Repeated use of request for confirmation in atypical interaction.

This study investigates a specific method for making possible the participation of participants with cognitive and communicative impairments in social face-to-face interaction. Non-impaired co-participants design close-ended questions that project who the next speaker is, i.e. the impaired co-participant. The questions also project what kind of response amongst alternatives the impaired co-participant is supposed to produce. Upon answers to these questions, the non-impaired co-participant requests the impaired participant to confirm the answer twice. Using conversation analytic (CA) methods, the study scrutinises what is achieved by requesting a confirmation of the provided answer - repeatedly so. The study argues that the practice may put the (deficit) competence of the participant with impairments in focus if the initial close-ended question works to establish an understanding of a prior action by the participant with impairments.

Introduction

At the heart of ordinary human life, we find social interaction. Yet, organisations such as the disability organisation of Denmark, Danske Handicaporganisationer, or Disability Rights UK point out that certain social structures may hinder some citizens from participating. These organisations, alongside researchers in the area of Disability research (Rapley, 2004; Robillard, 1999) and in medical care (WHO, 2013), have achieved political resonance and efforts are being made to empower these citizens; for instance by compensating for impediments of locomotion through installations of wheelchair ramps and for reading disabilities through making societal information available in various media and modalities such as Braille. The majority of these efforts are categorisable as external, contextual factors for (WHO, 2013) and ‘pre-conditions’ of participation. Thus, possibilities for participation are treated as being created and provided prior to and external to the concrete participation in social interaction and social life.

This political and scientific initiative towards the social sphere differs from the primary research focus within Ethnomethodology and Conversation Analysis (EMCA) (Sacks and Garfinkel, 1970). Since EMCA was founded in the late 60s/beginning of the 70s, research in this field has put members’ concrete participation in ordinary everyday interaction under scrutiny (Sacks et al., 1974). The vast amount of studies on everyday interaction has shown how participants conduct their ordinary everyday affairs through the detailed coordination and organisation of actions-for-interaction that are constructed by the coordination of available resources such as talk, gestures, bodily movements, gaze etc. (Depperman, 2013, Rasmussen et al., 2014). For the topic of this special issue, it is of particular interest that EMCA and EMCA related research describe how possibilities for participation are actually created and achieved by co-participants in and through the coordinated details of their social face-to-face interaction (Goodwin and Goodwin, 2004; Sacks and Garfinkel, 1970).

EMCA research has also shown how participants achieve participation in interactions that involve participants with different kinds of diagnosed impairments, including communicative ones (Bloch and Wilkinson, 2004; Brouwer, 2012; Laakso, 1999; Klippi, 2003; Pilesjö and Ramussen, 2011). This specific branch of research also investigates the challenges that participants in these interactions face as they rely on, and presuppose, a system that includes sequentially ordered talk as a central resource (Rasmussen, 2013; Wilkinson, 1999).

This article is a contribution to the latter line of investigation. It aims at investigating an interactional phenomenon, ‘repeated use of request for confirmation’ (RUR) upon answers to close-ended questions (Robinson and Heritage, 2005) through which non-impaired co-participants strive to settle or to nail down (compare to Liberman, 2011) the understanding, wishes, or needs of persons with communicative and cognitive impairments (PWCI). The phenomenon is accomplished by the use of talk and other bodily resources. The use of close-ended questions and practices of having confirmed that PWCI understands the questions and furthermore, that the non-impaired co-participant understands the response provided by PWCI are recommended in the clinical field (Bjerre, 2006). The aim of this practice, amongst others, is to create the possibility for PWCI to participate and be involved in social life on equal terms with non-impaired co-participants.

The analysis aims at describing what is achieved through the practice and how this is accomplished. Hence, the analytic focus is on the interactional phenomenon as it emerges across different types of cognitive and communicative impairments and settings, rather than on the frequency with which the phenomenon is found in specific types of settings involving specific types of impairments. The analysis is based on a collection of instances. It indicates that RUR is systematically used and recognisable in interactions involving participants with cognitive and communication disorders.

The analysis will reveal that the phenomenon is characterised by the non-impaired co-participant posing a close-ended question that receives a relevant answer by the PWCI. These two actions (1. question and 2. answer) are organised in terms of an adjacency pair (Schegloff, 2007), which is a small sequential exchange, in which the first part projects what is relevant as the second part as well as when the second part is relevantly produced and by whom. In addition, the first pair part may point specifically at one answer amongst relevant alternatives, i.e. it may point at a preferred answer (Pomerantz and Heritage, 2013). According to Schegloff (2007), preferred responses are more likely to close a sequence than are dispreferred ones. However, RUR is characterised by the fact that PWCI’s preferred answer to the co-participant’s close-ended question occasions several rounds in which the co-participant requests the PWCI to confirm the answer. The co-participant’s requests and the PWCI’s confirmations are also organised in the form of adjacency pairs with in-built preference structures. The speaking co-participants initiate then no less than three adjacency pairs before closing the RUR sequence:

1. Close-ended question

2. Preferred answer

1. First request for confirmation

2. Preferred answer (confirmation)

1. Second request for confirmation

2. Preferred answer (confirmation (+account))

The first part of the article concerns specifically a) what is actually achieved through requesting a confirmation upon receiving a preferred answer to a close-ended action? And b) what is achieved through requesting yet another confirmation upon having received the first (preferred) confirmation? The second half of the article addresses c) what are the sequential consequences of the use of RUR? It contrasts two different types of interactional work that is accomplished by the initial close-ended question: 1. The clarification or specification of a prior action conducted by PWCI and 2. the introduction of a new topic. The analyses in this part of the article indicate that the different types of work seem to occasion different types of RUR outcome. Finally, before summarising and concluding the study, the article discusses in which ways the use of RUR can be seen as local possibilities for participation and involvement in social interaction.

