can treat it as officially missing and
search for a reason. Was the greeting not heard? Was the person
otherwise engaged? Or was this an outright refusal to do a
greeting…and thus initiate a conversation?
Participants also use the normative expectations of adjacency pair
organization to interpret the next turn. A question (i.e. a request for
information) sets up an expectation for an answer. As such whatever talk
that occurs in the slot following a question will be inspected as a possible
answer. Thus, an adjacency pair serves as an interpretive frame.
A: How late does the hardware store stay open?
B: It's Monday.

Common adjacency pairs (Levinson 1983:303)
1PP greeting  2PP greeting
A: Hi
B: Hi
1PP question  2PP answer
A: What time does the party start?
B: I think around eight.
1PP summons 2PP reply
A: Excuse me
B: Yes?
* ((phone rings))
B: Hello:?*
*Note that “Hello” at the beginning of a telephone call is not a 1PP greeting, but a
2PP reply to the ringing of the telephone. This can be made clear in dialogues of
telephone calls by including ((ring, ring)) in the preceding line (Wong, 2002).

1PP offer/invitation  2PP acceptance
 2PP declination
A: Wanna lift?=
B: =You saved my life.
A: Have a cigarette.
(0.3)
B: um...I don't smoke thanks.
1PP request  2PP grant
 2PP refusal
A: Can I borrow your car?=
B: =Sure.
A: Can I borrow your car?
(0.4)
B: Well...I dunno…

1PP assessment  2PP agreement
 2PP disagreement
A: Isn’t he cute
B: O::h he::s a::DORable!*
A: This is interesting!
B: Yeah, really interesting!*
*This type of agreement is called “upgrade” (see Pomerantz 1984
A: This is GREAT!
B: Yeah, it’s ok**
**This type of “downgrade” agreement regularly foreshadows upcoming
disagreement.
*** Also note that "I agree" and "I think so too," which are both taught in
EFL texts, are heard as "same assessments" which also regularly
prefigure disagreement.

Discussion
In the following example, B orients to A's prior turn as an accusation and
issues a denial. What made A's turn hearable (to B) as an accusation?
A: You left the lights on.
B: It wasn't me.
Again B orients to A's turn as an accusation, but this time offers an apology.
What made A's turn hearable (to B) as an accusation?
A: It's half past six.
B: Sorry (we had an unexpected meeting)
Note: In doing CA it is not up to the analyst to decide what a given turn-at-
talk is doing. What matters is how participants orient to that talk.

1PP accusation  2PP denial
 2PP apology
A: You left the lights on.
B: It wasn't me.
A: It's half past six.
B: Sorry (we had an unexpected meeting)
1PP apology  2PP minimalization
A: Please excuse Dan
B: It’s nothing
Many of these adjacency pairs seem to be "universal" in the sense that
it's hard to imagine a culture in which a greeting wouldn't make relevant a
greeting or a request wouldn't make relevant either granting or denying.
However, the resources and manner for doing these actions may vary
significantly from language to language and culture to culture.

Complaint sequence (Dersley & Wootton, 2000)

1PP compliment  2PP minimalization
 2PP denial
 2PP acceptance
 2PP return compliment
A: This is delicious. You’re really a great cook!
B: My mother did most of the cooking.
A: You're a wonderful cook!
B: No, no, no, I'm useless in the kitchen!
B: Thank you.
B: So are you.
Unfortunately much that has been written about adjacency pairs in the pragmatics
and TESOL literature is distorted or misleading. They have sometimes been
mischaracterized as conversational “boilerplate” or “fixed routines” or
“conversational gambits.”

Adjacency pair expansion*
A base adjacency pair can be expanding in three locations:
Pre-expansion, insert-expansion, and post-expansion.
Each of these expansions is a sequence in its own right and
there can be multiple expansions in each position.
Schegloff (1990) analyzes one single action sequence,
which extends for over 100 lines of the transcript, which
turns out to be structured around one single adjacency pair,
namely, a request and its eventual granting.
*Schegloff (2007) spends almost 150 pages (p.22-168) on
the the nuts and bolts of sequence expansion. Sadly, I can
only provide a minimal overview here. 

Pre-expansion
The use of one or more pre-expansion sequences is done to check on
the likelihood that a possibly upcoming action will receive a preferred
response. If not, the base adjacency pair may not happen at all.
"…we should appreciated that pre-sequences or pre-expansion are
measures undertaken by the speaker – or the prospective speaker –
of a base first pair part to maximize the occurrence of a sequence
with a preferred second pair part." (Schegloff, 2007:81)
In terms of distribution, we empirically find more acceptances, grantings,
and agreements, not because people want to do these actions more
frequently, but precisely because considerable interactional work has
been done to avoid getting a dispreferred response.
Three common types of pre-sequences are pre-invitations, pre-requests,
and pre-announcements.

Not "pre" as in "before"
but
"pre" as in preliminary to.

Pre-invitations  Go-aheads
In the following examples a pre-invitation gets a go-ahead response that
acknowledges the possibility of an implied upcoming invitation. A go-ahead
is a preferred next action. In doing a go-ahead, speaker B indicates a
positive disposition towards a possible invitation.
A: Are you busy on Friday?
B: No, not really. (Not at all, No, I’m just…) GO-AHEAD
A: What are you doing around 8pm?
B: Nothing much. GO-AHEAD
A: Do you like Irish music?
B: Yeah, I love it! GO-AHEAD

Pre-invitations  Blocks
Conversely, speaker B may choose to block the potential invitation which
may result in the invitation not being done at all or being approached in
an alternative, less direct, format, e.g. “I heard there’s going to be live
Irish music on Friday and I was thinking of checking it out.” If speaker B
displays interest upon hearing this, an invitation to “join me” might then
be forthcoming. A blocking move is a dispreferred action and is often
packaged using the dispreferred turn shape.
(…)
B: Um…I’ve got some stuff I sort of have to do. BLOCK
A: Do you like Irish music?
(…)
B: iyeah…not so much. BLOCK
*Note: English learners regularly fail to hear these sorts of turns as pre-
invitations and instead orient to them as “questions,” to which they regularly
provide an “answer,” which might be taken as a blocking move.

Pre-invitations  Hedges
Another type of possible response to a pre-invitation is to adopt a
tentative orientation towards a possible upcoming invitation suggesting
that possible acceptance is contingent on the specifics of the invitation.
(…)
B: Just doing some odds and ends, why?
Following a hedged response the producer of the pre-invitation might
instead take the safer route of mentioning plans, safer in the sense that
"talk about plans" isn't subject to "rejection." If these plans are
enthusiastically received, an invitation may follow.
B: Not particularly, why?
A: A couple of us are thinking of having a BBQ at the beach…
B: That sounds like a lot of fun!
A: We'd love to have you join us.

Pre-offers  Go ahead / Block
Offers are similar to invitations. In fact, Schegloff (2007:35) suggests
that invitations may be seen as a sub-class of offers. As such we also
find pre-offers, which seek to see if the offer will be welcome.
Cat: I'm gonna buy a thermometer though [because I=
Les: [But-
Cat: =think she's [(got a temperature).
Les: [We have a thermometer.
Cat: (Yih do?)
Les: Wanta use it?
Cat: Yeah.
Pet: I'll see you Tuesday.
Mar: Right.
Pet: O[k a y Marcus ]
Mar: [You- you're al]right [you can get there.
Pet: [Ye-
Pet: Yeah
Mar: Okay ((<<- No offer is forthcoming))

Pre-requests: Go-aheads and blocking moves
Like invitations, requests are often prefaced with one or more pre-
request sequences. Again the pre-request seeks to check on the
likelihood of a preference response: granting.
Common types of pre-requests are “Are you free…” “Do you have…” “Do
you know how to…” and so on.
Even many seemingly “neutral” questions might actually be heard as
possible pre-requests, e.g. “Are you going into town this afternoon?”
"Have you ever fixed a computer?" Again, the pre-request can receive a
go-ahead or a block.
A: Do you have a hammer?
B: Sure.
A: Do you have a hammer?
(…)
B: iyeah, but…I’m not sure where it is at the moment.

Pre-announcements (or pre-telling)
This type of sequence is used with news announcements (i.e. "telling
news") both to check whether the news is really new (to this participant)
and for co-participants to display their acceptance of the role of news-
recipient, sometimes even when subsequent talk reveals that they do, in
fact, already know the news.
A: Guess what.
B: What.
A: Guess how much I paid for this?
B: How much.
A: Did you hear what happened at work yesterday?
B: No, what.
A: Hey, I got sump'n thet's wild. (We got good/bad news)
B: What.

Pre-announcements  Blocking / Preempting
An possible announcement might be blocked by the recipient displaying
that the news isn't news.
A: Did you hear what happened last night?
B: Yeah, shocking isn't it.

Is that a question or a pre-announcement?
Below is a classic, and particularly interesting, instance of how
participants might orient to a pre-announcement. It's also an illustration
of what conversation analysts call the "next turn proof procedure."
In line 2, Russ orients to Mother's prior turn as a pre-announcement,
accepting the role of news-recipient. However, in line 3 Mother rejects
this understanding and makes it clear that she meant her turn as an
actual question. In lines 4-7 Russ shows that he, in fact, already has a
very good idea regarding who will be at the meeting and that this really
wouldn't have been "news" to him at all.
01 Mother: Do you know who’s going to that meeting?
02 -> Russ: Who.
03 -> Mother: I don’t kno:w
04 Russ: Oh::, Prob’ly Missiz McOwen (‘n detsa) en prob’ly=
05 =Missiz Cadry and some of the teachers
06 (0.4)
07 Russ: And the counsellers

Pre-pre or action projections
Another type of pre-sequence is the pre-pre so named because it often
precedes other pre-sequences (Schelgoff, 2007:47). Common pre-pres
are “Could you do me a favor?” “Can I ask you a question?” or
formulations with "Let me…" In many cases the projected action (e.g. the
favor or the question) is not done in the next turn…or ultimately at all.
A: Could you do me a favor?
B: Sure.
A: Well, you know that I’ve been having problems with my car,
like the engine not starting. Well, it happened again this
morning
and I’ve got this really important meeting to get to this
afternoon,
but the garage guy says the car isn’t going to be ready until
maybe the day after tomorrow…
This extended turn seems aimed at getting B to making a pre-emptive
offer of assistance. Requests are unusual in that they are a dispreferred
first pair part. Another type of interactional work that a pre-pre can do is
to mark some upcoming favor or question as “delicate.”

