Cover Sheet

Author: Trine Heinemann

Address:Finnish Centre of Excellence in Intersubjectivity in Interaction

P.O. Box 4

00014 University of Helsinki

Finland

Tel. +45 21237862

Email: trine.heinemann@helsinki.fi

Author: Johannes Wagner

Address:Dept. of Design and Communication

University of Southern Denmark

Universitetsparken 1

DK 6000 Kolding

Tel. +45 65501445

Email: jwa@sdu.dk

Full title of Article: Recalibrating the context for reported speech and thought

Short title of the Article: Recalibrating the context

Word count (all inclusive): 7,794

Character count (with spaces): 46,851

1

Trine Heinemann (d. phil., University of York) is a Marie Curie Fellow at the

Finnish Centre of Excellence in Intersubjectivity in Interaction, University of Helsinki.

Her current research focuses primarily on how speakers of Danish indicate that they have

undergone a change-of-state. She is also working on requests and assessments in an American

shoe repair shop (with Barbara Fox) and on the action-implications of imperatives in Danish

(with Jakob Steensig). *Corresponding author: Trine Heinemann, Finnish Centre of

Excellence in Intersubjectivity in Interaction, University of Helsinki, Finland, E-mail:

trine.heinemann@helsinki.fi

Johannes Wagner (dr. phil., Odense University) is a Professor in the Department for Design and

Communication at the University of Southern Denmark. His research interests include second

language conversations, learning processes in interaction, and the role of objects in interaction.

He is currently working on two projects, Social Objects for Interaction and Learning (social-

objects.net) funded by the VELUX foundation 2014-2017 and Language Learning in the Wild

(languagelearninginthewild.com), currently funded as a research network by the Nordic Research

Council (NOS-HS). Johannes Wagner, Department of Design and Communication, University

of Southern Denmark, Denmark, E-mail: jwa@sdu.dk

2

Recalibrating the context for reported speech and thought

Abstract

This paper investigates how speakers who are about to produce, or in the midst of producing,

reported speech or thought (RT), temporarily abandon the production of RT to include other

material. Using Conversation Analysis, we identify three positions in which RT is abandoned

temporarily and describe the resources employed by speakers to make recognizable to the

recipient that what is about to be produced is something other than the projected RT. We then

demonstrate how such inclusions are done to recalibrate the larger context of RT, to ensure that

the recipient of RT is able to interpret the speaker’s underlying interactional project. We

conclude by discussing how the inclusions we have identified and described relate to other

practices that can be employed by speakers to adjust the design of talk in ways that best scaffold

the interactional purpose of that talk.

Keywords: Conversation Analysis, Reported talk and thought, Danish, Sequential analysis,

Recalibration

3

1 Introduction

Reported, or represented, talk and thought (henceforth RT) lends authenticity to an account

because the enactment of a reported event (Clark and Gerrig 1990) gives recipients access to that

event in a manner that enables them “to assess it for himself or herself ” (Holt 1996:229). In

determining how to understand, evaluate, and react to RT, recipients rely on the design and

content of RT, as well as on the context where RT is delivered. Speakers producing RT

consequently take care to provide the context within which RT was (allegedly) delivered,

employing a range of resources to convey their attitude towards either the content of the RT or

the original producer of the RT (Günthner 1997; Holt 2000). These resources include glosses

such as “I’m broiling about something” (Holt 2000), descriptions of cognitive states and or

stances (Rae and Kerby 2007), extreme case formulations (Pomerantz 1986), and script

formulations (Edwards 1995). Most of these resources are employed in the RT preface, i.e. the

talk leading up to the production of RT. Here they serve to help speakers “construct their version

of events as robust, hard to refute and in correspondence with reality” (Lamerichs and te Molder

2009:410). RT is consequently presented in a contextual frame, where “the reported speech gets

structured by the reporting context, by how the reporting speaker wants recipients to take the

reported speech” (Buttny 1998:52), so that RT ultimately “constructs ‘what was said’ (or

thought) for the speaker’s own interactional purposes” (Lamerichs and te Molder 2009:402). For

participants in interaction, RT is thus produced and treated not as an isolated phenomenon, but as

one in a string of actions, reported or represented in different ways, where RT “is coordinated

with other forms of reporting to create contexts for each other” (Rae and Kerby 2007:192).

Whilst speakers do work to prepare the context for RT in ways that indicate to the

4

recipient how the RT is to be understood, such preparation is not always successful and

recipients do not always orient to RT in the way presumably intended (Holt 2000). But speakers

may themselves signal that the contextual information preceding RT was not adequately

designed for the current interactional purpose. Thus, speakers who are about to produce RT or

are already in the midst of doing so sometimes abandon its production temporarily, to include1

other material before subsequently returning to and producing the projected RT. By displaying to

the recipient what contextual information was relevantly missing at the point at which RT was to

be produced, participants also make this matter available for analytic scrutiny, allowing us as

researchers to explore what participants deem relevant for their delivery of RT and why. It is

these matters and the resources employed to accomplish them that we will focus on in this paper.

The paper is organized as follows: We first present a brief description of the data that

forms the basis of our analysis (Section 2). We then identify three sequential positions in which

RT is temporarily abandoned to include other material (Section 3) and describe the resources that

are employed in the service of making recognizable to the recipient that what is about to be

produced is not the projected RT (Section 4). We then demonstrate how this inclusion of

additional material before the projected RT is done in order to recalibrate the larger context of

RT and thus ensure that the recipient of RT is able to interpret the speaker’s underlying

interactional project (Section 5). We conclude with a comparative discussion of other practices

that are employed by participants in interaction to adjust the design of their on-going talk.

