INT J LANG COMMUN DISORD, XXX VOL.

2014,

00, NO. 0, 1–14

Research Report Communicative strategies used by spouses of individuals with communication disorders related to stroke-induced aphasia and Parkinson’s disease Emilia Carlsson†, Lena Hartelius†‡ and Charlotta Saldert†‡ †Institute of Neuroscience and Physiology, Division of Speech–Language Pathology, Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden ‡University of Gothenburg Centre for Person-Centred Care (GPCC), Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden

(Received September 2013; accepted April 2014) Abstract Background: A communicative disability interferes with the affected person’s ability to take active part in social interaction, but non-disabled communication partners may use different strategies to support communication. However, it is not known whether similar strategies can be used to compensate for different types of communicative disabilities, nor what factors contribute to the development of a particular approach by communication partners. Aims: To develop a set of categories to describe the strategies used by communication partners of adults who have problems expressing themselves due to neurogenic communicative disabilities. The reliability of assessment was a particular focus. Methods & Procedures: The material explored consisted of 21 video-recorded everyday conversations involving seven couples where one spouse had a communicative disability. Three of the dyads included a person with dysarthria and anomia related to later stages of Parkinson’s disease, while four of them included a person with stroke-induced aphasia involving anomia. First a qualitative interaction analysis was performed to explore the strategies used by the communication partners when their spouses had problems expressing themselves. The strategies were then categorized, the reliability of the categorizations was explored and the relative frequency of the various strategies was examined. Outcomes & Results: The analysis of the conversational interactions resulted in a set of nine different strategies used by the communication partners without a communicative disability. Each of these categories belonged to one of three overall themes: No participation in repair; Request for clarification or modification; and Providing candidate solutions. The reliability of the categorization was satisfactory. There were no statistically significant differences between diagnoses in the frequency of use of strategies, but the spouses of the persons with Parkinson’s disease tended to use open-class initiations of repair more often than the spouses of the persons with aphasia. Conclusions & Implications: The types of strategies used by spouses of persons with neurogenic communicative disabilities seem to be more strongly associated with individual characteristics of communicative ability than with the type of disorder involved. The set of categories developed in this study needs to be trialled on larger groups of participants, and modified if and as necessary, before it can be regarded as a valid system for the description of such strategies in general. Once this has been done it may become a useful instrument in the assessment of the strategies used by communication partners of individuals with communicative disabilities. Keywords: conversational interaction, communication partners, repair, supporting communication, aphasia, Parkinson’s disease.

What this paper adds? Previous research has shown that communication partners of persons with communicative disabilities are able to adapt their use of different types of communicative strategies to support communication. However, the relative frequency of use of such strategies in relation to different types of communicative disabilities has not been fully explored. This study has yielded a reliable set of categories whose application shows that communication partners of Address correspondence to: Charlotta Saldert, Institute of Neuroscience and Physiology, Division of Speech–Language Pathology, Sahlgrenska Academy at the University of Gothenburg, Box 452, SE-405 30 Gothenburg, Sweden; e-mail: [email protected] International Journal of Language & Communication Disorders C 2014 Royal College of Speech and Language Therapists ISSN 1368-2822 print/ISSN 1460-6984 online  DOI: 10.1111/1460-6984.12106

2

Emilia Carlsson et al. persons with problems expressing themselves due to Parkinson’s disease with anomia or to stroke-induced aphasia may use the same types of strategies. In fact, the choice of strategies seems to be more strongly related to individual characteristics than to the type of disease. The results also indicate that the set of categories may be developed into a useful instrument not only in the clinical assessment of communication partners’ ability to support communication, but also in counselling and training of such partners.