Dataset and methods

The study draws on data in the form of face-to-face interactions in Danish between PWCIs and their relatives or speech-language pathologists (SLPs) . This specific study involves five adults who suffer from cognitive impairments, caused by an injury to the brain from a stroke. The cognitive impairments include (mild, moderate or severe) problems in remembering or understanding. The participants are also diagnosed with aphasia. Additionally, the study draws on two interactions involving two children/pre-adolescents, one of which is a boy diagnosed with Cerebral Palsy, and a severe intellectual impairment, e.g. problems in performing simple tasks and in understanding.

The data has been collected partly by myself, partly by students who were enrolled in the course Language, Cognition and Communication at the University of Southern Denmark. I thank all of them for helping me building up the corpus. All rights are with me.

An example from an interaction involving him will be presented. The total number of co-participating relatives and SLPs are 13. The interactions are carried out in institutional or private settings, and consist of various types of therapeutic activities as well as ordinary everyday conversations. As mentioned, RUR is similar across settings and populations.

The description of RUR upon preferred responses to close-ended questions is based on 17 instances. Five examples will be presented (1, 2, 3, 4, 5). The examples are divided into excerpts (1.1; 1.2; 1.3. 2.1; 2.2. 2.3; 3.1; 3.2; 3.3. and so forth). So, for instance, excerpts (1.2) and (2.2) are continuations of (1.1) and (2.1) respectively, whereas (1.3) and (2.3) are continuations of (1.2) and (2.2), respectively, and so forth.

Ethnomethodology and conversation analysis

The EMCA framework aims at capturing members’ perspective and understanding of social life. The aim is moreover to capture how members achieve an understanding inter alia of their everyday social affairs and how they establish their understanding as a social fact (Liberman, 2011). The focus is thus on members’ methods.

In studies of face-to-face interaction the EMCA interests concern specifically how members achieve an understanding inter alia of their actions, so-called ‘conversational objects’ (Sacks, 1995). The interests concern for example the methods used for organising and understanding an object as ‘a joke’ or ‘a question’, ‘an answer’, ‘a confirmation’ or ‘a request for confirmation’. Specifically, research interests concern participants’ actions as recognisable methods, which are socially (sometimes even normatively) established as that kind of method or action (Schegloff et al., 1977). In investigating the objects, EMCA researchers aim to take every possible detail into account, and not to treat any element as irrelevant to the interactional phenomenon a priori (Sacks, 1984).

For the purpose of analysis, EMCA analysts employ specific methods (ten Have, 2007): a) the data at hand are transcribed, including at best all details; b) analysis of the details of the immediate previous micro-context in which an action occurs; c) analysis of the details of the action in focus; d) analysis of details of how the action is treated by co-participant in a next turn (a so-called next turn proof procedure); e) building a collection of instances; f) analysis of similarities and differences of a) and b) in the collected instances; and g) analysis of deviant cases.

These methods have been employed for the study of RUR as presented in this article.

Transcription

The data were transcribed in accordance with the Jeffersonian transcript system (Atkinson and Heritage, 1984) to which the following notations have been added.

» Gaze towards/at

»« Mutual gaze

// Co-occurring resources for interaction in building a turn

A gloss is provided in a separate line. Unintelligible talk or vocalisations are though cited in the line:

PWCI3 #jah# (.) /n: #å: #h n(.)hoi:

yeah ‘n: #å: #h n(.)hoi:’

Analysis

When members of society interact, they take ordinary ways of organising the interaction and hence ways of achieving a common understanding for granted. The self-evident truth of sense-making methods is challenged in interactions in which a co-participant is not able to use one of the resources, which is treated as a, if not the most, fundamental part of sense-making, i.e. language (Rasmussen, 2013). It is tested even more so, if a co-participant has, presumably, also problems in understanding language or in understanding social actions due to cognitive impairments. As a consequence participants design methods (Drew, 2013) that are suited for this kind of interaction either by developing alternative methods or by selecting specific ordinary ones (Clarke and Wilkinson, 2007, 2009). In the data at hand, one method for dealing with this is for the speaking participant to employ close-ended questions, eventually two consecutive ones. Not only do these questions give PWCI a choice in the form of alternatives (Antaki et al., 2008), by designing them as close-ended questions or actions, the non-impaired participant may take into account PWCI’s sometimes very strong language impairments, which may hinder him or her in producing responses composed of more than a ‘yes’ or a ‘no’ (compare to Steensig and Heinemann, 2013). ‘Yes’ and ‘no’ namely conforms to the constraints laid out by the grammatical form of the close-ended question (Raymond, 2003).

Contextual environment of RUR

Close-ended actions form, as mentioned, the micro-context in which RUR may occur, making a limited range of next actions relevant (Robinson and Heritage, 2005). In example 1, excerpt 1, R1 asks his cognitively impaired older brother, PWCI1, which character he wants to be in a Play Station game before they actually play the game:

Excerpt 1.1

1 R1 → vil du være ha- ↑hende?do you want to be hi- her

2 PWCI1 j:aa

yes

»monitor

The question in line 1, do you want to be her, is constructed as a question which projects a specific class of answers (Raymond, 2003), ‘yes’ or ‘no’, as alternatives in a relevant next action (Schegloff and Sacks, 1973; Schegloff et al., 1977). It also projects who is supposed to deliver the answer, i.e. PWCI1, and when, i.e. upon the question having been terminated. PWCI1 orients to these projections as he responds with a yes.