Practicing a pre-pre
sequence
A: Could you do me a favor?
B: Sure.
A: [EXTENDED TELLING SEEKING PREEMPTIVE OFFER]
B: Preemptive offer
..

Insert-expansion
Quite often the talk that occurs in the next turn after a first-pair part is
hearably not a second-pair part. For example, a request might be
followed by a question asking for more details. (pre-second)
A: How long does it take to get there? 1PP
B: Are you going by car? 1PP
A: Yeah. 2PP
B: About 40 minutes. 2PP
However, insert-expansion can be heard as implicative of an upcoming
dispreferred action. This follows the general principle that anything that
occurs in the location where a preferred action should have occurred (i.e.
immediately) displays possible disaffiliation. (post-first)
A: Do you want to have Thai food for dinner?
B: What? REPAIR
A: Would you rather have something else?

Insert-expansion (post-first vs. pre-second)
"Whereas post-first inserts look backward, ostensibly to clarity the talk of
the first pair part, pre-second inserts look forward." Schegloff, 2007:106)
Consider the insert expansions in this 911 call.
A: Send an emergency to fourteen forty-eight Lillian Lane. 1PP
B: Alright sir. We'll have them out there. 2PP
Cal: 1PP Send 'n emergency to fourteen forty eight Lillian Lane.
Pol: PF Fourteen forty eight – [What sir?
Cal: [yesh.
Pol: PF Li[llian Lane?
Cal: [Fourteen forty eight Lillian.
Pol: PF Lillian.
Cal: Yeah.
Pol: PS What's th' trouble sir.
Cal: Well. I had the police out here once, Now my wife's got cut.
Pol: 2PP Alright sir. We'll have 'em out there.
Cal: Right away?
Pol: Alright sir.

Post-expansion
In adjacency pairs which receive a preferred second pair part, such as
accepting, granting, agreeing, the sequence typically ends and the talk
moves on, or the post-expansion is minimal.
A: Could I borrow your camera?
B: Sure, no problem.
A: Thanks. ( sequence closing third)
However, in the case of a dispreferred response, the sequence may be
re-opened in pursuit of the preferred response.*
A: Could I borrow your camera?
(…)
B: Um…I dunno…I sort of hate to lend it out to anyone…
A: I'll take good care of it and get it back to you tomorrow right.
*Schegloff (2007) provides a full discussion of the many subtle aspects and
specific practices involved in post-expansion.

Why sequence organization matters
“…a great deal of talk-in-interaction – perhaps most of it – is better
examined with respect to action than with respect to topicality, more for
what it is doing that for what it is about.” (Schegloff, 2007:1)
The primary context for understanding what is being said and done in the
current turn-at-talk is what was said and done in prior turns, often the just
prior turn.
Recognizing the type of sequence that is underway is of crucial
importance to producing a relevant next turn-at-talk.
Sequence organization lies at the core of a “grammar for interaction.”
Sequence organization has been almost wholly ignored in ESL/EFL
teaching materials.
The structure of sequences is intertwined with preference organization,
which is a primary resource for managing human sociality.

Expand the sequence
Re-imagine the following base pair (invite – accept/reject)
with one or more pre-expansions before K's turn, one or
more insert expansions after K's turn, and one or more
post expansions after M's turn.
1PP: Invite
2PP: Accept or Reject.
Feel free to include silences, delay, and other hedges
where you feel they might have occurred.

Talk-in-interaction syllabus (Part 1)
I. Opening a conversation
a. Face-to-face
b. One the phone
1. Summons-reply
2. Greetings AP
3. "Howareya" sequences
4. Stating the reason for the call (or proposing a first topic)
II. Closing a conversation
a. Pre-closing moves
1. "Well, I've got to be going now."
2. "Well, it's been nice talking with you."
3. Making plans for the future ("see ya tomorrow in class")
b. Closing

III. invitation/offer sequences
a. Making simple invitations
b. Accepting an invitation using a "preferred turn shape"
c. Rejecting an invitation using a "dispreferred turn shape"
d. Using pre-invitations to avoid rejections
1. Go-ahead" moves following pre-invitations
2. Blocking" moves following pre-invitations
3. Responding to blocking moves
IV. Request sequences
a. Making simple requests
b. Granting a request using a "preferred turn shape"
c. Refusing a request using a "dispreferred turn shape"
d. Using pre-requests to avoid refusals
1. Go-ahead moves following pre-requests
2. Preemptive offers following pre-requests
3. Blocking moves following pre-requests

V. Assessment sequences
a. Making simple assessments ("This's great!" “What a terrible movie.”)
b. Agreeing with assessments (preferred turn shape)
1. Upgrade assessments
2. Same evaluations
c. Disagreeing with assessments (dispreferred turn shape)
1. "weak agreements" + "weak disagreement" ("Yeah, it was
but...")
2. Downgrade assessments (“fantastic”  “OK”)
d. Using "tag questions" in assessments to solicit participation
e. Compliments (a special type of assessment)
1. Acceptance (Note: “Thank you” is a sequence closing third!)
2. Denial
3. Minimalization

VI. Telling news
a. Using pre-announcements ("Guess what?")
b. Responding to pre-announcements
c. Telling news
d. Responding to news
1. Oh + assessment
2. Really? (prompting further news-telling)
e. Using “How’s is going” as topic solicitation
1. OK/fine/good = no news to tell
2. Great!/Terrible! = news to tell

"Grammar, of course,
is the model of closely
ordered, routinely
observable social
activities."
(Sacks, 1992:31[Fall 1964, lecture 4])

Preference organization
There is a form of social organization, a grammar of sorts, which shapes
many aspects of talk-in-interaction. Conversation analysts call it
preference organization. It intersects with the concept of "face" but goes
further by specifying the practices involve.
Lerner, G. (1996). Finding "face" in the preference structures of talk-in-
interaction. Social Psychology Quarterly, December.
The organization of preference(s) affects how we tailor our talk for
particular recipients (recipient design). It influences how we describe
persons, places, and events, how we construct invitations as well as how
we respond to them, how requests are designed and how we shape the
responses, how we agree and disagree, how we ask questions and how
we answer them…and much more.
In short, it cuts across many of the bread-and-butter areas of language
instruction. Yet it is either entirely absent from language teaching
materials or only accidentally found usually in distorted form.

CA preference ≠ Personal preference
The concept of preference in CA is not about "preferring" this or that in
the everyday sense. It's not about "likes" and "dislikes" as in "I prefer
dogs to cats." It is not a personal or psychological preference. It is a
shared social preference grounded in observable practices.
Preference refers to a systematic (structural) and asymmetrical
relationship between alterative descriptions or actions. It is deeply
rooted in the norms, and reflected in the practices, of social interaction.
The CA terms preferred and dispreferred are somewhat similar to
unmarked vs. marked in linguistics. Broadly stated, preferred actions
reflect a no-problem orientation, while dispreferred actions suggest a
problematic stance.
In question formation, turns can be formatted such that they "build-in" a
bias towards a certain type of response. More on that tomorrow!

Preference for persons, places, and events
Some of the earliest considerations into what came to be called
preference emerged from the general principle of recipient design.
"If possible, select a description that you know the other knows." (Sacks,
1992: II:148)
In terms of person reference, there is an observed preference for a
recognitional ("John") vs. non-recognitional reference ("this guy I know").
An answer to the question "where do you live" may be selected from
among an open set of possible responses, including: the US, California,
Orange County, Orange, close to Disneyland, about a hour outside of
LA, just across the street from Sally, etc.
Events, to which one may be invited, can also be characterized in
multiple ways (dinner, drinks, coffee, watch a game, hang out, etc.).
Drew (1992, expanding on Sacks' earlier discussion of this in his
lectures) argued that these descriptions are ordered along what he
called the principle of "maximal property of descriptions."

Preference in adjacency pair organization
In the case of a 1PP greeting, a 1PP summons, or a 1PP question, there
is only one relevant 2PP.
Other 1PPs, such as invitations, offers, requests, assessments,
compliments, etc. have two or more relevant 2PPs. For example, an
invitation makes relevant either an acceptance or a rejection.
Yet, these alternative actions are not symmetrical. They are not treated
equally by participants. One of these relevant actions is oriented to as
systematically preferred, i.e. as the preferred next action. The other is a
dispreferred next action and this orientation is displayed ("marked") via a
set of interactional practices.
In the case of an invitation, acceptance is the preferred next action and
rejection is the dispreferred next action. Once again, we are not talking
about personal preference. You may strongly wish to reject the invitation,
but you will still structure your rejection along the lines laid out by
preference organization.

Preferred turn shape
Preferred next actions are regularly formatted ("packaged") in
characteristic ways (see Sacks, 1987; Pomerantz, 1984; Schegloff,
2007; Pomerantz and Heritage, 2013).
This is most immediately seen in the observed fact that the preferred
action is typically done a soon as possible. This orientation manifests
itself in two ways: 1) Next speaker begins speaking with little or NO GAP
and 2) the preferred action is done as early as possibly in the turn.
A: Do you want to go to a movie?
B: Sure, sounds fun. (Acceptance coupled with positive assessment)*
In this example, there is no gap whatsoever between the end of "movie"
and the onset of "sure." (Note that there is also no hesitation in the
production of "sure" as there is in "s:::ur::e) This single word, placed
precisely where it is, does the work of accepting the invitation.
*Compare with the stilted “yes, I do” short-answer format.

The importance of timing
1) Counting in circles
2) Do you want to play
tennis? Sure.

Dispreferred turn shape
In contrast, dispreferred next actions (e.g. rejecting, refusing,
disagreeing, and in general disaffiliating) are regularly packaged using a
well-documented…and teachable…set of practices.
These practices can be broadly described as delaying the doing of the
dispreferred action, and thereby providing warning.
As opposed to the "on-time" start of preferred actions, dispreferred
actions are regularly preceded by a silence. Even a micro-silence can
suggest an upcoming dispreferred response, if it is noticeable to
participants.
Dispreferred turns (and the dispreferred sequences that often turn into)
typically include pauses, hesitation tokens (um, er, well, I dunno), hedges
(sort of, kind of, a bit, etc.) repetitions, excuses, questions, and/or
apologies.
In many cases, they entirely avoid “saying no” (see Carroll, 2011a).