2. Data and methodological approach

5

Our study draws on approximately 32 hours of audio- and video recordings of naturally

occurring interactions in Danish, some of which can be found at the publically available

talkbank.org/SamtaleBank, while others come from a private database. Both sets of data have

been collected according to the relevant Danish ethical standards, i.e. with informed consent. The

32 hours of recordings consist of 107 interactions, both everyday and institutional, on the

telephone and face-to-face. Our initial procedure was to identify all instances of RT in this data,

after which we formed a smaller sub-collection in which RT was temporarily abandoned. The

ratio between the two is about 20:1, i.e. it is relatively rare occurrence for speakers to

temporarily abandon RT, something which further supports the observation that speakers usually

manage to prepare the context for the production of RT sufficiently for the recipient to

understand the interactional project within which it plays a part. The following analysis is based

on a total of only 15 cases of temporarily abandoned RT. All examples have been transcribed

following Jefferson (2004) and anonymized (where relevant) for publication. In each example

we have marked the initial projection or production of RT with ->1 and the place in which RT is

finally produced with ->2. Both the initial projection of RT and its completion are in addition

highlighted in bold typeface.

3. Three sequential positions for the temporary abandonment of RT

The temporary abandonment of RT occurs in three positions in our Danish data: (a) just after RT

has been projected, but not yet produced; (b) within a turn constructional unit that constitutes

6

RT; and (c) after a turn constructional unit of RT has reached completion but more RT is

projected. The following three examples illustrate the three positions. While we provide the full

interactional context in which the RT is produced in each example, the purpose of this section is

only to identify where RT is initiated, abandoned, and returned to. Each of the three examples

will be discussed in more detail in subsequent sections.

3.1 Abandonment of RT after a quotative

The first sequential position at which RT can be (temporarily) abandoned is after the production

of a quotative. The following example is an illustration of this. Mathias is relaying his

achievements in a computer game to his friend Malte. In the game, Mathias played Genghis

Khan who conquers Persia by using a Trojan horse trick. Mathias first explains how his character

in the game disguised his warriors as ‘trade cards’2 to get past the Shah’s guards (lines 04–12),

then how the Shah clearly understood the trade cards as gifts to ensure he caused the Mongolians

no further harm (lines 13–18), and finally how the warriors jumped out of the trade cards and

slayed the Shah (lines 19–24).

(1) Trade cards (TH/S2/19, telephone, everyday)

04

05

06

07

08

09

Mathias:

->1

->2

>.hh der har man så to træjt kards< there have one then two trade cards.hhh ogeh så- >så ska' man så bringe and then then shall one then bringdem hen te' ham der< sharen, .hh å' såthem over to him there shah-DET and thensi'r han s- >så si'r< eh d- de førstsays he then say they firstfordi de ska' forbi alle vagternebecause they shall pass all guards-DET.hh Jameneh vi kommer me' gaver te' yes-but we come with gifts to

7

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

Malte:

Mathias:

Malte:

Mathias:

Malte:

Mathias:

den store shar'n ikke,the big shah not

.hh then you have two trade cards<

.hhh andeh then- >then you need to bringthem over to that guy< the Shah, .hh and thenhe then says >then say< eh t- they firstbecause they need to get past all the guards.hh Yesbut we’re bringing gifts forthe great Shah right,

(0.1)

[Jerh yes Yeah

[.hhh >å' da vi så kommer der hen< and when we then come there overHelt hen te' shar'n, .hh såehm rh- si'rwholly over to shah-DET then sayshan (0.3) Nåh. Hva' moneh mongolerne harhe INT What wonder mongolians-DET have>givet te' mig i dag så'n så< jeg ikk' given to me today like-this so I notsmadrer dem ikke, crush them not

.hhh >and when we then get there<All the way to the Shah, .hh thenehm rh- sayshe (0.3) Right. Let’s see what the Mongolians have>given to me today so that< I don't crush them right,

Erh.=yesYeah.

=>.hh så ud a' de der< træjt kards then out of those there trade cardsder hopper der to champions,=>De allerthere jump there two champions the allaller bedste fodfolkthe all best foot-people

=>.hh then out of those< trade cards jumptwo champions,=>The very very best soldiers

Jerh.=yesYeah.

=>.hh< å' begynder så bare å' smadre and begin then just to crushham der sharen,him there Shah-DET

=>.hh< and then just start crushing that Shah guy,

8

The temporarily abandoned RT delivered by the gift-bearers Jahmen vi kommer med gaver…

‘Yesbut we’re bringing gifts…’ is initiated with the quotative så siger de først ‘then say they

first’ in line 07, but only produced in line 09, separated from the quotative by the inclusion fordi

de skal forbi alle vagterne ‘because they need to pass all the guards’ (line 08). In line 07 Mathias

also abandons the delivery of the RT allegedly produced by the Shah (and then he says), to

instead report what the warriors said (then say they first). The inclusion of the word ‘first’ in this

second quotative indicates that the shift from what the Shah said to what the warriors said is an

instance of “reordering repair” (see Schegloff 2013, example [34]). The Shah’s abandoned RT is

re-initiated in lines 14–17, so that the gift-bearers’ RT in effect constitutes an inclusion between

the two. It is relatively typical that temporary abandonments of RT occurs in clusters (see also

Examples 3 and 4), something we discuss in more detail in Section 3.3.

3.2 Temporary abandonment within a turn constructional unit of RT

The second possible place to include material other than RT is when a first few units of RT have

been produced. This is the case in the following example, where Ester is telling her sister, Fie,

about an exchange between Ester’s friend, Trine, and Trine’s husband. Ester organizes this report

by first informing Fie that Trine is leaving her young son with his father when coming to visit

Ester in her summerhouse (not shown here). She then reports how Trine’s husband, Ulrik,

offered to look after the daughter as well. It is within this report that the temporarily abandoned

RT occurs, in lines 13–16.