Introduction A communication disorder inevitably affects a person’s ability to interact socially with significant others and with society in general. Two common causes of such disorders are stroke and Parkinson’s disease (PD). Common sequelae of stroke include aphasia and dysarthria. Regarding PD, the communicative disorder most strongly associated with it is dysarthria, but linguistic and other cognitive difficulties become apparent in later stages of the disease and word-finding difficulties are common both in aphasia and in PD (Berg et al. 2003, Goodglass and Wingfield 1997, Hartelius and Svensson 1994, Miller et al. 2006). This has consequences for the achievement of mutual understanding in conversations between these individuals and their conversation partners. There are different ways for the participants in a conversation to manage such difficulties. Below, some such ways will be discussed focusing on the two main areas: how repair is managed in conversations and current knowledge about cooperation in conversation in relation to stroke-induced aphasia and PD. This is followed by a brief discussion of the need for reliable instruments exploring communication partners’ use of strategies in conversations with people with communication disorders. Repair in conversation Interactions with people who have communicative disabilities are often characterized by intermittent repair work due to frequently occurring problems with mutual understanding. The concept of repair has been investigated with conversation analysis (CA) (Sidnell 2010). In CA what is displayed by participants, for example, in terms of mutual understanding, is studied in the sequential organization of contributions in natural conversations. According to Schegloff et al. (1977), the structure of repair sequences is organized in a system for the recognition and management of issues that need repair. The repair is preceded by a trouble source that someone perceives as a problem. When this problem is highlighted by someone, a repair is initiated. An initiation of repair is usually followed by an attempt to solve the problem. Repairs may be initiated and performed either by the person whose contribution contained the problem (‘self-initiated self-repair’) or by the conversation partner (‘other-initiated other-repair’). Moreover,

repair initiated by one participant may be completed by the other (‘other-initiated self-repair’ or ‘self-initiated other-repair’). One relevant aspect in this context is the identification of the trouble source, given that knowing exactly where the problem is makes it easier to repair. Based on interactions involving people without communicative disabilities, Schegloff et al. (1977) have described different types of other-initiations of repair which vary in their ability to identify the specific trouble source. First, initiators such as ‘huh?’ and ‘what?’ do not really indicate where in the preceding utterance the trouble source was located. They have been called ‘open-class’ initiators (Drew 1997). ‘Closed-class’ initiators, for example question words such as ‘who?’, ‘when?’ and ‘where?’, indicate the type of referent and may thus be more effective. A ‘partial repeat’ of the conversational turn including the trouble source, supplemented with a question word in the appropriate place: ‘you went where?’, ‘the same what?’, is even more specific about the location of the trouble source. Finally, according to Schegloff et al. (1977), seeking ‘verification’ of a possible understanding of the preceding turn: ‘You mean you went to the market?’ is the strongest repair initiator and a very powerful way to pinpoint the trouble source. In fact, this type of repair initiator, where the other speaker only needs to confirm or deny with a ‘yes’ or a ‘no’, is often used in cases where people have difficulty expressing themselves due to aphasia (Simmons-Mackie and Kagan 1999). Cooperation in conversation In conversational interaction generally, self-initiated self-repair is more frequent than other-initiated otherrepair (Schegloff et al. 1977). However, as pointed out by Milroy and Perkins (1992), people with communication disorders often have difficulty performing effective self-repair, meaning that active participation in repair by their communication partners may have a beneficial effect on communication (Simmons-Mackie et al. 2010). In fact, it is an essential component of established theories of pragmatics and social interaction (Grice 1975, Schegloff 1982) that the quality of a conversation always is the result of cooperation between the communication partners involved. The effects that a communicative disability may have on the interaction and the ways in which conversation partners cooperate to express and convey meaning have been described, by means of

Communicative strategies used by spouses CA, in relation to both aphasia (e.g. Goodwin 1995, Oelshlaeger and Damico 2000, Wilkinson et al. 2003) and dysarthria caused by motor neuron disease (Bloch 2005, Bloch and Wilkinson 2009, 2011). Cooperation in word search related to aphasia has often been dealt with using CA. Lubinski et al. (1980) described collaborative repair work in a conversation between a man with aphasia and his wife as a ‘hint and guess’ cycle. The wife used semantic and syntactic clues contained in the contributions from her aphasic husband to offer guesses about what the target information might be. A collaborative hint-and-guess strategy was also demonstrated by Laakso and Klippi (1999), who analysed 24 conversations involving three aphasic participants. Ferguson (1993) also reported the existence of the practice of supplying words in seven dyads each including a person with aphasia, and Oelshlaeger (1999) described how the wife of a person with aphasia, when guessing target words, relied not only on linguistic features of her husband’s speech but also on their shared experience. Simmons-Mackie and Kagan (1999) described communicative strategies used by persons who had been rated as ‘good’ or ‘poor’ communication partners of persons with aphasia. The ‘good’ communication partners seemed more eager to help the persons with aphasia to avoid exposing their linguistic problems and focused more on the social aspects of the interaction than on the goal of conveying information. The strategies used to support the social aspects of communication included the use of response tokens to signal that the conversation partner was interested and willing to listen as well as adaptation to the means of communication used by the person with aphasia. Both the ‘good’ and the ‘poor’ communication partners used clarification questions, such as repetition (with rising terminal intonation) of the message as perceived by them, and guesses, such as the presentation of alternative versions of what the intended message may have been. Several different strategies used in relation to word search by conversation partners of people with aphasia have thus already been described. An attempt to categorize conversation partner strategies used in aphasia was made by Oelshlaeger and Damico (2000). They used CA and described different strategies used in participatory word search in eight video-recorded conversations between a man with aphasia and his wife: a ‘guess’ strategy where the communication partner provides a word; ‘alternative guess’, where a new guess is offered when the first one is rejected; ‘completion’, where the information provided is expressed more as an assertion and less as a question; and ‘closing’, where the conversation partner’s contribution is intended to move the conversation forward without the issue of the trouble source being resolved. Attempts to systematize and categorize strategies and other resources used by the communication partners of