Dispreferred turn shape
In the following example, a rejection gets postponed using a range of
practices, which both delay the dispreferred action within the turn as well
as push it “downward” within the sequence.
When the dispreferred action finally does appear, it is as the very last
word (and the very last resort) in the turn and the sequence. Ideally it
would not have occurred at all, if the inviter had backed down, for
example, by saying "Well, maybe some other time."
A: Would you like to go to a movie?
(0.2)
B: Um::..iyeah…I dunno…a movie?
A: Uh-HUH.
(0.3)
B: .hhh when?
A: On Saturday.
B: Iyeah…I dunno...I sorta gotta do some stuff…
so…um...I guess not.

Reacting to warnings
Quite often foreshadowing a possible upcoming (though not yet
produced) dispreferred action is enough to cause the prior speaker to
backdown or modify what was said.
B: . . . An' that's not an awful lotta frutcake
(1.0)
B: Course it is. A little bit goes a long way.
A: Well that's right
In the following example, Gene responds to the potential problem (as
displayed by Mary) by proposing a candidate reason himself ( i.e. you're
not available because you're working nights).
Gene: Whenih you uh: what nights'r you available.
(0.4)
Mary: .k.hhhhh[hh
Gene: [Are you workin' nights et all'r anything?
Mary: I do: I work, hhh a number o:f nights Gene. . .

Dispreferred sequences
Note that not only are there dispreferred next actions, there are also
dispreferred sequences. Schegloff (2007:82) identifies (at least) three
types of preference in sequences:
1)There appears to be a preference for offer sequences over request
sequences.
2)There appears to be a preference for noticing by others over
announcement-by-self.
3)There appears to be a preference for a party to be recognized (if
possible or relevant) over self-identification.
Participants often work to avoid doing a request at all, for example, by
telling trouble instead of formulating a request, which regularly occasions
a pre-emptive offer of assistance (see Carroll, 2012 for how this might be
taught). If a request is made (and the speaker feels lower entitlement), it
may be formatted using the characteristic shape of dispreferred turns.

Observations on requests (Schegloff, 2007:83)
1)Requests appear disproportionately to occur late in conversations…and seem
especially problematic and unlikely to occur in first topic position.
2)Requests are regularly accompanied by accounts, mitigations, candidate
"excuses" for the recipient, etc. which are done in advance of the request itself…
3)As with other "sensitive" or "delicate" conversational practices or actions, the
occurrence of one may "license" the occurrence of others.
4)One evidence of the treatment of requests as dispreferred is the masking of
them as other actions, often as ostensible offers…
5)The preferred response to a pre-request is not a go-ahead; this simply
forwards the sequence to a first part part – a request – which is the less
preferred way of implementing the project at hand. . . The preferred response…is
to pre-empt the need for a request altogether by offering that which is to be
requested.

Preference and adjacency pair expansion
An orientation to preference organization often lies behind the sort of AP
expansions that we looked at in the last session. Pre-expansions, (pre-
sequences) are aimed at securing a preferred response.
"…we should appreciate that pre-sequences or pre-expansion are
measures undertaken by the speaker – or the prospective speaker –
of a base first pair part to maximize the occurrence of a sequence
with a preferred second pair part." (Schegloff, 2007:81)
Insert expansion, in the form of repair (in particular "other-initiated"
repair), regularly foreshadows a dispreferred action, such as the
rejection of an invitation or refusal of a request.
Post-expansion seeks to keep an AP going after the provision of a 2PP
particularly when that 2PP is a dispreferred action. In other words, post-
expansion can "pursue" a preferred 2PP.

Preference organization:
Agreeing and disagreeing

Assessments and second assessments
Pomerantz (1984) examined assessments in conversation. Her findings
are worth summarizing here. As an initial observation, Pomerantz
noticed that assessments (“opinions”) are regularly followed by second
assessments as in the following instances (ibid.:59-61).
This may seem obvious, but EFL learners are frequently stumped for a
response when presented with an assessment.
J T’s- tsuh beautiful day out isn’t it?
L Yeh it’s jus’ gorgeous...
J: It’s really a clear lake, isn’t it?
R: It’s wonderful
B: Isn’t he cute
A: O::h he::s a::DORable
C: ...She was a nice lady--I liked her
G: I liked her too

“In general, dispreferred-action turn organization serves
as a resource to avoid or reduce the occurrences of
overly stated instances of an action. The preference
structure that has just been discussed - agreement
preferred, disagreement dispreferred - is the one in effect
and operative for the vast majority of assessment pairs.
Put another way, across different situations, conversants
orient to agreeing with one another as comfortable,
supportive, reinforcing, perhaps as being sociable and as
showing they are like-minded…
…likewise, across a variety of situations conversants
orient to their disagreeing with one another as
uncomfortable, unpleasant, difficult, risking threat, insult,
or offense.”
Pomerantz (1984:77)

Agreeing or disagreeing
The second assessment can either agree or disagree with the first
assessment. In most (but not all) situations agreement is the preferred
next action and disagreement is the dispreferred next action.
Pomerantz also noticed that preferred next action turns shared common
characteristics. For example, they are frequently produced immediately
after the 1PP assessment. Also these turns are typically fairly short and
direct. Disagreeing turns also shared common characteristics. They are
typically delayed and longer (or involve more turns) and regularly include
pauses, questioning repetitions, requests for clarification, repair initiators
(e.g. “What?”, “Huh?”), etc.
As Pomerantz (1984:70) states: “A substantial number of such
disagreements are produced with stated disagreement components
delayed or withheld from early positioning within turns and sequences.”

Three types of agreeing assessments
Pomerantz identified three main types of “agreeing” second
assessments in her data:
1) upgrades
****************************
2) same evaluations
3) downgrades
While all three are presented as agreement, Pomerantz found that same
evaluations and downgrades often prefaced actual disagreement and
therefore should be considered “weak agreement” forms, which
foreshadow possible disagreement.

1) upgraded assessments (upgrades)
Upgrades are uttered or worded in a way that is stronger than the first
assessment and, as such, display strong agreement.
a) The second assessment uses a stronger evaluative term.
A: Isn’t he cute
B: O::h he::s a::DORable
J: T’s- tsuh beautiful day out isn’t it?
L: Yeh it’s just gorgeous...
b) An intensifier is used (very, really, extremely, awfully…).
M: You must admit it was fun the night we=
=went [down
J: [it was great fun...
B: She seems like a nice little [lady
A: [awfully nice little=
=person

Upgrade activity #1
humungous
gigantic
enormous
huge
big

Upgrade activity #2
Itsy-bitsy, teeny-weenie
Itsy-bitsy
Teeny
Tiny
small

2) same evaluations
With same evaluations, the second assessment typically matches the
intensity of the first assessment.
C: ...She was a nice lady — I liked her
G: I liked her too
A: He’s terrific
B: He is.
Same evaluation can regularly be found in agreement turns, but they
also frequently preface disagreement and should therefore be
considered weak agreements.
J: Oh it’s [terribly depressing.
L: [oh it’s depressing.  same evaluation
E: Ve[ry
L: [but it’s a fantastic film.  disagreeing
*EFL learners are taught “I agree” and “I think so too.” which are same
evaluations, thus, displaying weak agreement suggesting disagreement.

3) downgraded assessments (downgrades)
With downgrades the speaker chooses a weaker wording. Downgrades
should also be considered “weak agreement” and, therefore, as
foreshadowing possible upcoming disagreement.
A: She’s a fox!
B: Yeh, she’s a pretty girl.
F: That’s beautiful
K: Isn it pretty
Downgrades frequently result in subsequent rebuttal as in the following
excerpt where a first assessment is offered (“a fox”), which receives a
downgraded evaluation (“a pretty girl”). This then causes the prior
speaker to reassert his position with an even stronger term (“gorgeous”).
A: She’s a fox.
L: Yeh, sh’e a pretty girl.
A: Oh, she’s gorgeous!

Basic patterns for disagreement
Pomerantz also identified two basic patterns for disagreement
(when disagreement is a dispreferred action).
Pattern 1) tokens/delays + strong disagreement
R: ...well never mind. It’s not important.
D: Well, it is important.
Pattern 2) tokens/delays + weak agreement + weak disagreement
R: Butchu admit he is having fun and you think it’s=
=funny.
K: I think it’s funny, yeah. But it’s a ridiculous=
=funny.

The impact of unstated disagreements
A silence after an assessment can be heard as a sign of potential future
disagreement. As a result prior speaker may choose to continue (Rule
1c) with something designed to deal with this potential disagreement
even before it occurs as in the following case where B states an opinion,
then backs down from the original assessment.
B: ...an’ that’s not an awful lotta fruitcake.
(1.0)
` B: Course it IS. A little piece goes a long way.

Disagreement as preferred next action
While agreement is a “preferred action” in most situations, it is not the
preferred action in all situations. One situation where it is not the
preferred action is after a “self-deprecating” assessment, i.e. when the
prior speaker has said something negative about him/herself.
Pomerantz (1984:81) states: “If criticizing a co-conversant is viewed as
impolite, hurtful, or wrong (as a dispreferred action), a conversant may
hesitate, hedge, or even minimally disagree rather than agree with the
criticism.”
B: ...I’m tryina get slim.
A: Ye:ah? [You get slim?
B: [heh heh heh heh hh hh
A: You don’t need to get any slimmah,
A: ...I’m so dumb I don’t even know it hhh! —
heh! B: Y-no, y-you’re not du:mb,...
These disagreements are packaged in the preferred turn shape.
.