9

(2) Nice guy (TH/M2/02, telephone, everyday)

13

14

15

16

17

18

19

20

21

22

23

24

25

26

Ester:

Fie:

Ester:

Fie:

Ester:

Fie:

Ester:

Fie:

Ester:

->1

->1

->2

->2

Men Ulr↑ik er så sød te' det >hanbut First-name is so sweet to that he [havde sagt te'< had said to

But Ulrik is so nice about that >he had said to<

[Nåhm' det' godt. INT-but that-is good Ohbut that’s good.

Trine . hhh hvorfor ta'r fordi hunFirst-name why take because shehavde sagt< Åhh jeg trænger så’n te'had said INT I need like-this to.hh å’ komme te' Rømø sam[men and come to Place-name together

Trine .hhh Why don’t >because she had said< OhI so need to .hh to go to Rømø

[Jerh yes Yeah

me' Ester å’ [så’n å’ så’nwith First-name and like-this and like-thiswith Ester and so on and so on

[Jerh yes Yeah

.hhh å’ så havde hun sagt å’ så and then had she said and thenve' jeg helst kun ha' Gitte med.will I rather only have First-name with.hh [Så siger han Hvorfor ta'r du ikk' then says he why take you not

.hhh and then she’d said and then I’d rather only bringGitte with me .hh Then he says Why don’t you

[°.Jerh° yes-INGR °.Yeah°

bare a' sted kun me' Ester.Just of place only with First-nameGo just with Ester.

Ester has both projected and initiated RT in lines (13–16), first by using the quotative han havde

sagt te’ Trine ‘he had said to Trine’, then by producing the first part of the RT Hvorfor ta’r ‘Why

take’. Instead of continuing the RT, Ester abandons it (line 16), includes extensive material

10

informing Fie of the context in which the RT was delivered, before returning to the RT in line

24, here re-projecting the RT with a new quotative Så siger han ‘Then he says’, after which the

full RT is delivered (lines 24–26).

3.3 Temporary abandonment after a turn constructional unit of RT

The third position in which RT can be temporarily abandoned is after the production of a first

turn-constructional unit of RT. In contrast to the two positions described so far, speakers cannot

in the same way be heard to be (temporarily) abandoning the RT, since a full turn of RT is in fact

delivered. If that RT constitutes a first pair part – for instance a question – the delivery of a

second pair part – for instance an answer – can however be said to have been projected (see Holt

2007:56, for similar observations about adjacency pairs in RT). If something other than that

second pair part is delivered next, this can thus be seen as an inclusion in the same way as

described above. The following example illustrates this. The example comes from another

telephone call between the two sisters Fie and Ester. Ester, who has recently moved, has just

informed Fie that her old dining table is now placed in her basement. She then reports on an

exchange she had with her son-in-law, Sten, when he inquired what Ester planned to do about the

table.

(3) Dining table (TH/S2/153, telephone, everyday)

05

06

07

08

Ester: ->1

->1

å'< .hh å' så'n al'så; Sten han sagdeand and like-this PRT First-name he saidNåhm' hva' me' dit spisebord, >du vedINT-but what with your dining-table you knowdet ville jeg jo ikk'- det ku' jegthat would I PRT not that could Iikk' ha' her vel,<not have here PRT

and< .hh and stuff; Sten he said Ohbut what about yourdining table, >you know I didn’t wantthat- I couldn’t have it here right,<

11

18

19

20

21

Ester:

->2

->2

->2

((ten lines omitted in which Ester complains abouthow Sten uses all the storage room in his house,leaving no space for Ester’s daughter))

=Så sagde han (0.8) Nåhmen det (.) then said he INT-but thatka' jeg da bare køre te' Nyhavn. Hva'can I PRT just drive to Place-name whatme' det spisebord, så si'r jeg M'with that dining-table then say I butdet ve' jeg beholdethat will I keep

Then he said (0.8) Ohbut that (.) can I surely justdrive to Nyhavn. What about that dining table, ‘then I say But that I want to keep

The RT allegedly delivered by Sten in lines 05–06 is brought to its possible conclusion

grammatically, prosodically and pragmatically, constituting a hearably complete turn in the form

of a question Nåhm’ hva’ me’ dit spisebord ‘Oh but what about your dining table’. Outside the

context of RT, such a question would constitute a first pair part that makes conditionally relevant

a second pair part in the form of an answer. While such conditional relevance might not apply to

the context of RT, the fact that Ester reports on Sten’s question does indicate that she will also

report her answer. Instead of doing so immediately, however, Ester includes material in which

she explains to Fie that there is no room for the table in her new apartment. Ester’s reported

answer is in fact only produced several turns-at-talk later in lines 20–21, after a complaint

sequence about Sten using all the storage space in the house he shares with Ester’s daughter (not

shown here). The fact that Ester here reproduces Sten’s reported question ‘What about that

dining table’ before reporting her answer further suggests that she herself is orienting to having

projected but abandoned the answer earlier in the sequence.

As the three examples above illustrate, speakers can abandon projected or initiated RT in

three different positions. The examples further illustrate that the included material varies, in

terms of both content and form, with some inclusions constituting just a single, clausal or

sentential turn construction unit, others consisting of several turns at talk. Moreover, the

12

examples show that speakers who abandon projected RT to include other material, may, but need

not, reproduce the quotative when returning to the RT. In the following we will consider first the

design of the inclusions (and the transitions between them and RT), then the interactional

function of these inclusions.