In example 2, excerpt 2.1, a close-ended question is followed by another one, which in itself presents an alternative to the first one. In the example, the SLP1 and PWCI2, who is diagnosed with aphasia and with dyslexia, are engaged in a therapeutic activity in which PWCI2 has accomplished the task of matching pictures with texts (short sentences). PWCI2 speaks but with the two (conventional Danish) words ‘ja’ (yes) and ‘nej’ (no). Upon completion of the task, the SLP1 asks PWCI2 how he managed to solve the task. Hence, she asks him if it is part of his strategy to read the whole sentence, do you read the whole sentence (line 1).

Excerpt 2.1

1 SLP1 → læser du he:le sætningen?

do you read the whole sentence

»PWCI2

points at card with index finger

2 (0.7)

3 PWCI2 dub bi dub dub[/dub dub

‘dub bi dub dub dub dub’

»card, /»SLP1

4 SLP1 → [eller læser du (.) /et af ordene.

or do you read one of the words

»cards     /»«PWCI2

5 (0.5)

6 PWCI2 nej

no

nods

As one sees, PWCI does not respond to SLP’s question in line 1 with an answer amongst the projected class of alternatives yes or no. Instead, he responds with unintelligible vocalisations, dub bi dub dub dub dub (line 3). Instead of seeking to make sense of PWCI’s response, SLP poses another close-ended question (line 4), which constitutes an alternative to the first one. Through the alternative, she pursues then the specific (and restricted) class of answers that was made relevant through the first question in line 1, a yes or a no answer. After a pause PWCI responds with a no (line 6), which is combined with nodding.

The environment may consist of other types of close-ended questions. In example 3, excerpt 1, the daughter (R2) suggests a candidate understanding (Antaki, 2012) of a movement that was produced in a prior action by her mother (PWCI3). The mother is diagnosed with aphasia and as PWCI2 she speaks but the two (conventional Danish) words ‘ja’ (yes) and ‘nej’ (no). The two of them are sitting at a kitchen table having a cup of coffee. The daughter brought cake, wrapped in cellophane, for the coffee table. The cake is finished and the cellophane lies on the table. The mother picks up the cellophane, holds it in up position, and gazes at her daughter, who responds as follows (line 1):

Excerpt 3.1

1 R2 → så er det man laver g(.)ække↑breve

and then one makes Easter letter riddles

»«PWCI3

2 PWCI3 #jah# (.)   /n: #å: #h n(.)hoi:

yeah ‘n: #å: #h n(.)hoi:’

»cellophane/»R2, »downwards

holds cellophane in up position

The daughter proposes, in line 1, how to understand the movement that the mother made as she lifted the cellophane that was wrapped around the cake: she, the mother, initiated talk about Easter letter riddles, which she supposedly indexes through the cellophane. In the Danish culture mostly children write anonymous letters containing riddles to (grand) parents at Easter. The number of letters in the name of the writer is indicated by an equivalent number of dots. In many families, it is tradition that the parents buy chocolate (in the form of the Easter bunny or eggs), if they do not guess the identity of the writer within a specific number of days. They don’t. Now, the daughter produces the suggested understanding as a declarative question (Stivers, 2010). Furthermore, by stressing then and Easter letter riddles (så og g(.)ække↑breve) and by raising the intonation (g(.)ække↑breve) the (grown-up) daughter treats the event as ‘exciting’/‘good’. Moreover she takes an affective stance towards it (Stivers, 2008). With the declarative question she, like the non-impaired co-participants in examples 1 and 2, makes a specific class of responses relevant in the form of a yes or a no.

The possible relevant answers to the close-ended actions in all examples are, however, not simply alternatives. The polar (opposite) yes/no questions and actions (Boyd and Heritage, 2006) are structured in ways, which point stronger at one rather than the other. In this CA technical non-psychological sense, the actions have an in-built ‘preference for’ one of the responses (Bolinger, 1978; Pomerantz, 1988; Raymond, 2003; Sacks, 1987), i.e. a confirmation in terms of a ‘ja’ (yes).

In excerpt 1.1, the speaker (self) repairs (Schegloff et al., 1977) an emerging question ‘vil du være ha-’ (do you want to be hi-) to ‘↑hende?’ (her). The repair, her, is produced with a pitch-reset (↑) through which her is being highlighted and thus pointed at as the preferred answer. As we saw, the brother (PWCI1) provided the answer in his subsequent turn. In excerpt 3.1 the daughter (R2) produces the turn (line 1) as a statement question (and then one makes Easter letter riddles) that offers a candidate understanding that is structurally designed to prefer its confirmation, yes, in the next turn. Furthermore, through the affective stance taking, the daughter’s action prefers an affiliating positive confirmation of her stance. Her mother delivers both in line 2: the confirmation of the question in terms of a yeah; the affiliating stance through the vocalisations or sound objects (Reber, 2012) ‘n: #å: #h n(.)hoi:’.

In excerpt 2.1, the client (PWCI2), does not deliver what his SLP´s question pointed at. The question or do you read one of the words is grammatically designed, in terms of the polarity of the interrogative (Horn, 1989), to provide for alignment in terms of a ‘yes’. In line 6 the client, however, provides a non-confirmative answer, no. This leads the SLP1 to conclude that the client then reads the whole sentence. If not one of the alternatives (or do you read one of the words), it is concluded that it is the other (do you read the whole sentence, line 1). Also, this subsequent concluding action is designed grammatically to provide for a positive confirmation, a ‘yes’, which the client delivers in line 8:

Example 2.2.