Conflicting ("cross-cutting") preferences
There can be situations in which multiple preferences are at work. For
example, there is a preference for agreeing with assessments. However,
doing self-praise is dispreferred. So how are participants supposed to
deal with compliments (which come packaged as assessments)?
Pomerantz (1978) demonstrates that it is the constraints imposed by
these cross-cutting preferences that account for the manner in which
most compliment responses are designed. In the following examples,
the compliment is downgraded or minimalized, but not outright rejected.
A: Oh, it was just beautiful.
B: Well thank you Uh I thought it was quite nice.
A: Good shot
B: Not very solid though
A: You're a good rower, Honey.
J: These are very easy to row. Very light.

Why preference organization matters
“Preference principles play a part in the selection and interpretation of
referring expressions, repair, turn-taking, and the progression through a
sequence of actions. In addition, there are contexts in which participants
orient to multiple preference principles. Sometimes, these multiple
principles are in conflicts with one another, and, in those circumstances,
participants use systematic practices to manage the conflicts involved.”
(Pomerantz and Heritage, 2013:210)
Preference organization is fundamental to the structure of talk-in-
interaction and to navigating our social worlds.
Preference organization has implication for the design of both questions
and their replies.
Guiding learners towards an understanding of the practices proficient
English speakers use to handle both preferred and dispreferred actions
will help them interact with fewer social problems.

Why preference organization matters
Research on preference organization demonstrates how inappropriate,
or just plain grammatically wrong, the so-called “short answers” common
to virtually every EFL textbook are, e.g. A: Do you want to go to a movie?
B: No, I don’t. This is “structurally incorrect” because a dispreferred
action has been packaged using the preferred turn shape. As a result
teachers and textbooks are actually teaching students how to act “curt”
and “hostile.”
It is vital that learners know how to reject as well as accept, how to
refuse as well as grant, how to disagree as well as agree. EFL materials
generally do a poor job of this. Quite often textbooks present an eternally
rosy world where everyone accepts, grants, and agrees. Or the
examples of doing rejecting, refusing, disagreeing, etc. do not reflect
observations from naturally occurring talk-in-interaction.

Challenges facing future study of preference
Pomerantz and Heritage (2013) argue that one current weakness in
research on preference is an overly simplistic association of one
preference with overly broad action categories, for example stating that
"accepting invitations/offers is preferred" or that "requests are
dispreferred." Not all invitations are alike. Not all requests are alike.
"Would you like the last piece of pie?" (Clearly prefers rejection)
Likewise, whether a given request is designed and oriented to as
preferred or dispreferred appears to hinge on what participants perceive
(and display to co-participants) as the level of entitlement. Requests
reflecting strong claims to entitlement may take the form of imperatives
("Pass the salt") while requests suggesting lower entitlement (higher
contingency) may receive different treatment. As such, future research
into preference will need to look deeper and at a more situated level.
For language teachers this means that we need to be carefully about
being overly dogmatic in our presentation of preference organization.

Preference in question design
Today we have addressed some of the ways that preference plays out in
sequential organization. For the most part, this dealt with preference
related to a 2PP.
However, it has long been known both in linguistics and conversation
analysis that preference organization is often implicit in the design of
questions (and questions as vehicles for other social actions), that the
formatting of a question can embed a preference for the type of
response it seeks to engender (see Bolinger, 1978; Heritage, 1984;
Heritage, 2002; Heritage, 2003; Heritage, 2010; Lee, 2013; Pomerantz,
1984, Pomerantz and Heritage, 2013; Sacks, 1987).
Furthermore, the replies to questions themselves reveal an orientation to
these embedded preferences.
But that's a topic for another day…specifically tomorrow!!! 

What is the CA attitude
towards grammar?

"Grammar, of course, is
the model of closely
ordered, routinely
observable social
activities."
(Sacks, 1992:31)

"…it should hardly surprise us if some of the
most fundamental features of natural
language are shaped in accordance with
their home environment in co-present
interaction, as adaptations to it, or as part of
its very warp and weft.
For example, if the basic natural
environment for sentences is in turns-at-talk
in conversation, we should take seriously the
possibility that aspects of their structure -
for example - their grammatical structure -
are to be understood as adaptations to that
environment."
(Schegloff, 1996:54)

Did the canyon make the river?

Or did the river make the canyon?

More than just grammar…
Experience with CA has led me, as a language
teacher, to re-evaluate almost everything I
thought I knew about "grammar."
What we need is grammar-in-interaction and
grammar-for-interaction, which extends beyond
the traditional borders of syntax.
Research in the field of CA is helping to build
this grammar-for-interaction one empirical
observation at a time.

CA insights on teaching conversation
The reason to teach conversation is not to practice the linguistic system,
but because it is the most ubiquitous form of human interaction and the
cradle of language development.
Talk-in-interaction is a skill in its own right not to be subsumed under
“speaking.”
Recognize that mundane conversation is the baseline, and most
complex, form of talk-in-interaction. Design activities that promote this
type of conversation.
Discussion, and in particular academic discussion, is a specialized genre
that should not be confused with conversation, although many of the
practices are shared.

Most talk is organized around action sequences rather than topics.
Therefore, conversation textbooks should not have units themed around
topics. Units are better organized around action sequences and their
expansions and alternative trajectories.
Sample conversations (authentic, simplified authentic, or scripted)
should ensure that natural turn-taking practices are accurately reflected.
For example, even in 2-party talk, the sequence of speakers is not
necessarily A > B > A > B.
More open-ended tasks allow for more natural interaction.
Shift focus away from the “structures of language” to the “structures of
practice” (Lerner, 1996). In other words, focus on how participants
manage talk-in-interaction.

Move away from typical textbook “information gap” activities which rarely
result in authentic interaction.
Minimize the use of traditional EFL dialogues, particular ones designed
to present a specific grammatical feature. If possible use authentic or
modified authentic examples and/or scripted examples, which conform to
empirical observations, for example about the shape of dispreferred
turns.
Many sequences can be presented in as few as two to four turns. We
might call these mini-logs. Any full dialogue will almost certainly never be
heard again. However, these mini-logs and the language in them are
often high frequency.
CA-inspired teaching materials should be designed in a way that
teachers with little or no knowledge of CA can still use the materials. It is
not advisable to explicitly use CA terminology with learners. The
teacher’s manual can help prepare the teachers.

Questioning Questions:
Questions and their
Replies as Social Actions

Traditional EFL
approaches to teaching
questions and answers

Questions, questions, questions
English language textbooks and teachers spend an inordinate amount of
time teaching the syntax of English questions and answers.
Textbooks almost always group questions according to their grammatical
form. First, there are the yes/no ("polar") questions using the present
tense forms of BE. In this unit, students usually also get their first taste
of the so-called "short-answer" format replies. Next up are polar
questions with present tense forms of DO. Sometime later they get the
past tense forms (was, were, did). Then come questions using the
"present perfect" (a label which communicates exactly nothing). These
are followed by the Wh-questions with these auxiliaries.
EFL textbooks are all about the GRAMMAR of questions.
Little or no thought is given to the communicative or socio-interactive
functioning of questions.

If you are like most EFL
teachers around the world…
…you have probably, without
thinking much about it, been
teaching questions through
transformation.

Transformational grammar lives on…and on…
A great many EFL teachers have never heard of Noam Chomsky or his
Transformational Generative grammar. Yet a great many of these
teachers and a great many EFL textbooks teach questions through
transformation…as if doing algebra! That is to say, students are
presented with a statement and required to transform it into a question.
He is a doctor.
Is he a doctor?
She works in a bank.
Does she work in a bank?
This is essentially having the students turn an answer into a question,
which is certainly the wrong way around.

English question formats and functions
A better approach is to teach students (i.e. "lead learners to learn") how
to match the intent of their question with the appropriate format.
INTENT FORM
? ABOUT HABIT Do you…
? ABOUT SPECIFIC PAST ACTION Did you…
? ABOUT EXPERIENCE (EVER) Have you…(ever)…
? ABOUT EXPERIENCE (ALREADY) Have you…(already)…
? ABOUT ABILITY Can you…
? ABOUT CURRENT ACTIVITY Are you V+ing…
? ABOUT INFORMAL FUTURE Are you going to…
? ABOUT FUTURE PLAN Will you…
? ABOUT OBLIGATION Do you have to / need to…
? A=B? Are you…

Grammar formation of Information questions
In English, the so-called Wh-questions are formed by combining
interrogative words with one of the forms of either an auxiliary verbs DO,
BE, HAVE or one of the modal verbs.
What…
Where…
When…
Why…
Who… / Whom…
Whose…
Which…
Which+Noun… (an unlimited number of these)
How…
How+Adjective… (an unlimited number of these)
Again, this is usually presented from a purely mechanical perspective
and often again in transformation drills of turning answers into questions,
e.g. He works in a bank.  Where does he work?

The EFL Grammar Mirror
Most EFL teaching materials treat responses to questions as a sterile
matter of grammar transformation…as a simple reflection of the
grammar (and vocabulary) of the question. This is not what we normally
find in naturally-occurring talk.
A: Where do you live?
B: I live in Tokyo.
A: What is your favorite color?
B: My favorite color is blue.
A: Where did you go last weekend?
B: Last weekend, I went to my friend's house.

Activity: Non-mirrored answers
1) Q: Who did you go with?
2) Q: How much did it cost you?
3) Q: Where do you live?
4) Q: When are you going to the party?
5) Q: Where did you go last night?
6) Q: Where are you from?
7) Q: What type of sports do you prefer?
8) Q: Where is your piano concert?
9) Q: How did you get there?
A) R: Well, I grew up in southern California.
B) R: Well, it’s supposed to start at nine, I think.
C) R: Well, my hometown is, or was, Takamatsu.
D) R: Well, we started the evening at my friend’s house, his apartment.
F) R: Well, I ended up going alone.
G) R: Well, I had planned to take the train but…
H) R: Well, initially, it was scheduled to be held at the city hall.
I) R: Well, actually, I got a huge discount.
J) R: Well, I’m more into reading and watching movies.

Replies to polar questions: The short answers
The format as presented in EFL textbooks begins with either a "yes" or a
"no" token followed by a clause consisting of a subject (usually a
pronoun or "there") plus repetition of the question's auxiliary or modal
verb (mirroring the tense in the question). The polarity of the verb
matches the turn-initial token, e.g. Yes, I do. / No, she wasn't. / Yes, they
have. / No, he can't. / Yes, we did. / No, it won't.
[yes / no] + [subject pronoun / there] + [auxiliary / modal]
A: Have you been to Egypt?
B: Yes, I have. / No, I haven't.
A: Do you want to go to a movie?
B: Yes, I do. / No, I don't.
The short-answer format is DEEPLY problematic and rarely found in
naturally occurring interaction…as we shall see later.