4. Transition into and out of the inclusion

In this section we take a closer look at the features employed by speakers to establish and make

recognizable that something is an inclusion rather than a piece of projected or already initiated

RT. The ability to distinguish between RT and other types of components within the same

sequence is a participants’ problem that applies more generally, because RT is often employed as

part of a larger narrative in which RT is just one representation of action within a string of

actions that constitutes the narration. Work on RT has recognized the participants’ “binding

problem” (Levinson 2006) in having to parse “the often fast-paced switches between those

narrative components grounded in the immediate context and those grounded in the context

being recounted” (Bangerter et al. 2011:187) and identified various resources that are employed

to mark the boundaries between RT and other material. These include prosodic features such as

voice quality, pitch, volume, and speed (Couper-Kuhlen 1998; Günthner 1999; Klewitz and

Couper-Kuhlen 1999); changes in dialect or language (Golato 2000); the use of various linguistic

items such as pronominal, temporal and locational references, verb tenses, interjections and

discourse particles (Bolden 2004); and gesture, gaze, and facial expressions (Sidnell 2009;

Streeck 2002). Here we consider how these resources are employed in relation to the

participants’ “problem” of making recognizable that something is not projected or initiated RT,

13

but an inclusion. Since the solution to this problem differs, depending on whether RT has merely

been projected or has already been initiated, we address the two alternatives in turn.

4.1 Transitions into and out of inclusions after RT has been projected but not initiated

Entry into RT is frequently marked prosodically, for instance by a step up in pitch, pitch resets

and changes in volume and speed of delivery (Bolden 2004; Klewitz and Couper-Kuhlen 1999).

In order for something to be heard not as the projected RT, but as an inclusion, speakers

consequently mark their inclusions as prosodically – and syntactically – integrated with the

previous units. Example (1), here presented in an abbreviated version, illustrates one way in

which speakers accomplish such integration.

(1) Trade cards (TH/S2/19, telephone, everyday) Abbreviated

07080910

Mathias: ->1

->2->2

>så si'r< eh d- de førstfordi de ska' forbi alle vagterne.hh Jameneh vi kommer me' gaver te' den store shar'n ikke,

>then say< eh t- they firstbecause they need to get past all the guards.hh Yesbut we’re bringing gifts forthe great Shah right,

Prosodically, Mathias here integrates the inclusion (‘because they need to get past all the

guards’) with the previous unit, the quotative ‘then say they first’, through producing the

inclusion with no delay, no obvious change in pitch and with the same speed. Syntactically, the

integration is accomplished through the subordinate conjunction fordi ‘because’ which explicitly

ties the inclusion to the preceding unit by a causal relationship, and through the pronoun de

‘they’ which shares the referent of the quotative. By contrast, when Mathias produces the

14

projected RT (line 09), this is formulated as prosodically and syntactically distinct from the

preceding units: the RT is preceded by an inbreath and initiated with the particle combination

ja+men ‘yes+but’, both as indicators that something new is now about to be initiated, and the

latter specifically indicating that it is a responsive action. In addition, the pronoun used is now in

the first person plural (vi ‘we’), thus shifting the perspective in the telling from narrating “about

them” as was done in the previous units to narrating “as we.” Finally, the RT is distinguished

from the preceding units by a pitch reset on the pronoun.

4.2 Transitions into and out of insertions after RT has been initiated

Once RT has been initiated, speakers face the task of distinguishing inclusions from the previous

unit, rather than integrating them. Examples (2) and (3), here presented in short, both illustrate

how this can be accomplished. In Example (2), the inclusion is placed at a point where the RT is

not yet complete, i.e. after the production of the first two units of the RT, hvorfor ta’r ‘why

don’t’. This RT is delivered with a very dynamic, animated pitch contour, whereas the

subsequent inclusion fordi hun havde sagt ‘because she had said’ is delivered in monotone, with

a much narrower pitch span and less dynamics than the speaker’s regular voice. Once the RT is

reintroduced in lines 23–24, the speaker returns to using the dynamic, animated voicing,

moreover demarcating the shift back to RT by reproducing the quotative.

(2) Nice guy (TH/M2/02, telephone, everyday) Abbreviated

14

1617

24

Ester:

Ester:

Ester:

->1

->1

->2

>han havde sagt te'<

Trine .hhh hvorfor ta'r fordi hunhavde sagt< Åhh jeg trænger så’n te'

.hh Så siger han Hvorfor ta'r du ikk'

15

26 ->2 bare a' sted kun me' Ester.>he had said to<Trine .hhh Why don’t >because shehad said< Oh I so need to.hh Then he says Why don’t yougo just with Ester.

A similar pattern of switching from animated, dynamic pitch movement in the RT and to

what can best be glossed as a “monotonous” delivery in the inclusion, can be found in Example

(3), where the inclusion is placed between one unit of RT and a next projected unit. Here, the

speech reportedly delivered by Ester’s son-in-law in lines 06 and 20, as well as Ester’s own

reported response in line 21, are animated and produced within broad pitch spans, whereas the

inclusion in which Ester explains the lack of space for the table in her new apartment is delivered

fast and with a very flat prosodic contour.