7 SLP1 hele sætningen

the whole sentence

8 PWCI2 → ja

yes

First request for confirmation

In the interactional phenomenon of interest here, sequences in terms of preferred responses to close-ended questions are expanded. The subsequent action is though not what Schegloff (2007) has found to be sequence-closing thirds (like okay or yes), which are designed to possibly finish the sequence. Nor are they elaborations by PWCI of his or her response (compare to Steensig and Heinemann, 2013). Instead, the non-impaired co-participant requests a confirmation of it. The request constitutes a new first pair part. It is constructed as re-workings, re-doings or enhancements of the sequence initiating close-ended questions (Schegloff, 2007) by reference to its response (i.e. its second pair part). It may also be built by a tag question (Sacks et al., 1974) or through what Jefferson (1981) described as ‘post-response response solicitations’. In example 3, excerpt 3.2, a tag question (line 3) occurs upon the preferred affiliating response provided by the mother (PWCI3), yeah ‘n: #å: #h n(.)hoi:’ (line 2):

Excerpt 3.2 (lines 1 and 2 replicated)

1 R2 →   så er det man laver g(.)ække↑breve

»«PWCI3

and then one makes Easter letter riddles

2 PWCI3   #jah# (.) /n: #å: #h n(.)hoi:

yeah ‘n: #å: #h n(.)hoi:’

»cellophane/»R2, »downwards

holds cellophane in up position

3 R2 ik os

right

smiles, eyebrows lifted

»«PWCI3

4 PWCI3   °jo.h°

yes

»«R2, »cellophane

smiles, puts cellophane down

The action, right, (line 3) is inundated with emotional displays (a smile and a eyebrows lifted). It is thus emphatically delivered and affiliates with the stance that the mother took in the prior affiliating turn (line 2) through the sound objects ‘n: #å: #h n(.)hoi:’. The action is thus produced with reference to the preferred response (line 2) to the prior action (line 1). It is, as mentioned, designed as a tag question, which itself constitutes a close-ended question, which does not contribute to a development of the on-going topic, nor does it re-do or re-work the prior sequence initiating action (and then one makes Easter letter riddles). Instead, it works exclusively to re-offer the turn to the mother to, in principle, work on the preferred answer that she provided – to either confirm it or actually to repair it and hence to confirm or repair the daughter’s candidate understanding as suggested in line 1.

Of course, the question comes in the form of a positive tag (Heritage and Raymond, 2005), i.e. it is also designed to prefer an aligning or confirming response. Exactly this turns the action into a request for confirmation, which is then characterised by

re-offering the turn to PWCI to reconsider the preferred response while

pointing to a preferred (confirmative) response.

The mother responds with a preferred answer yes (°jo.h°) in line 4, which is combined with a smile and putting down the cellophane on the table. The latter may indicate an orientation to a termination of the sequence-expansion and hence of the sequence.

Requests for confirmation may, as mentioned, also be re-workings of the prior close-ended question. Example 2, excerpt 2.3, is an instance of this. It is an expansion of excerpt 2.2, in which the SLP1 asked the client (PWCI2) whether he reads the whole sentence. Recall that the client delivered the preferred yes:

Excerpt 2.3

9 SLP1 → kan du godt læse han /spiser et æble?

can you read ‘he eats an apple’

»text /»PWCI2

points with index finger at text card

10 PWCI2 ja

yes

nods

»text card

The request for confirmation (line 9) is a re-worked version of the prior close-ended question, do you read the whole sentence (excerpt 2.1 above), followed by the declarative question (the whole sentence, excerpt 2.2. above). Also, this request constitutes itself a close-ended question. The verb ‘læser’ in ‘læser du’ (do you read) is replaced with a modal verb can as well as the Danish modifier and confirmatory godt (good/well; you can do that alright). Furthermore, the item sentence in the prior first pair part is replaced with an example of a sentence he eats an apple. The sentence example is produced in a reading voice (Couper-Kuhlen and Selting, 1996) and what is read is a whole sentence with a subject, a verb and an object. Also, this close-ended question, which is produced in ‘a light tone of voice’, invites the client to reconsider the prior preferred yes while at the same time it is designed to prefer an aligning and positive response, a positive confirmation. The client provides this in his subsequent turn in line 10, yes.

A request for confirmation may also address the preferred response explicitly, which invokes the prior close-ended question that it depends on (Schegloff, 2007). In excerpt 1.2 it draws on prior turns in which the brother (R1) asked his older brother (PWCI1) if he wanted to be hi-her (line 1) in a game and in which he received a preferred yes (line 2):

Excerpt 1.2 (lines 1 and 2 replicated)

1. R1 → vil du være ha- ↑ hende?

do you want to be hi- her

2. PWCI1 j:a:

yes

»monitor

3 R1 → er du he::lt sikker?=

are you absolutely sure

»adult, smiles

4 PWCI1 ja

yes

»R1, »downwards

In line 3, R1 in effect re-offers the turn to his older brother to reconsider the preferred answer yes in line 2. Also, this way of requesting a confirmation is accomplished by employing the technique of a close-ended polar question, which as we have already seen, is designed to prefer an aligning answer, a confirmation. Also, in this case, it is designed (are you absolutely sure) to prefer a positive confirmation, which the older brother delivers in line 4, yes.