But what are questions DOING in conversation?
With all this focus on the grammar of questions, EFL textbooks and
syllabi seem to lose sight of the much more significant matter of what
questions are actually doing in interaction.
For participants there is always the interpretive issue of why some
particular question, in that particular form, was done at this particular
time.
WHY?
THAT?
NOW?
A: Do you want to get something to eat?
B: Sure, what do you have in mind?
A: I don't know. What time is it?

"…even though information (or
confirmation) may be part of what
a question is built to get, this
seems to be virtually never…what
questioning in interaction is
centrally about."
(Steensig and Drew, 2008:9)

Turns which assume the grammatical
format of questions regularly serve
as "vehicles" for other social actions.

…and not all questioning actions take
the grammatical form stereotypically
associated with questions.

Think of some “questions” that
are primarily hearable as doing
some other social action.
Type them in the chat!
I’ll give you a few minutes.

What social action is being being done?
A: Have you ever seen such a beautiful sunset?
B: __________________________
A: Would you like to come to a party?
B: __________________________
A: Why weren’t you at work yesterday?
B: __________________________
A: Are you an idiot?
B: __________________________
A: Aren’t we going to be late?
B: __________________________
A: Are you gonna eat that?
B: __________________________
A: Isn’t he just the cutest thing?
B: __________________________

Question design
vs.
Question formation

Alternative questions (X…or Y)
There are three basic question types (typologically speaking) in English:
Information questions, polar questions, and alternative questions. For
some reason many EFL learners seem to have a hard time with
alternative questions, which they tend to answer with "yes" or "no"
instead of selecting one of the alternatives.
A: Would you like coffee, or tea.
B: Yes.
Note that sometimes the two alternatives are separated by a silence.
This can be analyzed as B withholding talk at the point of first possible
completion (after "coffee") resulting in silence. This may serve as a
warning to A of an upcoming dispreferred action (namely rejection of the
offer of coffee), which prompts A to produce an alternative offering.
A: Would you like coffee? (0.2) or tea.
A: Would you like coffee? (0.2) or tea? (0.2)

Tag-questions
Stivers (2010) states that tag-questions were rare in her data, comprising
only 6% of polar questions. Furthermore, 77% of those tag-Qs
employed a simple lexical tag (e.g. "right" or "huh") rather than the more
complex clausal formats taught in EFL textbooks.
The tag can be produced with either final rising or final falling intonation
– despite the textbook characterization of them always having rising
intonation. Given that the tag is regularly produced both ways, what is
the interactional difference?
Tag-Q with rising final intonation
He's arriving at eight, isn't he?
Tag-Q with falling final intonation
He's arriving at eight, isn't he.

Tag-questions
You often hear that the Japanese final particle "ne" has no equivalent in
English. But is this really true?
Kim: (Very good.)
(6.5)
Mark: Not bad for free huh?
(0.3)
Kim: Hm mm.
What interactional work is being done by the addition of "ne" or "huh" at
the end of a turn?
Like all questions, tag-questions don't merely ask. They are also
pressed into service in pursuit of other social actions. They are subject
to why-that-now inspection. What do you feel is going on in the following
instance? Is it a tag-question or something else?
Jess: That's kind of a lot for breakfast don't-ya think?.
Mike: Nah::, I thin' iz- I think it's great.

Appendor questions
Some questions are built to appear as extensions of a prior turn-at-talk.
They are grammatically parasitic. These are referred to as appendor
questions (Sacks, 1992:660; Schegloff, 1997).
Ingrid: I know that it's just too hard b't_ Like it is too
hard_ It's disrupting both of our li:ves.
Zoe: ((laughing loudly about other conversation))/(1.5)
Mandi: Tuh be friends, or tuh be (0.2) more than friends.
Ingrid: Tuh be more than friends is disrupting both of our
lives.
I have never seen an EFL textbook that teaches appendor questions as
a strategy for producing questions, although it would seem to be quite a
simple and productive method.
A: I went shopping.
B: For shoes?

Questions as vehicles for social actions
It turns out that a very great number of “questions” in conversation,
perhaps most, serve as vehicles for other actions and it is the social
action, which determines what sort of response is relevant.
Some turns-at-talk (which assume the grammatical shape of questions)
will be heard by participants first and foremost as invitations, other
questions as requests, other questions as complaints, other questions as
assessments or compliments, and still other questions as part of pre-
sequences or insert sequences, and so on. As such, these first-pair part
actions call for something other than an “answer.” Even a seemingly
neutral question such as “When does the shop close?” might be heard
by a tardy friend as a mild rebuke and elicit a response such as “Don’t
worry, we have plenty of time” rather than the provision of a time.
From the perspective of conversation analysts looking at data, the
syntactic formatting of a turn-at-talk is of secondary importance to co-
participants' displayed understanding of the action it performs.

Questions that are not
(primarily) doing
questioning

Fairly obvious…
(Common speech acts*)

INVITATIONS / OFFERS
Do you want to…
Would you like to…
Do you feel like V-ing…
Do you want me to…

REQUESTS
Can you…
Could you…
Would you…
Would you mind V-ing…
Can I…
Could I…
Would you mind if I…

COMPLAINTS
Why can't I…
Why didn't you…

Not so obvious
(Interactional practices)

PRE-INVITATIONS
Are you busy…
What are you doing…
Are you free…
Do you like…
Have you…

PRE-REQUESTS
Are you busy…
What are you doing…
Do you know how to…
Have you ever…
Do you have a…

PRE-ANNOUNCEMENTS
Did you hear the news?
Did you hear Wh-Q…?
Do you know Wh-Q…?
Do you wanna know Wh-Q…?

PRE-PRE
Can/Could I ask you a
question?
Can/Could you do me a favor?
Can/Could you help me?

"Question design in
conversation"
(Hayano, 2013)

How do turns-at-talk get heard as questions?
We naively think of questions as being identifiable by their grammar
and/or prosody. That's certainly how EFL textbooks present questions.
However, detailed study of talk-in-interaction (in numerous languages)
demonstrates that grammar and prosody are frequently not reliable as
markers that a particular turn-at-talk will be heard (and responded to) as
"doing questioning."
At the very core of "questioning" lies a recipient-tilted epistemic
asymmetry (Stivers & Rossano, 2010), which in simpler terms means
that the recipient has been assumed to have fuller information.
It this excerpt, the interviewer (IR) asks no question, but the talk
nevertheless occasions an answer by the interviewee (IE).
IR: So in a very brief word David Owen you in no way regret
what you did er dispite what has (happened) in Brighton
this week in the Labour Party.
IE: n- In no way do I regret it. ( Repeats the wording. Why?)

How do turns-at-talk get heard as questions?
This type of epistemic asymmetry can even make a "why" or "how come"
interrogative unnecessary in prompting an account, as in the following
except. Participant A produces a statement. However, the reason lies
within B's purview and thus she "answers" (provides the reason) for a
question that hasn't been asked.
A: . . . Yer line's been busy.  Statement / Observation
B: Yeuh my fu(hh) .hh my father's wife  Answer / Account
called me. .hh so when she calls
me::. .hh I always talk fer a long
time. Cuz she c'n afford it'n I
can't. Hhhh heh .ehhhhhh
Once again, we see that it is epistemic asymmetry that is primarily
responsible for allowing a prior turn-at-talk to be heard as doing
questioning.

The epistemic gradient
The term "epistemic" refers to "states of knowledge." In terms of doing
questioning we can think of these states of knowledge existing on a
continuum…or as a gradient, the steeper the slope, the greater the
epistemic asymmetry.

Presuppositions
Questions embed presuppositions. The question "where did you eat?"
quite obviously presupposes both that the recipient has, in fact, already
eaten and that the eating was done elsewhere. Sometimes, the
presuppositions can be more subtle, objectionable, or insidious.
These embedded presuppositions can present a problem:
"...answerers face a choice between: (a) providing a relevant answer
and accepting the presupposition(s); or (b) bringing the
presuppositions to the surface of the interaction, rejecting them but
not answering the question, and thus potentially being heard as
"evasive." (Hayano, 2013:402)
In such cases, the recipient may need to employ a range of interactional
practices to navigate these troubled waters.

Preference in polar question design
We've seen that preference organization plays a significant role in the
formatting of 2PP actions…and sequences as a whole. We have also
noted that certain 1PP can be dispreferred, such as requests.
It is less obvious, however, that preference organization is often implicit
in the design of questions (and questions as vehicles for other social
actions), such that the formatting of a question can embed a preference
for the type of response it seeks to engender (Heritage, 1984; Heritage,
2002; Heritage, 2003; Heritage, 2010; Lee, 2013; Pomerantz, 1984,
Pomerantz and Heritage, 2013; Sacks, 1987).
These empirical studies of naturally occurring conversational data have
identified constructions that normally embed a preference for an aligning
“yes” response as well as those formats preferring an aligning “no”
response (see Hayano 2013:405).
In more formal linguistics this phenomenon has been studied under the
rubric of "polarity."

(+) PREFERRED / (-) DISPREFERRED
The following formats embed a preference for a (+) positive or affiliating
response, which is to say, they seek to elicit a positive response.
Positively formulated straight interrogative (Have you heard from her?)**
Positive statement + negative tag (You’ve heard from her, haven’t you?)
Positively formulated declarative questions (You heard from her?)
(-) PREFERRED / (+) DISPREFERRED
In contrast, the following formats embed a preference for a (-) positive
response.
Negative declaratives (You haven’t / never heard from her.)
Negative declarative + positive tag (You haven’t heard from her, have you?)
Positive interrogative + negative polarity item (any, ever, at all, yet, etc.)
Positive interrogative + negatively tilted adverbs (seriously, really, etc.)
** Are you going to eat that? ( Doesn't this prefer a "no")

Reversing polarity in search of a preferred action
One demonstration of the relevance of preference in polar question
design is that when a dispreferred response seems to be in the offing (as
displayed, for example, by silence) the questioner (A) can reverse the
polarity of the question in hopes of eliciting an affiliating response as in
this example (Schegloff, 2007:71):
A: D’they have a good cook there? (+ PREFERRED)
(1.7) (POSSIBLE TROUBLE)
A: Nothing special? (- PREFERRED)
B: No, Everybody takes their turn. (PREFERRED AGREEING)
Given how much time and energy is devoted to questions in EFL
materials, it would be worthwhile for teachers and textbook writers to
learn more about how questions are actually handled in real-world
interaction.
For outstanding overviews of CA research on question design see
Stivers (2010) and Hayano (2013).