3 persons with dementia has also been made (Orange et al. 1996). When it comes to cooperation in conversation in relation to PD, there are only a few studies of conversational interaction including persons with dysarthria due to PD. Those studies have reported problems with turn-taking, for example, due to delays or failures to respond or due to word searching and reliance on minimal acknowledgements, as well as difficulties in topic initiation and topic maintenance, referencing and repair (Griffiths et al. 2011, Hartelius et al. 2011). Griffiths et al. (2012) used CA to explore everyday communication in relation to PD, reporting on a tendency for overlap due to dysarthria, which led to repair. They also described how opportunities to repair the disordered talk were often not followed up, which may lead to a deletion of PD turns at talk, thus reducing participation in conversation. Griffiths (2013), using CA on natural conversations involving persons with PD and dysarthria, described a pattern of other-initiation of repair in line with the typology of Schegloff et al. (1977). The qualitative aspects of word search in relation to PD have not yet been explored to the same extent as in aphasia although pragmatically atypical word choices used in conversational interactions have been described as affecting mutual understanding (Saldert et al. 2014). The analysis of conversational interaction alone does not allow inferences to be made about the causes of dysfluencies and delayed responses. Although word-finding difficulties are common in PD at later stages, dysfluencies and gaps in interaction may also be related to memory loss and attention deficits as well as the speechinitiation difficulties associated with PD. In speech and language pathology, a distinction is traditionally made between communication disorders affecting language abilities, such as aphasia related to stroke or traumatic brain injury, and speech disorders (or dysarthria) due to, for example, stroke or progressive neurological diseases such as PD. This is a relevant distinction in many ways. The conversation partner of someone with a communicative disability needs to adjust to the individual resources and problems of that person, and they may vary depending on the type of disorder. However, the strategies available for the management of problems in communication are finite, irrespective of whether persons with communicative disabilities are involved. Further, people suffering from stroke or PD may be affected by both dysarthria and anomia, and that this may affect the choice and use of strategies by their conversation partners even though it is not always possible to identify the origin of specific problems in terms of mutual understanding. Bloch and Beeke (2008) used CA as a means of describing and comparing conversations in two dyads where one member of one dyad had dysarthria due to a motor neuron disease and one member of

4

Emilia Carlsson et al.