Second request for confirmation

As mentioned, preferred responses are more likely to close a sequence than are dispreferred ones. However, the phenomenon of interest is characterised by the fact that the sequence is also not being closed after PWCI has provided a preferred, confirming, response to a request to preferably confirm his or her prior answer. RUR emerges when yet another sequence expanding request for confirmation follows the provided confirmation to the first one. The second request for confirmation is re-workings, re-doings, repetitions or other re-instantiations of the first one. The former is produced with reference to the latter, which it expands. In line with the first pair parts of the two prior adjacency pairs, the second request for confirmation, and first pair part in a third adjacency pair, is designed as a close-ended question. Examples (2-6), excerpts 2.4, 3.3, 4, 5.1 and 6 are instances of this:

In example 2 below (excerpt 2.4) the SLP1 re-instantiates the first request for confirmation can you read he eats an apple (excerpt 2.3) in the second request for confirmation as she replaces ‘he eats an apple’ with the demonstrative pronoun ‘det’ (it) (line 11). Grammatically speaking, the Danish pronoun ‘det’ (it; neuter) does not refer to the ‘sætningen’ (the sentence; common gender) in the question-answer adjacency pair that was constructed prior to the first request for confirmation (examples 2.1 and 2.2):

Excerpt 2.4 (lines 9 and 10 replicated)

9 SLP1 → kan du godt læse han /spiser et æble?

can you read ‘he eats an apple’

»text  /»PWCI2

points with index finger at text card

10 PWCI2 ja

yes

nods

»text card

11 SLP1 → ka du /godt læse det=

you can read it?

/nods

»«PWCI2

12 PWCI2 =ja dub dub dub di /dub

yes ‘dub dub dub di /dub ’

leans back in the chair, produces vocalisations,

which are accompanied with left hand gestures,

points towards cards on the table,

points towards his mouth. Palm covers mouth

In example 3 (excerpt 3.3) below, the daughter (R2) re-instantiates the first request for confirmation (excerpt 3.2), which was delivered in the form of a tag question, right, upon the preferred answer, yeah to the statement question and then one makes Easter letter riddles (line 1, excerpt 3.1). The first request for confirmation received itself a preferred yes (line 4) combined with a smile and putting down the cellophane on the table. The re-worked request is constructed as a close-ended statement question (line 5), which is combined with nodding and gazing at PWCI3:

Excerpt 3.3. (lines 3 and 4 replicated)

3 R2 → ik os

right

smiles, eyebrows lifted

»«PWCI3

4 PWCI3 °jo.h°

yes

»«R2, »cellophane

smiles, puts cellophane down

5 R2 → jeg tænkte at det måske var det du ville sige

I thought that maybe that was what you wanted to say

nods, eyebrows lifted

»the cellophane, »PWCI3

6. PWCI3 °ja°

yes

lifts cellophane slightly from the table, »«R2, puts it down, hand rests on it, swifts gaze direction »table, facial muscles relax

In example 4 below, the daughter also produces a second request for confirmation (is it, line 7) after the mother (line 6) has provided a preferred answer (yes, produced through a nod and a smile) to the first one (yes? line 5). The first request for confirmation, yes?, was produced by reference to a confirming response (nods, line 4) to a prior close-ended statement question (thirty five years ago, lines 1-3), through which the daughter suggested that her mother danced for the first time to a specific tune thirty-five years ago. The two of them hummed that tune prior to the daughter’s question:

Example 4

1 R2 femogtredve,

thirty five

»«PWCI3

2 PWCI3 nods and smiles, »«R2

3 R2 år siden

years ago

»«PWCI3

4 PWCI3 nods shortly twice, smiles, »«R2

5 R2 → /ja?

yes

nods, lifts eyebrows, smiles, »«PWCI3

6 PWCI3 /jah

yes

/nods twice, smiles, »«R2

7 R2→ ↑er det det

is it?

smiles, »«PWCI3

8 PWCI3 jah

yes

nods, smiles, »«R2

Finally, in example 5 (excerpt 5.1) an SLP produces a second request for confirmation (line 5 did you?) of the client’s (PWCI3) preferred answer yes (line 2) to the SLP’s close-ended question of whether the client once played the flute (how was it did you play the flute once, line 1) and upon a preferred confirmation of the answer (line 4).

Example 5.1

1 SLP3 var der noget med at at var det dig har du spillet fløjte engang

was it so that that was it you have you played the flute once

frowns, »«PWCI3

2 PWCI3 ja

yes

nods

3 SLP3 → ja?

yes

»«PWCI3

4 PWCI3 ja

yes

nods

5 SLP3 → har du det?

did you

»«PWCI3

6 PWCI3 ja=

yes

All second requests for confirmation are sequence expanding first pair parts (Schegloff, 2007) that build on first requests with an inbuilt preference for a confirming response in the second pair part. All delivered responses to the second requests are also preferred ones. Thus, all PWCIs orient to the preference structures that are part of the actions in turns that simultaneously offer them an opportunity to reconsider previously delivered –preferred – answers as responses to an initial question and to a following first request for confirmation.

Close-ended questions and repeated use of request for confirmation: One set of methods, two possible sequential consequences

When RUR is followed by repair, accounts and eventually ‘giving up’

In some cases the PWCI finally makes use of the possibility to reconsider his or her prior response. The data indicates a possible explanation of when reconsiderations of prior responses are oriented to in subsequent turns and when this possibility does not have that kind of sequential consequence.