(3) Dining table (TH/S2/153, telephone, everyday) Abbreviated

05060708

Ester: ->1->1

Sten han sagdeNåhm' hva' me' dit spisebord, >du veddet ville jeg jo ikk'- det ku' jegikk' ha' her vel,<

Sten he said Ohbut what about your dining table, you knowI didn’t want that- I couldn’t have it here right,

18192021

Ester:->2->2

=Så sagde han (0.8) Nåhmen det (.)ka' jeg da bare køre te' Nyhavn. Hva'me' det spisebord, så si'r jeg M'det ve' jeg beholde

Then he said (0.8) Ohbut that (.) can I surely justdrive to Nyhavn. What about that dining table,then I say But that I want to keep

The examples above illustrate that speakers who (temporarily) abandon projected or

initiated RT to include other material do so in ways that show their orientation to the

participants’ problem of understanding that an inclusion is being produced in place of what was

otherwise projected to come. The resources employed by speakers to manage this problem are

the same, whether the inclusion is produced in place of RT that has merely been projected or in

16

place of RT that has already been initiated, and include both prosodic and syntactic means. But

the practical use of these resources differs depending on the position of the inclusion. When

inclusions are placed after a quotative, but before RT is produced, speakers face the task of

integrating the inclusion with the preceding unit, so as to mark it as something other than the

projected RT. By contrast, when inclusions are placed after a turn or sequence of RT has already

been initiated, speakers instead face the task of demarcating the inclusion from that of the

preceding units, i.e. the RT.

5. Recalibrating the context of RT with inclusions

Having established that, how, and where speakers introduce inclusions within a turn or sequence

of RT, we now turn to consider how such inclusions serve as a speaker’s resource to recalibrate

the context in which a particular piece of RT is to be produced, in order to meet particular

interactional projects and solicit the relevant kind of response from the recipient. Speakers may

address and recalibrate (at least) three different aspects of the context for RT through inclusions,

and we will consider each of these aspects in turn.

5.1 Recalibrating the context to introduce a source

A basic prerequisite of RT is that there is an (alleged) source being “quoted.” In referring to third

persons, however, speakers have available a range of different terms, the choice of which varies

depending both on whether and how the referent is known to the participants, as well as on the

larger interactional project within which the person referred to plays a part (Enfield and Stivers

17

2007). In the context of RT, the identification of the source being quoted is an important part of

preparing the context for RT and hence one aspect that may need to be addressed through an

inclusion. Example (4) is an illustration of this. This example comes from a televised debate. The

editor-in-chief (Seidenfaden) of a Danish newspaper has printed a book manuscript by the

former EU commissioner Ritt Bjerregård, without her permission and after she had withdrawn

the manuscript from the intended publisher. The example begins where the interviewer, Jersild,

has inquired how Ritt Bjerregård reacted when informed that the newspaper would print the

manuscript. In response to this inquiry, Seidenfaden produces the quotative hun siger ‘she says’,

but then corrects the reference by explicating that his source is the head of cabinet, Laurs

Nørlund (lines 19–23).3 Having identified the source of the RT, Seidenfaden then goes on to

once more project the RT-answer with the quotative han siger ‘he says’ in line 25, but abandons

the RT temporarily to include additional material, here to specify his special relationship with the

referenced source, before finally, in lines 29–30, producing the projected RT.

(4) Kabinetchefen (Seidenfaden, Face-to-face, Media interview)

18

19

20

21

22

23

24

25

Seide:

Jersild:

Seide:

Seide:

Jersild:

Seide:

->1

->1

h:un siger. (0.5) at eller hun sigershe says that or she saysingenting fordi jeg taler ikk med hende°nothing because I talk not with hermen hendes mand .hhhbut her man

she says (0.5) that or she doesn’t say anything becauseI don't talk to her but her spokesman .hhh

Laurs [Nørlund ikk?First-name Last-name not?Laurs Nørlund right?

[han- he-

°Laurs Nørlund. (.) kabinetche:fen.° First-name Last-name cabinet-boss-t Laurs Nørlund the head of her cabinet.

mmm,

han siger, (0.8) .hhh vi: i øvrigt

18

26

27

28

29

30

Jersild:

Seide: ->2

he says we-are by the-waygamle venner°, så det enold friends so it-is auformel samta:leinformal conversation

he says (0.8) .hhh we are by the way old friendsso it’s an informal conversation

mmm,

men han siger? .hhh (0.7) det der, (0.4)but he says this thereer øh gør Ritt Bjerregaard utilpas,is makes First-name Last-name uncomfortable

but he says hhh (0.7) the thing (0.4) that makesRitt Bjerregård uncomfortable,

The recalibration done by Seidenfaden may find its relevance in the specific context in

which the RT is to be produced. Seidenfaden already faces the risk of being subpoenaed by Ritt

Bjerregård (as he indeed was). Falsely quoting her in public may cause further problems. The

overt repair of the referent, in combination with the inclusion of his relationship with the true

source, Nørlund, may help Seidenfaden to portray himself as a journalist and editor who under

normal circumstances balances the journalistic tensions between ethics and truth perfectly, but

who in this specific case has weighed the importance of truth higher than that of behaving

ethically correct and respecting Ritt Bjerregård’s wish not to have her manuscript printed.

5.2 Recalibrating the context to organize a logical, true order of events

A second issue that is regularly addressed by inclusions is the recalibration of the order of

events, so that RT is eventually produced in the (alleged) original context in which it was

delivered. Such recalibration seems specifically concerned with establishing why the RT was

originally produced and in response to what, which in turn may have consequences for how the

original source of the RT is evaluated by recipients. Example (1) serves to illustrate how a

19

speaker, in recalibrating the order of events, may attempt to guide the recipient’s understanding

of those events.