the other had aphasia. There were several similarities in the practices used, and Bloch and Beeke suggested that further non-disorder-specific research should be carried out. Furthermore, they propose that coconstruction is one way in which communicative competence is accomplished in relation to communication disorders. Milroy and Perkins (1992) showed that less overall collaborative effort is required if the partner with no communicative disability cooperates in repair work with the person who has a communicative disability. However, prolonged repair sequences may expose and highlight a person’s linguistic incompetence, thereby exposing him or her to face-threat, and avoidance of repair at the expense of understanding is not uncommon (Bloch and Beeke 2008, Goffman 1955, Wilkinson et al. 2003). Intervention Training of communication partners has been shown to have the potential to facilitate communication for persons with a communicative disability due to aphasia, and there is a growing interest in establishing training programmes that may improve communication for persons with dysarthria (Griffiths et al. 2011, Forsgren et al. 2013, Simmons-Mackie et al. 2010). However, in order to be able to provide effective interventions, more knowledge is needed about strategies to support communication in relation to communicative disabilities (Oelshlaeger and Damico 2000). There is also a need for instruments to measure behavioural changes in the communication partners without disabilities. Global ratings scales like the Measure of skill in supported conversation (Kagan et al. 2004), may provide a comprehensive measure of the conversation partner’s ability to adapt, but it needs to be complemented with measures of what more specifically it is that has changed. Furthermore, to be able to use global rating scales in a valid and reliable way, further knowledge about patterns in conversational interaction is warranted. The aim of the present study was to develop a reliable set of categories to describe the strategies used by communication partners without disability (CP) of persons who have problems expressing themselves due to different types and degrees of communicative disabilities (PwCD) related to either PD or stroke-induced aphasia. The specific research questions asked were: (1) What strategies used by CPs can be observed in cases of communicative problems during conversations with PwCDs, and can those strategies be reliably categorized? (2) What is the relative frequency of use of different strategies (according to the categories yielded by the present study) by CPs in cases of communicative problems during conversation? (3) Are there any differences

in the use of strategies that may be associated with the type of diagnosis of the PwCD? Method Participants The study was based on video-recorded conversational interactions within seven dyads, each consisting of one PwCD and his or her spouse (CP) (table 1). Three of the dyads (PD dyads) included a person with a communicative disability related to PD in later stages, which had resulted in dysarthria as well as anomia. Four of the dyads (SA dyads) included a person diagnosed with aphasia, including anomia, related to a stroke. None of the participants had a dementia diagnosis. All participants reported having Swedish as their primary language as well as hearing and vision within functional limits (aided or unaided). The CPs had no known neurological disease or brain damage. The study was approved by the regional research-ethics board at the University of Gothenburg. Written informed consent was obtained from all participants. PD dyad 1 consisted of Robert, a former medical doctor, and Sonja, a former audiologist. Robert’s ability to articulate fluctuated and he often spoke in a weak voice. He had intermittent anomia and often tried to use other words or circumlocutions when he was unable to find the words he needed. PD dyad 2 consisted of Sten, a retired economist, and his wife Ingrid, who had worked as an office assistant before retirement. Sten’s speech was fast, often with imprecise articulation. In cases of problems expressing himself, Sten often seemed to give up the planned utterance completely or finished his contribution in a weak voice with imprecise articulation. PD dyad 3 consisted of Carl, a retired journalist, and Mary, who had worked as a nurse. Carl often spoke in a weak voice. He had intermittent anomia, and often used semantic paraphasias. SA dyad 4 consisted of Ted, a retired dentist, and Lisa, a teacher. Ted had moderate-to-severe Broca-type aphasia. His speech was non-fluent and characterized by anomia and frequent phonological paraphasias. He was sometimes able to produce short phrases, although his articulation was imprecise. Ted also often pointed and sometimes gestured with his left hand. SA dyad 5 consisted of Sara, who had been a shop assistant before the onset of aphasia due to a stroke, and Hans, who was an electrician. Sara had severe mixed aphasia. In addition to ‘yes’ and ‘no’ responses and variations in prosody and facial expressions, she used a communication book with pictures and words and she often tried to make drawings to provide clues to what she wanted to say.

Communicative strategies used by spouses

5 Table 1. Participant characteristics

Dyad

PD 1

PD 2

PD 3

SA 4

SA 5

SA 6

SA 7

Age (PwCD/CP) Gender (PwCD/CP)a Years as a couple Years since diagnosis Stage of PD/ degree of aphasiab Contextual speech intelligibilityc Word fluency (Phon/Sem)d Token teste

76/73 M/F 49 13 IV 75% 28/9 175

79/73 M/F 49 18 IV 37% 26/19 249

72/72 M/F 49 19 III 87% 31/26 237

73/54 M/F 8 4 Moderate–severe – 4/0 155

63/61 F/M 37 1 Severe – 0/0 108

45/47 F/M 4.5 2 Mild–moderate – 7/16

71/69 M/F 25 0.4 Severe – 2/2 14

f

Notes: a M = Male, F = Female b According to the Hoehn and Yahr (1967) five-grade scale for PD and according to a diagnostic aphasia test (A-ning; Lindstr¨om and Werner 1995) for aphasia. c Intelligibility was measured (for PwCDs with PD only) as the percentage of correctly perceived words by a naive rater of 100 words uttered by the PwCD, in context, in a video-recorded conversation. d Phonological/semantic word fluency was measured as the number of words beginning in F, A and S, the number of animals, and the number of activities mentioned during 1 min per category. e De Renzi and Vignolo (1962) token test; maximum of 261. f The PwCD in dyad 6 discontinued the test session, but her comprehension seemed adequate in conversation.