Example 6 below contributes to the development of this point. It is an interaction between a young woman diagnosed with aphasia and her SLP. Prior to the excerpt, the client (PWCI4) has repaired the Danish verb ‘bade’ (bath) in a sentence that was initiated with a personal pronoun ‘han’ (he) (‘han bader’ (he takes a bath) unsuccessfully. The transcript starts where her SLP makes a suggestion, as to which verb she is looking for, are you thinking of ‘to make’ (line 1). Upon the client’s delivery of the preferred answer in line 2, yes and a nod, the SLP produces a first request for confirmation in line 3 (was that the one?), which she reworks in the second request, line 5 (make), upon the client’s preferred confirmation of the first one (yes and a smile, line 4):

Example 6

1 SLP2 tænker du på at lave

are you thinking of ‘to make’

2 PWCI4 ja=

yes

nods, »«SLP2

3 SLP2 → =var det det

was that the one?

»«PWCI4

4 PWCI4 ja

yes

smiles

5 SLP2→ lave,

make

6 PWCI4→ ja (.) m:(.)ad

yes food

»«SLP2

7 SLP2 han laver mad,

he makes food

smiles, »PWCI4

As mentioned, the suggestion made by SLP2 (are you thinking of ‘to make’) emerges in the context of the client’s lack of success in repairing a word (bade). Hence, the suggestion in the form of a close-ended question points backwards in the interaction as it picks up on PWCI4’s work in constructing the prior action. In this way it concerns PWCI4’s performance. Upon the repeated (second) request for confirmation (RUR), the client then does more than simply provide a positive confirmation. In addition to confirming that she was actually thinking of the verb make, she adds food (in line 7). Through this grammatical and semantic expansion, she provides evidence that the verb make was the right one, which she of course already indicated in her answer (line 2, yes and nods) to the initial question, are you thinking of ‘to make’ (line 1), and confirmed (yes and smiles) subsequently (line 4).

Also, examples 2 and 3 are instantiations of how PWCI ends up reworking his/her answer to the initial question that somehow puts the performance of PWCI into play as it picks up on PWCI’s prior action: in example 2 the statement question (the whole sentence) addresses how the client (PWCI2) in a previous task matched texts with pictures; in example 3 it addresses the mother’s (PWCI3) previous action that she constructed by lifting up a piece of cellophane towards the daughter (R2). In both examples PWCI reacts to the co-participant’s continuous pursue of having his or her initial (close-ended) question confirmed. In example 2 (excerpt 2.4) we saw that the client added ‘dub dub dub di /dub’, which was co-coordinated with gestures:

Excerpt 2.4 (replicated)

11 SLP1 → ka du /godt læse det=

you can read that?

/nods

»«PWCI2

12 PWCI2 =ja dub dub dub di /dub

yes ‘dub dub dub di /dub’

leans back in the chair, produces vocalisations,

which are accompanied with left hand gestures,

points towards cards on the table,

points towards his mouth. Palm covers mouth.

In the sequential context in which he has twice confirmed his confirmation of the SLP’s initial question (do you read the whole sentence) the addition of talk and gesturing indicates a kind of account, expansion or explanation of his confirming response to the second request (line 12). He seems to indicate that in the micro-context of having received two preferred responses to actions that projected them while at the same time giving him the opportunity to reconsider them, his co-participant must be after something more (or something else?) than the provided asked-for preferred answers. He delivers that extra something and actually reworks or re-does this in a confirmation to a third request for confirmation (the only example of a third found in the data):

Excerpt 2.5

13 SLP1 det ka du seh

you can see that

»«PWCI2

14 PWCI2 → ja /de dub de dub dub

yes ‘de dub de dub dub’

»«SLP1

/points towards mouth

In line 14, the client constructs his response in similar ways as his response to the second request for confirmation (excerpt 2.4. line 12, above). Hence, he not only confirms the preferred prior confirmation, he also confirms what seems to be his account of it.

In example 3 (excerpt 3.3), the mother (PWCI3) too adds to her confirming response to the second request for confirmation (line 6):

Excerpt 3.3 (replicated)

5 R2 jeg tænkte at det måske var det du ville sige

thought that maybe that was what you wanted to say

nods

»the cellophane, »PWCI3

6 PWCI3→ °ja°

yes

→lifts cellophane slightly from the table, »«R2, puts it down, hand rests on it,

swifts gaze direction »table, facial muscles relax

In line 6, the mother indicates an orientation to the possibility of repair. In responding to the second request for confirmation, she provides a yes and then lifts up the cellophane again slightly, as she achieves mutual gaze with the daughter. She then changes trajectory of the movement as she puts the cellophane back on the table, rests her hand on it, and changes gaze direction from the daughter to the table. Notice, that she also relaxes her facial muscles (from a smile to a non-smile). The fact that she changes a reworking of her prior movement to a non-reworking one may be understood as if she indicates giving-up and the fact that she changes a smiling face to a non-smiling one may be understood as if she indicates sadness. As a matter of fact the daughter treats her mother’s action along these lines in her subsequent turn (excerpt 3.4 below). She relaxes her facial muscles too, which is remarkable, since she has previously lifted her eyebrows and smiled (3.2. and 3.3 above). In other words, she too changes from behaviour that can be understood as indicating happiness to behaviour that may indicate sadness as she displays her understanding of her mother’s change in the production of her response to the second request for confirmation: she, the daughter, did not understand the mother’s prior action (as suggested in line 1, excerpt 3.1):

Excerpt 3.4

7 R2 → det var det /ik

it wasn’t

muscles relax

»PWCI3 /»«PWCI3

An alternative understanding of the mother’s behaviour is that she gives up on finding another way of telling her daughter that the initial action (lifting up the cellophane) actually meant to convey what the daughter suggested (and then one makes Easter letter riddles) – for all practical purposes. This was what for example the client (PWCI4) in example 6 worked at by adding the lexical unit food to the verb make (line 6), as did the client (PWCI2) in excerpt 2.4 (line 12) as he added ‘dub dub dub di /dub’ and gestures to his confirmatory yes (see above).