(1) Trade cards (TH/S2/19, telephone, everyday) Abbreviated

07080910

13

14

15

16

17

Mathias:

Mathias:

->1

->2

Så si'r han s- >så si'r< eh d-de først fordi de ska' forbi alle vagterne.hh Jameneh vi kommer me' gaver te' den store shar'n ikke,

then says he t- >then say< eh t- they firstbecause they need to get past all the guardshh Yesbut we’re bringing gifts forthe great Shah right,

[.hhh >å' da vi så kommer der hen< and when we then come there overHelt hen te' shar'n, .hh såehm rh- si'rwholly over to shah-DET then sayshan (0.3) Nåh. Hva' moneh mongolerne harhe INT What wonder mongolians-DET have>givet te' mig i dag så'n så< jeg ikk' given to me today like-this so I notsmadrer dem ikke, crush them not

.hhh >and when we then get there<All the way to the Shah, .hh thenehm rh- sayshe (0.3) Right. Let’s see what the Mongolians have>given to me today so that< I don'tcrush them right,

Retrospectively, it should be quite evident that there are two features of Mathias’s telling

that are of great import to how his adventures within the game should be assessed: the general

indefeasibility of the Shah and the clever use of the Trojan horse trick. However, when Mathias

initiates the reported speech, he begins by projecting what the Shah said upon first receiving the

gifts (line 07), skipping the information about how the gift-bearers came to be in the Shah’s

vicinity, information which is crucial for capturing the exquisite duplicity of their behavior. By

initiating “reordering repair” (Schegloff 2013) Mathias shifts from reporting what the Shah said,

to what the gift-bearers said, but though this reordering portrays a more correct version of the

event, he has yet to include the fact that the gift-bearers had to get the hidden warriors past the

20

Shah’s guards. The RT projected in lines 07–08 with ‘then they say first’ risks being heard as

talk that the gift-bearers delivered to the Shah himself. By abandoning the projected RT and now

adding the information that the gift-bearers needed to pass the guards, Mathias recalibrates the

context further to specify that the original recipient of the RT was ‘the guards’, rather than ‘the

Shah’, thus indicating how the RT was in fact part of a larger cunning plan to trick both the Shah

and his guards. In this way Mathias manages to show the recipient of his story “that deception

was under way and how precisely the very words which had been used constituted deception,” in

a similar way to that described by Rae and Kerby (2007:194) for adolescent males who

interpolate the description of events and cognitive states with RT in order to convey their own

tactical behavior during the reported event. But whereas the adolescents reported on by Rae and

Kerby (2007) apparently set out from the beginning to design reports of their own clever or

devious actions in such a way that “that action is intelligible to their audience without making it

so obvious that its cleverness is lost” (192), Mathias’s initial path of telling his story was not so

designed and the cleverness and deviousness of his reported actions would have been lost had he

not temporarily abandoned the projected RT and included the explanation that his characters

needed to get past the guards.

5.3 Recalibrating the context to describe states of mind

A third matter that is frequently done in inclusions is to describe, specify or speculate about the

state of mind of one or more persons who allegedly participated in the event in which RT was

originally produced. As noted by Rae and Kerby (2007) such descriptions are an important

feature of a telling in which RT is produced, since they help to make the action that is being done

21

intelligible to the recipient. Speakers who are in the midst of producing RT may find that they

have not sufficiently indicated how the RT is to be interpreted and hence abandon the RT

temporarily to include a description of their own or a co-participant’s state of mind at the time at

which the RT was originally delivered. This is illustrated in Example (3), which is here provided

in an extended version.

(3) Dining table (TH/S2/153, telephone, everyday) Extended

05060708

Ester: ->1->1

Sten han sagdeNåhm' hva' me' dit spisebord, >du veddet ville jeg jo ikk'- det ku' jegikk' ha' her vel,<

Sten he said Oh but what about yourdining table, >you know I didn’t wantthat- I couldn’t have it here right,<

1819202122

23

24

35

36

37

38

39

40

41

42

43

44

Ester:

Ester:

->2->2->2

->2

->2

=Så sagde han (0.8) Nåhmen det (.)ka' jeg da bare køre te' Nyhavn. Hva'me' det spisebord, så si'r jeg M'det ve' jeg beholde Jeg troedehan .hh mente om vi sku' sætte dethe meant whether we should put iti den blå avis så sagde jeg Nahmenpaper-name then said I no-butdet beh↑older jeg.that keep I

Then he said (0.8) Ohbut that (.) can I surely justdrive to Nyhavn. What about that dining table,then I say But that I want to keep I thought he .hh meantwhether we should advertise it then I said No but I’ll k↑eep that.

((eleven lines omitted in which Ester provides herreason for keeping her dining table))

å' så: så sagde han Nåhmen så kørerand then then said he INT-but then drivingjeg te' Nyhavn me' den jeg ka'I to Place-name with it I canbare li:' (.) ta' bordet me' deropjust just take table-DET with there-upå' så ka' jeg ligge det op påand then can I lay it up onloftet over i (.) pumpehuset (eller)attic-DET over in pump-house-DET or.hh >så sagde jeg< Nej. Det ve' then said I no that willjeg ikk' ha'..hhh Jæ va- Hva' så, I not have what thenSå sagde jeg Jahmen så ve' jeg ha'then said I yes-but then will I havedet ned i min egen kælder.>.hh< Jait down in my own basement yesd*et var da* helt åndsvagt det sku'

22

45

46

47

48

Fie:

Ester:

that was PRT wholly stupid it shouldstå dernede å' fylde.=Nej sagde jegstand there-down and fill no said IDet ska' ikk'- .hhhh=it shall not

and then then he said Oh but then I’ll drive to Nyhavnwith it I can just (.) bring the table there and thenI can put it in the attic in (.) the pump house (or)hh >then I said< No. I don’t want that. .hhh Yes wh-What then, Then I said Yes but then I want it down inmy own basement. >.hh< Yes *that was certainly* totallystupid that it should be down there taking up space.=No Isaid It shan’t- .hhh