SA dyad 6 consisted of Tina, a pre-school teacher, and Paul, a salesman. Tina had suffered a stroke and had mild-to-moderate mixed aphasia with intermittent anomia but good comprehension. Tina was often able to express herself effectively and had developed a strategy for self-prompting of verbal production in which she wrote the first letters of the words she wanted to say in the air or on a table. She also frequently used gestures and made drawings in the air or on the table. SA dyad 7 consisted of Thomas, who had worked as a lawyer before retirement, and Maria, who had worked as a manager of a dental clinic. Thomas had been diagnosed with Wernicke’s aphasia, and presented with fluent speech characterized by frequent semantic and phonological paraphasias and neologisms as well as severe comprehension problems. He usually made no attempts to initiate repair of his own utterances, and his paraphasic speech was very difficult to understand. Procedure The material consisted of nine video-recorded conversations involving the PD dyads and twelve involving the SA dyads. All couples were future participants in a training programme for communication partners, and they were all video-recorded at home on three different occasions. The couples were instructed to talk and act as they usually did during 15 min of recording. They were told that while they were free to choose any topic they wanted to, they should preferably discuss something they needed to talk about anyway. To this was added that they could stay quiet together as much as they wanted to. The video-recording equipment was set up by a research assistant, who then left the couples alone during the recording and returned after 15 min.

Data analysis In the first part of the study, the aim of which was to identify categories of strategies, the middle 10 min of the video-recorded interactions were analysed qualitatively. The middle section was chosen because it was felt that the couples would be less self-conscious about being video-recorded after a few minutes, and hence that section would be more representative of their typical interaction. The methods of analysis used are based on an inductive analytic approach that has been described as ‘interpretive description’ (Thorne et al. 2004). This is a qualitative method that constitutes an alternative to grounded theory, phenomenology and ethnography when it comes to generating grounded knowledge suitable for applied health and clinical problems. The aim of using the inductive analytic approaches characteristic of interpretive description is to understand and illuminate the characteristics and patterns of a phenomenon in a theoretically—and in this case also practically—useful manner. The analysis of the conversational interaction was influenced by the principles and findings of CA (Sidnell 2010), but it should be noted that the focus of this study is what happens when one participant’s ability to express him- or herself is impaired—including instances where the conversation partners do not actually treat the signs of impairment as a problem. That is, the instances analysed in this study are those deemed by the researchers to manifest symptoms of impairment (see the exact definition below). This means that the method used for the analysis of the conversations does not take a completely conversation-analytic approach. Reflecting this, the instances of symptoms of impaired ability to express oneself are called ‘signs of problems’ in this study instead of ‘trouble sources’, which is an established term but would imply that the conversation partners always treat the turn concerned as a problem.