How to discourage questions in class
How many times in class have you asked the following question…and
gotten no response?
Does anyone have any questions?
Ironically, the negative polarity ("anyone" and "any") of this formulation
embeds a preference for a negative response. You are as much as
telling students that you don't want them to ask any questions!
In contrast, the same basic question formatted with positive polarity
embeds a preference for a positive response.
Does someone have a question?
Of course, such subtle variations may not be enough to override EFL
student reticence to speaking up in class, but such orientations to
embedded preferences have been documents in CA research, for
example, in doctor-patient interactions (see Heritage et. al., 2007).

Designing requests
Several CA studies have investigated alternative request formats from
the perspective of entitlement, both in terms of doing requests and in
having them granted.
Lindström (2005) looked a Swedish data.
Requests in imperative form = Entitled to have request granted
Requests in question form = Not entitled to have request granted
Heinemann (2006) looked a Danish request formats.
Negative interrogatives (e.g. Can't you…) = Entitled
Positive interrogatives (e.g. Will you please…) = Not entitled
Curl and Drew (2008) looked at English requests.
Can you… / Could you… = Entitled, non-contingent
I wonder if… = Not entitled, contingent

.
Doing questioning is much more than grammar
EFL textbooks tend to present questions from an almost entirely
grammatical perspective. They are taught as syntax not social actions.
This does language learners a great disservice.
There is still much that could be said here about how "doing questioning"
is actually accomplished in situated interaction. And there is still much to
be investigated.
But I hope you have come away with a strong feeling that doing
questioning, in all its many ways, is a fundamental skill in the
management of social interaction.
Next, we move on to the design of responses which like questioning is of
great importance in interaction and due greater consideration that it is
given in EFL materials and syllabi.

Clausal vs. phrasal
responses to information
questions

Phrasal vs. sentential responses
Teachers often insist that students answer questions using a “full
sentence,” the pedagogic purpose being to drill some grammatical
structure.
However, CA research (Fox and Thompson, 2010, cited from Lee, 2013)
reveals that unmarked, or in CA parlance preferred (or canonical or
default) responses to Wh-questions are typically packaged as lexical or
phrasal units, that is to say, not as full sentences (i.e. "clauses"). These
lexical or phrasal responses align with the presuppositions and
relevancies of the question. In other words, they treat the question as
unproblematic. In the following example of a preferred phrasal response
(from Fox and Thompson, ibid.:137) note that the response is provided
immediately (with no gap) upon the completion of the question as is
typical of the preferred turn shape described by Pomerantz (1984).
Molly: .hhh How far up the canyon are you.=
Feli: =Ten miles. ( phrasal response)

A full sentential response suggests some type of misalignment with the
presuppositions of the question (Fox and Thompson, 2010).
Q: Where are you from?
(.)
R: I live in New Zealand. ( sentential response)
Contrast this with the alternative phrasal response below, which aligns
seamlessly with the question.
R: New Zealand. ( phrasal response)
Minor troubles can still be displayed in phrasal responses through
various delays and hedges. However, sentential responses display more
significant trouble, are less frequent, and often incorporate delays and
prefaces. In short, lexical or phrases responses provide a “no problem”
answer, while a sentential length answer suggests disaffiliation towards
the presuppositions of the question.

A full sentential response suggests some type of misalignment with the
presuppositions of the question (Fox and Thompson, 2010).
(.)
R: I live in New Zealand. ( sentential response)
As further talk revealed, R was actually born in Pakistan, but had
immigrated to New Zealand with his family as a young child. His full
sentential response resists his perception of an embedded
presupposition that the formulation "where are you from" is tied to ethnic
identity.

Inferring purpose of a query
Pomerantz (2017) examines question-answer sequences in which the
reply does more than just answer the literal question (also see Walker,
Drew, and Local, 2011). Instead, the response addresses the inferred
reason behind the question. This practice reflects nicely the
WHY…THIS…NOW challenge constantly facing the recipient of a
question.
Shi: hhh Suh how'r you?
Ger: Oka:y d'dju just hear me pull up?=
Shi: =.hhhh NO:. I wz TRYing you all day. 'n the LINE=
=wz busy fer like hours.
Ger: Ohh:::::::, ohh:::::
Shirley's answer provides an reason as to why (from Geri's perspective
as indicated by her inclusion of "just") this phone call came immediately
after she returned home.
A: _______________________? (One student writes a question)
B: Why?
A: _______________________. (Another student gives a reason)

Crafting responses to
polar questions

Responses to polar questions
As with information questions, a great many polar (“yes/no”) questions
are actually doing something other than questioning, i.e. requesting
information. This is obvious from the fact that 70% of the questions in
Stivers (2010) data were built as polar questions.
An inquiry such as “Do you like pizza” is rarely oriented to as a neutral
question about “likes and dislikes” (as is typical in beginning units of EFL
textbooks). Instead, proficient language users will more likely hear this
as initiating a pre-invitation sequence, making relevant a go-ahead or a
blocking move (or a hedged response).
In fact, the design of responses to polar questions is much more
interesting, much more varied, and much more important than the
unnatural (and problematic) formats taught in EFL textbooks.

Non-answer responses
A question makes relevant an answer. Not responding at all to a
question, i.e. responding with silence, (as many Japanese students do)
may be a socially and/or interactionally sanctionable act.
Furthermore, there is a demonstrated preference in conversational data
for answers over non-answer responses, which respond without actually
answering (Stivers, 2010; Stivers and Robinson, 2006). Stivers (2010)
found non-answer responses occurring in fewer than 20% of the
hundreds of question sequences she examined. Not only are answers
hugely more frequent, they are also provided much more quickly.
The following except provides an instance of a dispreferred non-answer
response. This follows the dispreferred turn shape, including various
delays and with the non-answer (“She hasn’t said”) only provided in turn-
final position.
J: But the train goes. Does th’train go o:n th’boa:t?
M: .h.h Ooh I’ve no idea:. She ha:sn’t sai:d.

Stand-alone "yes" or "no" responses
A stand-alone “yes” response can sound curt and may shut down the
sequence. Similarly, Ford (2000) showed that co-participants regularly
withheld talk following a stand-alone “no” response and that this silence
often prompted an expansion of the “no” response. In other words, a
stand-alone “no” is treated as an insufficient response.
EFL learners frequently respond with only "yes" or "no." When pressed,
they tend to minimally expand their “yes” or “no” with the stock “short
answer” formats…as they have been taught to do.
As we saw in the section above on question design, polar questions can
embed a preference for the type of response it seeks. In light of this,
both “yes” and “no” can be preferred responses…or they can both be
dispreferred responses. If they are preferred, the turn will usually
assume the characteristic preferred turn shape. If dispreferred, then the
turn (as well as sequence) may include various delays and hedges with
the “yes” or “no” occurring later in the turn…or never!

Type-conforming vs. non-type-conforming
Research has also demonstrated a further fundamental form of
preference in responses to polar questions (Raymond, 2003; Heritage
and Raymond, 2012). Raymond called responses prefaced with either
“yes” or “no” tokens* type-conforming responses, in that the design of
polar questions sets up an expectation for one of these two response
tokens. Type-conforming responses accept the presuppositions of the
question.
Raymond (ibid.:948) found that type-conforming responses follow the
typical design of preferred turns (see Pomerantz, 1984; Schegloff, 2007),
as in the following instance from a health visitor (HV) interview with a
new mother.
HV: How about your breast(s) have they settled do:wn
[no:w.
Mom: [Yeah they ‘ave no:w yeah.
*A "token" might be any variant such as "yeah" "uh-HUH" "yep" "nah."

Type-conforming vs. non-type-conforming
However, Raymond also found numerous instances of non-type
conforming responses, in which a response to a polar question was not
prefaced by either “yes” or “no.” His analysis demonstrated that non-type
conforming responses are dispreferred and indicate a problem with the
presuppositions of the question (Raymond, ibid.:948).
.
HV: Mm.=Are your breasts alright.
(0.7)
Mom: They’re fi:ne no:w I’ve stopped leaking (.) so:
Mom’s non-type conforming response resists a presupposition that there
was never any problem.
Non-type conforming responses are typically entirely missing from EFL
textbooks. We don't even give students the option of omitting a turn-
initial "yes" or "no."

Don't be short:
The occurrence and non-
occurrence of short-
answer format replies

Should we be teaching the short-answers?
Found in virtually every EFL textbook the world over, the so-called "short
answers" are treated as bread-and-butter grammar and the default
format for responding to polar questions. This following examines the
occurrence, and more significantly, the non-occurrence of this format in
naturally occurring L1 English speaker conversations.
One immediate observation is that these short answers are rare and,
from a CA perspective, rarity is often an initial indication that a practice is
interactionally marked, which is to say that it does some special work.
This raises several interesting questions. If short-answers are not
common, what is found in their stead? When the short-answer format is
used, what does the short-answer format accomplish that the far-more-
common stand-alone "yes" or "no" token would not?

Response types to polar questions
Lee (2013) provides an outstanding summary of CA research on
response design. Type-conforming responses generally accept the
presuppositions and constraints of the question and according to
Raymond (2003) and Stivers (2010) occur in about 75-80% of question
sequences in English conversation. Non-type conforming responses
resist the presuppositions and constraints, thereby framing the question
as problematic.
Repetitional responses – Repeat words in the question and display
greater agency relative to type-conforming responses.
Marked confirmations – Turn-initial tokens such as "oh," "absolutely,"
can contest the relevance of asking the question. Responding with "of
course," can reject entirely one alternative (Stivers, 2011b).
Transformational responses – Display even more resistance by
shifting some aspect of the presuppositions of the question (Stivers and
Hayashi, 2010).