In example 6, the client (PWCI4) and her SLP (SLP2) succeed in establishing what the client attempted at saying (makes). In that context, they continue the activity in terms of talking about the kind of food that ‘he makes’ (not in the transcript). However, in examples 2 and 3, PWCI’s accounting or reworking of actions, which occasioned the co-participant’s first and sequence initiating question, does not contribute to a more profound understanding of that action, i.e. an understanding of what PWCI really meant. Neither did of course the positive responses to the repeated requests for confirmation of their (preferred confirming) answers to questions that addressed it. In that context, the attempt at really understanding PWCI’s action (prior to the close-ended question and the RUR sequence) is given up: In example 3 (excerpt 3.5 below) the mother (PWCI3) lifts up the cellophane once again and shakes it while she nods and produces (unintelligible) talk (line 8). Her daughter (R2) treats this as an invitation to throw it into the wastebasket (line 9):

Excerpt 3.5

8 PWCI3 føj eh forøg

‘føj eh forøg’

nods, lifts cellophane, shakes it

9 R2 → ska jeg smide det ud?

want me to throw it out

»«PWCI3, nods

10 PWCI3  ja

yes

smiles, nods

In example 2 (excerpt 2.6 below), the SLP initiates the termination of the sequence as she – after a third request for confirmation – in an emergent fourth request, changes trajectory mid-turn and produces a very positive assessment of the client’s achievements in reading (line 19):

Excerpt 2.6

15 SLP1 ka du se (.) altså- læser du:

can you see you know do you read

16 PWCI2 .hhh

17 (0.5)

18 SLP1 ka- ka- du-

can you

»text card

points with small finger towards text card

19 SLP1 → altså de- de- /jeg synes bare det er så fantastisk=

you know I just think that it is so fantastic

»cards  /»«PWCI2

20 SLP1 =fordi du ku slet ik ↑læse nogen sætninger i starten

because you couldn’t read any sentences in the beginning

The production of what seems to be a fourth request for confirmation is abandoned by the SLP1 (line 18). Notice, that the client’s only response to the unfinished fourth attempt is a deep in-and outbreath ‘.hhh’ (line 16), which may be understood as indicating frustration or giving-up (as did the mother, PWCI3, excerpt 3.3) in an attempt to tell her what he has already told her a couple of times in different ways and accounted for. In this context, the SLP1 produces a positive assessment ‘you know I just think that it is so fantastic’ (line 19) of his reading skills. Thus, through this assessment she changes the trajectory of the interaction as the assessment does not deal with his possible account of how he can read the whole sentence. In the context of client’s repeated responses, the SLP1’s positive assessment comes off as a redefinition of what she has been pursuing by the repeated re-workings of prior requests and questions or at least a redefinition of what the client takes her to having done as indicated in his final responses and accounts.

When RUR is followed by topic development

The use of RUR upon preferred responses to close-ended questions does not always result in repairs, accounts or re-workings (in some cases before continuing the main business of the interaction). Instead, the sequence may be terminated and further talk developed. In these cases the close-ended questions prior to the first request for confirmation with preference for specific responses do a specific job that differs from the one described above. Example 5 (excerpt 5.1) represents such examples:

Excerpt 5.1 (replicated)

1 SLP3 var der noget med at at var det dig har du spillet fløjte engang

was it so that that was it you have you played the flute once

frowns, »«PWCI3

2 PWCI3 ja

yes

nods

By means of her close-ended question (line 1), the SLP3 initiates a topic (Button and Casey, 1984). The question prefers a yes, which PWCI3 delivers in line 2. A request for confirmation of this response offers as any other request for confirmation of course the turn to PWCI3 to reconsider her yes. The preferred answer is, however, not a confirming response to a question of what PWCI3 did in a previous turn. Hence, a repeated use of request for confirmation does not provide her the possibility to potentially reconsider how she performed previously in the interaction. Instead, it works to prompt or proffer the initiated topic (Schegloff, 2007) and in this sense to point forward in the interactional business. In responding to this proffer, PWCI3 may (as in excerpt 5.1) align with the topic by employing a positive response token, yes:

Excerpt 5.1 (replicated)

3 SLP3 → ja?

yes?

»«PWCI3

4 PWCI3 ja

yes

nods

5 SLP3 → har du det?

did you?

»«PWCI3

6 PWCI3 ja=

yes

PWCI3’s aligning response encourages the topic and as the subsequent sequence expanding requests continue prompting talk on the topic (yes? line 3; did you? line 5), so do the responding 4aligning confirmations by PWCI3. In these cases it may of course become the job of the non-impaired co-participant to produce actions that contribute to the development of it:

Excerpt 5.2

7 SLP3 → =sådan en der

one like this

points at a picture of a flute in a newspaper, »picture

8 PWCI3 ja

yes

taps on the picture with index finger nods

9 SLP3 okay

okay

In line 7, the SLP3 builds on the syntactic structure of the initial close-ended question, did you play the flute once as she constructs a (further) request for confirmation. Notice, that this request does not expand the prior (second) request for confirmation sequence (did you? line 5). Instead, it develops the topic of whether the client played the flute once into a matter of which kind of flute she played. This choice amongst alternative flutes is accomplished through the use of stress ‘sådan en der’ (one like this).