6 In the initial analysis, all occurrences of signs of problems in the PwCDs’ ability to express themselves were coded. Occurrences of signs of problems related to word search or other cognitive impairments as well as speech problems were included. An occurrence of a sign of problems was defined as follows: It had to occur either in the PwCD’s contribution or when the PwCD had explicitly been chosen as the next speaker. Signs of possible problems included pauses and gaps in the flow of the conversation; hesitation sounds; cut-off words; circumlocutions; paraphasias; or imprecise articulation. When co-occurring with other signs, use of gestures, gaze or other body communication and soft voice was also interpreted as a sign of problems. The final type of sign involved the PwCD making the problem explicit by means of a verbalized question (e.g. ‘what’s it called?’) or inviting the CP to take part in a word search in some other way. Examples of signs of problems are presented in the ‘Results’ section below in the form of transcribed extracts from conversations. For each occurrence of a turn including a sign of problems, the action by the CP that either overlapped the problematic turn or followed in the next turn was analysed and described. When the CP used several different strategies in relation to the turn including the sign of problems, each of those actions was treated as a separate strategy. That is, each occurrence of a sign of problems may result in the application of several different strategies. When an action performed by the CP was followed by further attempts at self-repair from the PwCD, that PwCD contribution was coded as a new occurrence of a sign of problems. To identify categories into which the strategies used by the CPs could be divided, a three-step procedure was used. The first step was based on transcriptions of two of the three video-recordings of one PD dyad (1) and one SA dyad (4). All contributions by the respective CP following a sign of problems were discussed by the authors. Categories were defined and redefined until each case could be assigned to one (and only one) category. In the next step, two other PD dyads (2–3) and three other SA dyads (5–7) were subjected to the same procedure: all occurrences of signs of problems were identified and then discussed by the authors. The categories from the first step were adapted and redefined to fit the different strategies used by the CPs in all seven dyads. In the third and final step, the inter-rater reliability of the use of the set of categories was analysed using Cohen’s kappa, as well as per cent agreement to obtain separate measures of the inter-rater reliability of the individual categories. The first and third authors independently categorized, according to the final set of categories and using written definitions, the signs of problems occurring in 33% of all the video-recorded material: data from one video-recording of each dyad

Emilia Carlsson et al. which had not previously been coded using the set of categories. Cohen’s kappa was also used to obtain a measure of intra-rater reliability, based on ratings by the first author of 19% of the material on two occasions one and a half weeks apart. When the reliability of the set of categories had been examined (and found to be satisfactory), the second part of the study was initiated. This involved recategorization by the first author, using the final set of categories, of all actions by the spouses without a communicative disability following signs of problems in the material. The relative frequency of the different types of strategies used by the CPs was also analysed in relation to whether the communicative disability of the respective PwCD was related to stroke-induced aphasia or PD. The frequencies of the various strategies were calculated as percentages of all occurrences of strategy use in all dyads, in PD dyads and in SA dyads separately. Then the relative frequencies were compared as between the different diagnoses using the non-parametric Mann–Whitney Utest, with the alpha level set to 0.005 in accordance with a Bonferroni correction due to the performance of multiple comparisons. Results This section begins with a presentation of the categories that emerged from the qualitative analysis of the strategies used in the material and the calculations relating to the reliability of the ratings. This is followed by an account of how often the CPs used the various strategies and how these were distributed among the different dyads. Finally, the distribution of the strategies between the two types of diagnosis is compared. The categories The final set included nine categories (table 2). Each of these was related to one of three themes: No participation in repair; Request for clarification or modification; and Providing candidate solutions. The categories are presented below, all illustrated with examples taken from transcribed conversations. The extracts have been translated into English, but the Swedish originals can be obtained from the corresponding author. Numbers in single parentheses in the extracts indicate the length (in tenths of a second) of a pause while non-verbal activities are indicated within double parentheses. Three of the categories (1–3) covered contributions from the CP that were not related to the initiation of or participation in repair. In the first of these (category 1: response tokens), the CP’s contribution is restricted to the production of response tokens signalling that he or she is taking part in the interaction but does not intend

Communicative strategies used by spouses

7

Table 2. Themes, categories and inter-rater reliability (per cent agreement) for the strategies used by CPs in conjunction with signs of PwCDs having problems expressing themselves Themes No participation in repair Request for clarification or modification

Providing candidate solutions

Categories 1 2 3 4 5 6 7 8 9

Response token Contribution for flow Topic shift Open-class initiation of repair Request for alternative or augmentative communication Request for specified clarification Request for verbal repetition or use of verbal modality Guess/completion/suggestion Elaboration/specification

to initiate or participate in repair work. In the extract below, the PwCD is talking about doing the laundry before going to her speech therapist. Her speech is halted by several short pauses and hesitation sounds, but the CP waits and only produces a response token: PwCD: so eh (0.5) one more (0.5) CP: mm hm PwCD: machine then I go (0.3) and I’m late for (0.5) therapist

When using the second type of strategy, the CP maintains the flow of the conversation by contributing a comment or a question making the conversation proceed within the area of the topic initiated (category 2: contribution for flow). In the extract below, the PwCD is describing his day at the daycare centre. His contribution is left unfinished, and after a 3-s pause the CP asks what was served for lunch at the centre: PwCD: yes I had time to read for a while then so it was it (3.0) CP: yes then it was lunch and then what did you get today then?