The data
The talk from which instances of polar questions and their replies were
drawn consists of a set of classic telephone calls between L1 English
speakers, recorded in the 1960s (Newport Beach = NB). These calls
were originally transcribed by Gail Jefferson, but later re-transcribed for
CLAN by Kraesten Nyman and Johannes Wagner. For illustrative
purposes a small number of instances have also been drawn from other
sources. The calls were retrieved from TalkBank as CLAN files from
which Word.docx files were created.
There are a total of 25 calls which together yield over 120 pages of
transcript, from which a master collection of more than 120 candidate
polar question sequences was culled. These sequences were then
sorted into sub-collections according to the response type.

Instances of the short-answer format

Oh, sure, I do.   Yes, I do.

Yeah following newsmark / news receipt
Following newsmark (verb first)
Following news receipt (verb last)

Yeah following “you know what I mean”

No following a newsmark
No to offer

Non-type conforming responses
Raymond (2003) found numerous instances of non-type conforming
responses, in which the response to the polar question was not prefaced
by either “yes” or “no.” His analysis demonstrated that non-type
conforming responses are dispreferred and indicate a problem with the
presuppositions of the question.
Furthermore, this type of preference is independent from other
preference structures operational in polar questions.

Non-answer answers – I don’t know

(Oh) + non-type conforming response

Non-type conforming following news
receipt

Polar questions for other social actions

Non-type conforming short answers

So what are short answers doing?

Conclusion
The results of the analysis of the NB calls demonstrate that far from
being the default method of responding to polar questions, the so-called
short-answer format is exceedingly rare in natural conversation.
Furthermore, when it is employed, it appears to be doing specific
interactional work and in that sense it is a marked response type. In in
many contexts. For example, following an invitation, a positive short-
answer sounds stilted and a negative short-answer is hearably rude or
curt (since its formatting runs counter to the dispreferred turn shape).
Even as a response to a question, the short-answer can come off
sounding well…short…that is to say, abrupt or dismissive, as was the
case in the Bonny and Jim example.
Given the rarity of the short-answer format and its marked nature, should
teaching short-answer grammar really be given such “pride of place” in
our EFL textbooks and classrooms? I think not.

Why questions and their responses matter
“Depending on such factors as how [questions] are designed, who is
asking them, and the social and sequential context within which they are
asked, questions bring about diverse interactional consequences. In
fact,…questions are a powerful tool to control interaction: they pressure
recipients for responses, impose presuppositions, agendas and
preferences, and implement various initiating actions…” (Hayano,
2013:395-396)
Using questions effectively for a full range of interactional purposes is a
vital conversational skill.
The majority of questions found in talk-in-interaction are actually
primarily concerned with doing some other (or at least additional) social
action than simply “questioning.” They are heard by recipients (and
responded to) as invitations or offers or requests or compliments or
complaints or assessments, etc.

Why questions and their responses matter
EFL materials, in most cases, fail to go beyond the grammar of question
formats. For example, textbook after textbook presents “Do you like…” in
a unit about random “likes and dislikes” without informing students that
one of the prime uses of this “question” is in pre-sequences.
Preference organization extends into the design of both questions and
their replies and this needs to be incorporated into the teaching of
conversation.
The preferred (unmarked) format for replies to information-questions is
lexical or phrasal rather than a full sentential unit, which does the
additional work of displaying problems with the question. This needs to
be reflected in EFL materials.

On classroom observation…
For both pre-service and in-service teachers, classroom
observation has long been a valuable learning tool. It is
useful for new teachers (and pre-service student-teachers)
to observe more experienced teachers.
It is equally valuable for new teachers to observe
themselves while teaching, which of course can only be
done by recording the class.
Afterwards, these recordings can be reviewed, ideally with a
group of fellow teachers, and discussed.
This is a casual sort of informal self-evaluation rather than
formal classroom research.

Is CA the right
methodology for the
job?
CA vs. Ethnography
Ethnography + CA

What is “classroom interaction”
In the simplest sense any interaction that happens in a
classroom might be called ”classroom interaction.”
However, that’s not really very informative. Lots of different
sorts of interactions take place in a classroom. What
teachers are primarily interested in is “pedagogic
interaction,” that is, interaction that involves learning.
There has been quite a lot of research on the organization of
traditional teacher-fronted instruction. There has been less
research on less traditional student-centered learning in the
classroom.
Recently there has also been a growing research interest in
so-called “learning in the wild” which examines more
informal learning interaction outside of institutional settings.

The Practicalities of
Using CA for Analysis of
Classroom Interaction

The practical realities of recording
Good quality recordings. The good news is that record
devices from camcorders to smartphones are getting better
and better.
Unfortunately, the environments in which we use them are
staying the same. You will need to be aware of ambient
noise that would make doing transcription a nightmare. This
includes, for example, sounds coming from air conditioners,
the next classroom, or from outside.
Depending on the nature of the interaction, you may also
need multiple recordings from multiple positions.
To the extent possible, you want your recording(s) to capture
what the participants themselves would have been aware of.

Are audio-only recordings enough?
Sadly, no. Not if you are investigating co-present interaction.
Spoken elements of an interaction and what CA analysts call
“embodiment” are inseparably intertwined and cannot be
analyzed independently. A momentary silence in the talk is
not a cessation in the interaction.
Silence can be “filled” with embodiment and talk is virtually
always accompanied by embodiment.
Working only with an audio recording may allow certain
limited sorts of observation, but it cannot capture the whole
picture. You would be like a blind man watching a movie.
That said…

Are audio-only recordings enough?
That said…
…some of the earliest CA research was based on audio-only
recordings.
However, this is really not suited to classroom research.
On the other hand, if you have audio recording devices
available it is quite handy to supplement video recordings
with audio recordings so you can check something that
might have been unclear in the video.

Where to put the camera?
This is always a question facing the analyst. Ideally you
want the camera(s) to capture anything that might have
been relevant for the participants themselves.
For recording traditional teacher-fronted instruction, you
should have at least two cameras, one in back facing the
teaching and one in front facing towards the students.
Having two cameras in front (one in each corner) would be
even better.
Recording small group activities can be even trickier,
particularly if the participants are going to be moving around.

Conversation analysis
and
multimodality

Multimodal interaction
Overwhelmingly talk takes place during co-present
interaction. It is embedded in, and inseparable from, other
semiotic systems, including the resources of the body and
structure in the environment.
Back in the 1960s when CA got started, the only affordable
means of recording was with large reel-to-reel tape
recorders. For this reason, the field came to be called
CONVERSATION analysis. But as time and technology
have moved on, most conversation analysts now work with
video recordings for interaction that occurred face-to-face.
This has led to many wonderful insights, but it has also
made the analyst's task more challenging.

 "…when joined together in local
contextures of action, diverse
semiotic resources mutually
elaborate each other to create a
whole that is both greater than, and
different from, any of its constituent
parts."
 "…much existing research has
avoided the crucial issues posed by
the heterogeneous semiosis that sits
at the center of actual human action
by focusing on the analysis of
individual semiotic systems as self-
contained wholes."
 "…too often, analysts regard talk as
their starting point, even when talk
appears late in the order of things
accomplished in face-to-face
interaction."

The challenge of
unpacking multimodal
interaction

A question of granularity
 Schelgoff (2000) spoke of the “granularity” needed in analysis.
 Conversation analysts work with “fine-grained” transcripts of talk.
 Embodied aspects of interaction are often treated more coarsely.
 What level of granularity is needed in understanding embodiment?
 Heritage (1984, p. 241):
“...no order of detail can be dismissed, a priori, as disorderly,
accidental or irrelevant.”
 Gestures are infinitely malleable and interactively unfolding.

The challenge of
transcribing multimodality

01 M: what is that lines?
02 J: tables.
03 ((J weakly points past M to other tables))
04 ((J switches to three taps on table))
05 M: cables?
06 ((J lightly re-taps 3-4 times))
07 J: ˚tables˚
08 ((0.6)
09 M: [(unintellagble in overlap with teacher)
10 ((M points to drawing))
11 ((M begins to rotate paper))
12 ((J reestablishes control of paper))
13 M: [ah! did you mean those-
14 ((rotates to point to other tables))
15 [these two tables?

32 -> A: spring, summer,
K
A
S
33 S: ah [faru hah hah hah A::-H] $autumn$
34 K: [huh huh falu hah huh]
K
A
S
35 A: autumn,[autumn, °autumn°]
36 S: [( )yeah f]all
K
A
S
37 (0.20)

CLAN transcription
software
The analyst still does the transcription!

Transana transcription
software
The analyst still does the transcription!

23 S: .hh I::[y
24 A: [an’
25 (0.73)
26 other staff-u
27 K: mm
28 (1.00)
29 K: a-huh
30 (0.95)
31 A: cor-u ((“call”))
32 (1.17)
33 ?: [[un: ]
34 A: [[other] staff call-u[:
35 S: [you?
36 (0.18)

Red arrow locks in the instance of the frame grab

“Augmented transcripts” are needed to capture multimodal action

Goodwin, C. “The Semiotic Body in its Environment””

Embodied participation
 Interaction involves skilled, local negotiation utilizing the
resources of co-present bodies, structure in the
environment, as well as talk.
 Turns-at-talk (and turns-at-action) are often “pre-
conditioned” or “foreshadowed” by strategic positionings of
the body.
 Likewise, through the visible withholding of embodied
engagement, participants minimize active participation.
 The degree of embodied engagement appears to be a very
good predictor of verbal participation.

Embodied disengagement
 If a potential participant is physically disengaged from the
interaction, that individual may not be able to re-engage quickly
enough to step into the rapidly changing flow of events.
 Deep physical disengagement also renders the individual
“opaque” to other potential co-participants and thus calls into
question his or her status as a “participant.”
 Degrees of physical engagement are to some extent culturally
and individually constructed, but nevertheless can act as a barrier
to effective interaction.
 Limited physical engagement can therefore become a barrier to
effective language learning.

What about translanguaging and
transcription?

DISCUSSION TIME!!!
Now is your chance to ask
questions, comment, and/or
discuss your ideas on CA and
language teaching.