In sum, in cases in which the initial close-ended question initiates a new topic the use of RUR seems to establish the relevancy and the newsworthiness of the topic, which is then developed subsequently. Hence, in this sense the initial close-ended question points forward to further interactional business without addressing any potential issues in understanding a performed action, or issues in performing actions, on the co-participants’ part. RUR works to support this.

Discussion: RUR upon preferred responses to close-ended questions and local possibilities for participation in interaction

Participation in interactional activities and thus involvement in social life is accomplished locally through concrete actions for interaction. In interactions between PWCIs and their co-participants, close-ended questions and subsequent repeated use of request for confirmation-confirmation adjacency pairs may be seen as methods for achieving this. The local details of the interactional context are, however, crucial to the ways in which participation and involvement are achieved. When close-ended questions are used to initiate topic, RUR is recognisable as a method for ‘insisting on’ inviting the PWCI to engage in the talk on the topic by prompting or proffering it. The possibility to reconsider the provided preferred answer is thus a possibility to reconsider how further talk may contribute to developing the topic.

When close-ended questions are used to address the understanding of some prior action produced by the PWCI, RUR is recognisable as a method for pursuing to nail down what he or she ‘really’ meant. The possibility to reconsider the provided answer is thus a possibility to reconsider how further talk may help to establish the sense of the prior action. In this way, the method in this micro-context addresses the performance and thus the competence of the PWCI.

The RUR sequences develop in similar ways independently of the micro-context of the close-ended question, in that they expand the original adjacency pair after the delivery of its second pair part and in that the second pair parts of the requests in the RUR sequences are preferred responses (confirmations). However, the interactional implications and consequences differ, in that RUR when employed for the purpose of ‘really’ understanding makes it relevant for the PWCI to produce her or his action in a different way or to demonstrate (including explanations of and accounts for) what is meant by it. The close-ended question (plus RUR) points backwards in interaction, as it addresses a prior action that belongs to the PWCI. In interactions in which the PWCI’s means for constructing actions are restricted, the employment of RUR risks putting both participants checkmate, the PWCI may not be able to produce (or demonstrate) the action in different ways through the means that are available. This leaves the co-participant, whom the turn is re-allocated to, little possibilities for contributing to a further development of the interaction without by-passing the efforts of the PWCI or ignoring the interactional work that they have both been engaged in in doing RUR. This work was an attempt at establishing what the PWCI ‘really’ meant to say.

Pilesjö and Rasmussen (2013) found that the use of communication boards combined with talk makes it possible to employ techniques through which the co-participants co-construct turns and action that belong to the minimally speaking co-participant.

When a close-ended question (and RUR) is employed to initiate a topic, it points forward in the interaction. Not just as it projects a next action (a second pair part) but also as it provides a topic for possible further talk. The request for confirmation adjacency pairs then addresses an action (the close-ended question) that belongs to the non-impaired co-participant. When the turn is (repeatedly) re-allocated to her or him upon for instance minimal confirmations (yes), she or he is then offered the opportunity by PWCI to develop the topic.

One may hold that the use of RUR in the two analysed environments provides possibilities for participation in interaction and thus involvement in social life in different ways and with different qualities. RUR-to-‘really’-understand seems to be used to help the PWCI talk for her- or himself and thus to be involved in social life on equal terms. By contrast the use of RUR-to-proffer-topics opens up – easily – for the non-impaired participant to, more or less, speak on behalf of the PWCI and thus to categorise social life for her or him. The former is without doubt the ideal of social life, but the use of RUR for that purpose is risky. The latter may be less attractive in terms of empowerment, but the use of RUR for proffering topics at least gives the PWCI the possibility to participate at all and the (both) co-participants to interact locally and in the here and now of the interaction without risking to have to deal with the delicacy of lack of understanding, which is potentially due to insufficiently available ordinary resources for interaction.

Clinical implications

Close-ended questions and subsequent requests for confirmation should be used, trained, and recommended with a view to the micro-context in which they occur for them to work to achieve specific purposes.

Summary and conclusion

Close-ended questions and repeated use of request for confirmation are in accordance with recommendations of techniques for securing intersubjective understanding in atypical interaction. When answers to the close-ended question are preferred the use of RUR develops sequentially in the same way independently of the local context of the close-ended question and thus of the job that is accomplished by posing it: a request for confirmation (itself a close-ended question) gets the preferred confirmation – repeatedly so. Furthermore, they may make further talk/action relevant as the RUR sequence develops. The difference in the micro-context of the close-ended question, however, has a consequence for how the close-ended question and RUR are understood and hence on the subsequent talk: When used by the non-impaired co-participant for topic initiation, the employment of RUR makes a confirmation of and further talk on the topic relevant. The non-impaired participant her- or himself may provide the further talk. When used by non-impaired co-participant for understanding a prior action produced by the PWCI, RUR makes a confirmation of and a demonstration, account or reworked version of the prior action relevant – by the latter. This moves the performance and competence of the PWCI into focus, risks the effortless ease with which the interaction may be conducted, while being understandable as an attempt to treat the PWCI as an otherwise competent member of society. Close-ended questions and RUR may thus be used as methods for creating possibilities for participation in interactional activities and for involvement in social life. As this study has shown, the micro-context of the close-ended question is though decisive for how the PWCI may participate and be involved in the interaction and hence in social life.