In this case the CP cooperates by moving the conversation forward, without trying to resolve the problem, that is what has been termed a ‘closing’ by Oelshlaeger and Damico (2000). The third type of strategy used by CPs that was not related to the initiation of or participation in repair involved changing the topic (category 3: topic shift). The couple in the extract below have been talking about the achievements of the participants in a dance course they are giving. After the CP’s positive assessment the PwCD agrees but the last ‘yes’ she produces is prolonged and the fact that she starts drawing in her sketchbook signals that she wants to express something more. Her husband looks at her drawing for a couple of seconds and then shifts the topic to the planning of their garden without giving any feed back to her drawing. He signals the topic shift by saying that he has been thinking about their home: CP: but yes but they are good anyway PwCD: yes yees ((starts drawing))

Inter-rater reliability 96% 85% 100% 100% 0% 100% 100% 60% 100%

CP: what eh (2.0) well I have been thinking about home now about the tiling and all that

In all three cases, the signs of the PwCDs having problems expressing themselves are left without the CPs taking active part in repair. Four of the categories (numbers 4–7, belonging to the theme of Request for clarification or modification) related to different types of repair initiation by the CP. Category 4 included occurrences of open-class initiation of repair (Drew 1997) where the CP did not specify what part of the PwCD’s contribution needed to be clarified—by saying, for example, ‘huh?’ or ‘what do you mean?’. In the extract below, the PwCD is speaking in a soft voice and repeats the first syllable when he is responding to the CP’s question about what he did during a meeting with the researcher; after a brief pause, the CP produces an open-class initiation of repair: PwCD: I I had to say words CP: (0.3) what? PwCD: had to say words

Category 5 (request for use of alternative or augmentative communication) included cases involving the initiation of use of artefacts such as a communication book or pencil and paper. In the extract below, the PwCD has been trying to ask her husband a question and she has been drawing in a sketchbook as a means to express herself. The CP is trying to guess what she wants to ask, and when his guess is rejected he encourages her use of paper and pen, suggesting that she should use a new page in the sketchbook for her drawing: CP: it is Friday today if you want to know no PwCD: no (1.0) eh (2.0) yes ((draws)) CP: take a new sheet now then or? ((turns page to a new sheet))

In category 6 (request for specified clarification), the request for clarification was specified, for example, by the repetition of the parts that were understood or by the production of a wh-question (e.g. ‘who?’; ‘when?’; ‘where?’), as described by Schegloff et al. (1977). In the extract below, the PwCD has just initiated a new topic

8

Emilia Carlsson et al.

and produces a negative assessment about the instruments used to measure cognitive ability in a study he is participating in. However, his wife seems to have difficulty understanding what the new topic is, and she asks him to specify what he is talking about: PwCD: no but it is on (0.3) it is on coarse (0.6) measure (0.5) instruments (0.9) CP: me- which?

Category 7 (request for verbal repetition or use of verbal modality) included cases where the CPs asked the PwCDs to try to express orally what they meant or to use a specific target word after the PwCDs had tried to express themselves with other means, such as gesturing or circumlocutions. Pedagogical cueing, such as providing the first sound of a target word, was included in this category. However, this type of action occurred only in one of the dyads, in whose conversations pedagogical sequences sometimes occurred. In the extract below, the PwCD has been telling her husband that she has been painting at her daycare centre. The CP corrects her, and when she says she cannot say the word and instead writes the first letters, he urges her to say the word by providing the first sound (note that the normal Swedish word for ‘watercolour’ as a painting method is akvarell; this is not as clearly a technical term as the corresponding English loan from the French aquarelle): CP: but isn’t it water colour but is it called something else PwCD: m I can’t I like this ((writes the first letters on the table with her finger)) CP: yes you can (1.0) aPwCD: aquarelle CP: m

The third and last theme, Providing candidate solutions included occurrences of CPs trying to guess, by providing a target word or specific alternatives, what the PwCDs were trying to express (category 8: guess/completion/suggestion). In the extract below, the PwCD is telling the CP about a visit he has made to the church. He is trying to tell her that someone played the organ, and when he cannot find the words, as signalled by the hesitation sound and the pause, the CP provides two alternatives: PwCD: who played on eh (0.8) CP: on the organ or the piano?