Thank you all for attending
this seminar series!
I hope you found it
interesting and potentially
useful for your own
teaching and future
research.

Selected references
Carroll, D. (2016). Natural responses: Breaking the grammar mirror. In D. Tatsuki & D. Fujimoto
(Eds.) in Back to Basics: Filling in the Gaps in Pragmatic Teaching Material. JALT Pragmatics
SIG, special publication book. Pp. 23-35
Heritage, J. and Raymond, G. (2012). Navigating Epistemic Landscapes: Acquiescence, Agency
and Resistance in Responses to Polar Questions. In J.P. de Ruiter (ed.), Questions: Formal,
functional and interactional perspectives. Cambridge: Cambridge University Press: 179-192.
Lee, S. (2013). Response design in conversation. In J. Sidnell and T. Stivers (Eds.), The Handbook
of Conversation Analysis. West Sussex: Wiley-Blackwell.
Fox, B. and Thompson, S. (2010). Responses to Wh-questions in English conversation. Research on
language and social interaction, 43(2), 133-156.
Raymond, G. (2003). Grammar and social organization: Yes/no interrogatives and the structure of
responding, American Sociological Review, 68: 939-67.
Stivers, T. (2010). An overview of the question-response system in American English conversation.
Journal of Pragmatics, 42(10), 2772-2781.
Stivers, T. and Robinson, J.D. (2006). A preference for progressivity in interaction. Language in
Society, 35(3) 367-392.

Keeping it real: Re-
imagining textbook
dialogs

Authenticity and granularity
Sentences such as the following were common in EFL materials
even in the late 20th century and are still found today. They are
intended to illustrate aspects (usually grammatical aspects) of the
language and authenticity is rarely a concern.
"My shopping must come to an end."
-- NHK English program

Authenticity and granularity
If we examine recent EFL textbook dialogues, for example the
following from Basics in Speaking (Rost, 1998:42), we find what
appears to be reasonably authentic English.
Jim: Do you want to go with us?
Sue: Where are you going?
Jim: To a karaoke bar.
Sue: Oh, gosh, no. I can’t sing karaoke.
Jim: What? Everybody can sing karaoke.
Sue: No, really, I can’t. It’s impossible. I’m a terrible singer.
Jim: Oh, come on. I’ll teach you some easy songs.
Yet, from the perspective of conversation analysts EFL dialogues
offer only an impoverished and often distorted caricature of talk-
in-interaction (Bernsten, 2002; Wong, 2002, 2003, 2007).
Specifically, such dialogues lack the level of fine-grained order
that typifies real-world interaction.

In what ways is this EFL
textbook dialog unnatural?
Kelly: Hello?
Emily: Hey, Kelly? It’s Emily. What are you doing?
Kelly: I’m studying. What are you doing?
Emily: I’m studying too. What are you studying?
Kelly: English. How about you?
Emily: I’m reading a book about China.
Kelly: That’s nice. Actually, I’m not studying.
Emily: Yeah, me neither. I’m watching a movie.
Kelly: Me, too! What are you watching?
Emily: I’m watching Harry Potter.
Kelly: No way! Me too!

Most obviously, it is contrived to model the
present continuous tense.
Kelly: Hello?

Unpacking the TCUs ("turn construction units")
Kelly: Hello?

Unpacking the TCUs ("turn construction units")
Kelly: Hello?
Emily: Hey, / Kelly? / It’s Emily. / What are you doing? /
Kelly: I’m studying. / What are you doing? /
Emily: I’m studying / too. / What are you studying? /
Kelly: English. / How about you? /
Emily: I’m reading / a book / about China. /
Kelly: That’s nice. / Actually, I’m not studying. /
Emily: Yeah, / me neither. / I’m watching a movie. /
Kelly: Me, too! / What are you watching? /
Emily: I’m watching / Harry Potter. /
Kelly: No way! / Me too! /
* The / marks a point of possible speaker transfer.

Applying SSJ turn-taking
1a. Current speaker may, but need not, select next speaker.
1b. If current speaker does not select next speaker, a next
speaker may, but need not, self-select by speaking first.*
1c. If no other participant self-selects, current speaker may,
but need not, speak again. This can also be worded as
Current speaker may, but need not, continue speaking.
2. Rules 1a, 1b, and 1c continue to apply in this order at
every possible transition relevance place until speaker
transfer occurs.
*Research suggests that participants may also make claims on next speakership
via various forms of embodiment.

>> ((ring, ring)) Summons
Kelly: Hello::? [called]
Emily: Hey, Kelly? 1b
(0.5) Not Recognized
It’s Emily. 1c
Kelly: Oh, hi Emily! 1b, Late
Emily: What are you doing? 1b, Pre-inv.
Kelly: >> Nothing much. I’m just studying. 1a, Go-ahead
>> (0.5) No invitation?
What are you doing? 1c
Emily: Me? Oh, I’m studying too. 1a
>> (0.5) No up-take
What are you studying? 1c
Kelly: English. 1a
>> (.) No up-take
How about you? 1c
Emily: I’m reading a book about China. 1a
Kelly: Uh-huh. That’s nice. 1b
>> (.) No up-take
Actually, I’m not studying. 1c
Emily: Yeah, me neither. 1b
>> (.) No up-take
I’m watching a movie.= 1c
Kelly: =Me, too! 1b
>> (.) No up-take
What are you watching? 1c
Emily: (I’m watching) Harry Potter. 1a
Kelly: No way! Me too! 1b

Blocking out the sequences
Anthony: Are you hungry?
Beth: Hungry? I’m starving?
Anthony: Do you like pizza?
Beth: No, not really. Do you like Chinese food?
Anthony: Yes, I do. I love it!
Beth: Let’s have dumplings.
Anthony: Great idea!
Beth: OK! Let’s go.

Upgraded version of prior dialog
Anthony: Are you hungry? (Anthony’s pre-invitation #1)
Beth: Hungry? I’m sta::rving! (upgraded go-ahead)
Anthony: Do you like pizza? (Anthony’s pre-invitation #2)
(…) (indication of upcoming block)
Beth: Eh, not really. / not so much (hedged format)
It’s not my favorite. (dispreferred, non-type-conforming reply)
(.) (Brief inter-turn silence, SSJ Rule 1c)
Beth: Do you like Chinese food? (Rule 1c, Beth's pre-invitation #1)
Anthony: Oh, I lo::ve it! (go-ahead, replacing “Yes, I do.”)
Beth: Let’s have dumplings. (invitation / suggestion)
Anthony: Great idea! (acceptance)
Beth: OK! Let’s go. (sequence closing third)

Any particular real-world conversation is the by-
product, the precipitate left behind so to speak, of
two or more people engaging in social interaction.
As such the chance that any particular
conversation would ever re-occur as it did is
astronomically unlikely.

RE-IMAGINING A DIALOG STEP-BY-STEP
B: Hi, Can you help me?
A: Sure. What can I do for you?
B: I want to change some money. Is there a bank in the neighborhood?
A: Yes, there is. There's one across the street.
B: Across the street?
A: Yes.
B: OK. Thanks!

Step #1: Unpacking the TCUs
B: Hi. / Can you help me? /
A: Sure. / What can I do / for you? /
B: I want to change some money. / Is there a bank / in the neighborhood? /
A: Yes, there is. / There's one / across the street. /
B: Across the street? /
A: Yes. /
B: OK. / Thanks! /

Step #2: Turn-timing and inter-turn gaps
B: Hi.=Can you help me? ("rush-through" then pre-pre)
A: Sure. (Go-ahead to pre-pre)
(...) (A waits for B)
A: What can I do for you? (A proceeds)
B: I want to change some money. (B tells trouble)
(...) (A doesn't respond)
B: Is there a bank in the neighborhood? (B follows with question)
A: Yes, there is. (A gives minimal response)
(...) (B "waits" for more info)
A: There's one across the street. (A elaborates)
(...) (Possible trouble)
B: Across the street? (B initiates "repair")
A: Yes. (A resolves the repair)
B: OK. / Thanks! (B provides a SCT)

Step #3: Considering alternative trajectories
B: Excuse me.
A: Yes?
B: Can you help me?
A: Sure.
(...)
A: What can I do for you?
B: I want to change some money.
(...)
A: There is a bank just down the street. (Pre-emptive offer)
B: Oh, is there?
A: Yeah, it's about half a block, next to a restaurant.
(...)
B: A restaurant?
A: Yes, a Mexican place, I think.
B: OK, great, thanks!

Step #3: Considering alternative trajectories
B: Hi=Can you help me.
A: Sure.
(...)
B: I want to change some money.
(...)
A: oh. (No immediate offer of help or suggestion)
(...)
B: Could you tell me if there is a bank around here?
A: Yeah, it's about half a block.
(...)
A: in that direction.
B: OK, thanks!

Myths about teaching conversation
You can’t teach conversation to beginners.
FALSE! Conversation is the most fundamental form of language. We
should strive to contextualize language from the very earliest moments
of language learning.
Conversation classes should come after the students have “learned
the basics.”
FALSE! Conversation is the basics!
Doing casual conversation is the same as having a discussion.
FALSE! Discussion and negotiation are highly specialized speech
genres with their own norms. One common model for conversation
classes is the “topic” based discussion class. Many EFL “conversation”
texts consist of a set of topics assumed to stimulate students’ desires to
“practice speaking.” But the ability to do discussion is predicated on the
practices of mundane conversational interaction.

Myths about teaching conversation
The teaching of conversation does not require any specialized
knowledge.
FALSE! To teach conversational interaction you have to understand the
practices and resources that participants employ. CA provides that
necessary knowledge.
You have to be a native speaker to teach conversation.
FALSE! This assumes that the conversation teacher’s main role is to
model speech, engage students in conversation, and “lead discussions.”
It’s not. Native speaker teachers are no more consciously aware of the
norms and practices of conversational interaction than nonnative
speaker teachers. Both need to develop awareness in this area.
.

Conversation Analysis and its Relevance to Language education

More Related Content

What's hot (20)

Similar to Conversation Analysis and its Relevance to Language education (20)

Recently uploaded (20)

Conversation Analysis and its Relevance to Language education