=N*[ejh:* >.h< Det bli'r ås' no it become also=N*oh:* >.h< It also becomes

[Han ble' v↑ed, å' ble' v↑ed he became on and became on He kept ↑on and kept ↑on,

Ester’s report of her exchange about the dining table with her son-in-law can be heard as

an illustration of the son-in-law’s kind nature; having heard that Ester wants to keep the table but

has no room for it (lines 20 and 24), he selflessly offers to store the table in his summer house in

Nyhavn, even transporting it to the storage place himself (lines 18–19 and lines 35–39). We can

see from the conclusion of this example (line 48), in which Ester reports that Sten ‘kept on and

on’ about the table, however, that she does not consider his solution and offer a kindness. In

order to indicate a less benign understanding of Sten’s reported actions and talk, Ester engages in

a number of recalibrating inclusions. First, she reorders the event to show that Sten’s offer (first

reported in lines 19–20) was in fact made in response to Ester’s expressed wish to keep the table

(line 21). Having reported this wish as an answer to Stens’ inquiry (‘What about that dining

table’, line 20), Ester does not return to the offer, but instead reports to Fie what she was thinking

when being asked about the dining table, namely that Sten was asking because he wanted to

know whether they should advertise the table for sale in a newspaper (lines 21–23). By

describing her own thinking and repeating her reported answer after this description, Ester

indicates that at this point she had not herself realized that there was anything ominous about

23

Sten’s inquiry and thus, also, that she perfectly innocuously replied to this inquiry. Having

delayed the reproduction of Sten’s offer by an extended account of why she wants to keep her

dining table though she cannot use it in her new apartment (not shown here), in lines 35–39 Ester

once more reports on Sten’s “offer.” At this point, however, Ester has, through her inclusions,

recalibrated the order of events and described her own thinking while the exchange was on-

going, so that the context in which Sten’s offer is made is now significantly different from when

it was initially reported, and Ester’s reported rejection (in line 40) now comes off as perfectly

reasonable. The remainder of the reported exchange (lines 41–48) serves as an illustration of

Sten’s impertinent persistence in wanting to store Ester’s table in his summer house rather than

letting her store it herself in her own basement.

The examples in this section illustrate that by abandoning RT and instead producing an

inclusion, a speaker can in various ways recalibrate the context in which RT is delivered, so as to

help, guide and in other ways indicate to the recipient of the RT how this is to be interpreted and

responded to. Our data suggests at least three different matters that can be addressed in such

recalibrations: the source of the RT, the logical or real order of events in which the RT was

(allegedly) produced, and the state of mind of the participants reported to be part of an RT

exchange. Example (3), along with examples (1) and (4) furthermore illustrate that inclusions

and recalibration often occur in clusters, so that in the few cases of RT in which inclusions are

produced, we frequently find more than one such inclusion, placed at different positions in

relation to the RT and doing different jobs in terms of recalibrating the context. This points to the

possibility that recalibration through inclusion is occasioned by the speaker’s realization that he

or she had not from the beginning designed the context of RT adequately for its interactional

24

purposes and that once such a flawed RT is underway, getting it exactly right might take more

than one try.

6. Discussion and conclusion

In this paper we have considered a small number of cases in which projected or initiated RT is

temporarily abandoned to include other material. We have identified three different positions in

which this may occur: either right after RT has been projected, after RT has been initiated but not

completed, and after one turn of RT has been delivered but the projected next RT turn is yet to be

produced. We then described some of the prosodic and syntactic resources which speakers

employed to distinguish the inclusions from the projected or initiated RT. Finally, we illustrated

that a speaker who is about to produce some piece of RT may temporarily abandon it to

recalibrate aspects of the context in which the RT is to be delivered, for instance who is the

quoted source of the RT, where the RT was originally produced in a larger order of events and

what the original RT producer or recipient of RT was thinking or feeling at the time of delivery.

The fact that speakers may include or in other ways produce other material than that

which was projected by their immediately preceding talk is of course not particular to RT

sequences but is something that speakers do in a range of situations and for a range of

interactional purposes. In his overview of ten different operations for self-initiated, same-turn

repair, Schegloff (2013) argues that one outcome of repair is to alter an ongoing turn-at-talk “in

some interactionally consequential way” (43). In particular the repair operation that Schegloff

names “parenthesizing” seems reminiscent of our RT-inclusions, as they “add to the turn-in-

progress something other than a next-due element” and are “ordinarily composed of clausal

25

TCUs” (51).4 Similarly, Mazeland (2007) describes how what he terms “parentheticals” are used

to halt an ongoing turn constructional unit to insert a short remark before returning to and

producing the remainder of the halted unit. The parentheticals described by Mazeland share

many of their features with the RT inclusions we have described here: like inclusions,

parentheticals are typically syntactically and prosodically distinctive from the otherwise

projected talk; moreover, parentheticals are employed for the purpose of “providing background

knowledge that is relevant to understand the action that is being done” (1836). As we have

sought to demonstrate, the recalibrations of RT-contexts we have explored appear to be

implemented in order to bolster the recipient’s understanding and interpretation of the reported

event in which RT occurs. While we are loath to suggest that these recalibrations are repair

operations, the similarities between Mazeland’s “parentheticals,” Schegloff’s “parenthezising”

and our “inclusions” – in terms of both positioning, syntactic, and prosodic features and overall

interactional function – to us suggests at least a family resemblance among the three, as similar

practices for accomplishing what could be glossed as “recalibrating talk-in-progress for

interactional purpose.”

References

Bangerter, Adrian, Eric Mayor, & Simona Pekarek Doehler. 2011. Reported speech in

conversational storytelling during nursing shift handover meetings. Discourse Processes

48(3). 183–214.

Bolden, Galina. 2004. The quote and beyond: Defining boundaries of reported speech in

conversational Russian. Journal of Pragmatics 36. 1071–1118.

Buttny, Richard. 1998. Putting prior talk into context: Reported speech and the reporting context.

26

Research on Language and Social Interaction 31(1). 45–58.

Clark, Herbert H. & Richard J. Gerrig. 1990. Quotations as demonstrations. Language 66 (4).

764–805.

Couper-Kuhlen, Elizabeth. 1998. Coherent voicing. On prosody in conversational reported

speech. In Wolfram Bublitz, Uta Lenk & Eija Ventola (eds.), Coherence in spoken and

written discourse. How to create it and how to describe it, 11–33. Amsterdam: John

Benjamins.

Edwards, Derek. 1995. Two to tango: Script formulations, dispositions, and rhetorical

asymmetry in relationship troubles talk. Research on Language and Social Interaction

28(4). 319–350.

Enfield, Nick J. & Tanya Stivers (eds.). 2007. Person reference in interaction: Linguistic,

cultural and social Perspectives. Cambridge: Cambridge University Press.

Golato, Andrea. 2000. An innovative German quotative for reporting on embodied action: Und

ich so/und er so ‘and I’m like/and he’s like’. Journal of Pragmatics 32. 29–54.

Günthner, Susanne. 1997. The contextualization of affect in reported dialogues. In Susanne

Niemeier & René Dirven (eds.), The language of emotions: Conceptualization,

expression and theoretical foundation, 247–275. Amsterdam: John Benjamins.

Günthner, Susanne. 1999. Polyphony and the “layering of voices” in reported dialogues: an

analysis of the use of prosodic devices in everyday reported speech. Journal of

Pragmatics 31(5). 685–708.

Holt, Elizabeth. 1996. Reporting on talk: The use of direct reported speech in conversation.

Research on Language and Social Interaction 29(3). 219–245.

Holt, Elizabeth. 2000. Reporting and reacting: Concurrent responses to reported speech.

27

Research on Language and Social Interaction 33(4). 425–454.

Holt, Elizabeth. 2007. ‘I’m eyeing your chop up mind’: Reporting and enacting. In Elizabeth

Holt & Rebecca Clift (eds.), Reporting talk: Reported speech in interaction, 47–80.

Cambridge: Cambridge University Press.

Jefferson, Gail. 2004. Glossary of transcript symbols with an introduction. In Gene H. Lerner

(ed.), Conversation Analysis: Studies from the first generation, 13–31. Amsterdam: John

Benjamins.

Klewitz, Gabriele & Elizabeth Couper-Kuhlen. 1999. Quote-unquote? The role of prosody in the

contextualization of reported speech sequences. Pragmatics 9 (4). 459–485.

Lamerich, Joyce & Hedwig F. M. Te Molder. 2009. ‘And then I’m really like…’:‘Preliminary’

self-quotations in adolescent talk. Discourse Studies 11(4). 401–419.

Levinson, Stephen C. 2006. On the human “interactional engine”. In Nick J. Enfield & Stephen

C. Levinson (eds.), Roots of human sociality: Culture, cognition and interaction, 39–69.

Oxford: Berg.

Mazeland, Harrie. 2007. Parenthetical sequences. Journal of Pragmatics 39. 1816-1869.

Pomerantz, Anita. 1986. Extreme case formulations: A way of legitimizing claims. Human

Studies 9. 219–229.

Rae, John & Joanne Kerby. 2007. Designing contexts for reporting tactical talk. In Elizabeth

Holt & Rebecca Clift (eds.), Reporting talk: Reported speech in interaction, 179–194.

Cambridge: Cambridge University Press.

Schegloff, Emanuel A. 2013. Ten operations in self-initiated, same-turn repair. In Makoto

Hayashi, Geoffrey Raymond & Jack Sidnell (eds.), Conversational repair and human

understanding, 41–70. Cambridge: Cambridge University Press.

28

Sidnell, Jack. 2009. Coordinating gesture, talk and gaze in reenactments. Research on Language

and Social Interaction 39 (4). 377–409.

Streeck, Jürgen. 2002. Grammars, words, and embodied meanings: On the uses and evolution of

so and like. Journal of Communication 52(3). 581–596.

Wilkinson, Sue & Ann Weatherall. 2011. Insertion Repair. Research on Language and Social

Interaction 44(1). 65–91.

29

1We use “include” as a general gloss for a speaker producing verbal material other than what was

projected. Our original gloss was “insert,” but as one anonymous reviewer noted, this is also a

technical term that denominates a particular type of repair (Wilkinson and Weatherall 2011;

Schegloff 2013). As we do not intend to claim that the phenomenon we describe here is a repair

operation, we have chosen the less technical term “include.” We will, however, discuss the

similarities and differences between our phenomenon and other related practices, including that of

repair, in Section 3.

2 ’Trade cards’ are – as we understand it – one part of the computer game: in order to get to see the

Shah the player first has to earn these ’trade cards’ by accomplishing certain tasks.

3Seidenfaden’s shift from one referent to another appears to be an instance of “replacing repair”

(Schegloff 2013) rather than the “reordering repair” that Mathias employed in Example (1).

4In fact, Example (15) in Schegloff (2013:52) seems to concern the temporary abandonment of RT

that has already been projected, just as is the case for our Example (1). See also Wilkinson and

Weatherall (2011), footnote number 7, for a similar example of a “parenthetical” repair that

involves RT.