The second type of strategies included in the final theme involved the CP trying to narrow down, sum up, or expand on the information given by the PwCD (category 9: elaboration/specification). In the extract below, the PwCD is describing how he finds himself being impatient nowadays. The pauses and the somewhat imprecise vocabulary indicate that he is having trouble expressing

himself, and the CP sums up her perception of what he is trying to express: PwCD: I don’t want to are not doing (0.5) that often doing (0.8) one thing CP: no PwCD: but have several things going (0.4) so it becomes petty and (1.8) CP: you jump from one activity to another PwCD: yes

In both these cases the participants cooperate in repair by providing suggestions for what the PwCD may have wanted to express. Reliability of the categorization The calculation of intra-rater reliability showed almost perfect agreement, according to the norms provided by Landis and Koch (1977), for the entire set of categories and all seven dyads: κ = 0.91 (p < 0.001). For the individual dyads, κ varied between 0.86 and 1.0. Inter-rater reliability was also very satisfactory when all categories were included: κ = 0.87 (p < 0.001). It was also mainly satisfactory in relation to the individual categories. The material used to assess reliability contained a single occurrence of category 5 (request for use of alternative or augmentative communication), and that occurrence was only perceived as such by one of the researchers. Category 8 (guess/completion/suggestion) had the clearly lowest rate of agreement of the remaining eight categories: 60%. However, the material used to assess reliability contained only five occurrences of this category, of which the raters agreed upon three while two were categorized by one rater as category 9 (elaboration/specification). The remaining seven categories had rates of agreement in the range of 85–100% (table 2). Distribution of the use of strategies To assess the distribution of the use of the different strategies, a total of 724 occurrences of strategy use from the interactions within the seven dyads were analysed and categorized using the set of strategies described above. Regarding the total number of occurrences per dyad, it can be noted that there does not seem to be a clear correlation between the severity of the communicative disability and the number of occurrences of strategy use. Among the PD dyads, one including a person in stage IV of PD (PD 2) had fewer occurrences than the other two dyads, where the person with a communicative disability was in stage III and IV, respectively, according to the five-grade scale of Hoehn and Yahr (1967). Among the SA dyads, one including a person with mild-to-moderate aphasia (SA 7) had more

Communicative strategies used by spouses

9

Table 3. Number of occurrences of strategy use within the different dyads PD dyadsa Number of occurrences

PD 1 69

PD 2 63

SA dyadsb PD 3 74

SA 4 58

SA 5 155

SA 6 147

SA 7 158

Notes: a PD dyad = dyad including a person with Parkinson’s disease. b SA dyad = dyad including a person with stroke-induced aphasia.

Figure 1. Distribution among the PD and SA dyads of strategies belonging to either of the three different themes: No participation in repair, Request for clarification or modification; and Providing candidate solutions.

occurrences than one including a person with severe aphasia (SA 4) (table 3). Regarding the distribution and frequency of the individual categories, the two most commonly used ones (categories 1 and 2) both belong to the theme of No participation in repair (figure 1). The largest numbers of occurrences of category 1 (response token) were found in SA 6 and SA 7, where the PwCDs both had fluent speech. By contrast, the other two SA dyads, where the PwCDs had non-fluent speech, had fewer occurrences of response tokens than the three PD dyads (table 4). A particularly large proportion of the response tokens were produced by the CP in SA 7, where the PwCD had Wernicke’s aphasia with very fluent, but paraphasic speech. During the PwCD’s extended sequences of incomprehensible speech, the CP mostly produced response tokens. Her requests for clarification

or attempts to elaborate or specify what she believed the PwCD had tried to express usually resulted in more trouble, owing to the comprehension problems. The highest relative frequency of use of contribution for flow (category 2) was seen in PD dyads 2 and 3. In PD dyad 2, where the PwCD had quite severe dysarthria, the CP used contribution for flow in 62% of all occurrences of strategy use. The third category in the first theme, topic shift, accounted for only 0–3% of all occurrences of strategy use across the seven dyads. The third-most commonly used type of strategy across the seven dyads (19% of all occurrences) was for the CP to try to provide candidate solutions by means of an elaboration or a specification of the information the CP believed that the PwCD was trying to convey (category 9). This category belongs to the theme of Providing candidate solutions. By contrast, the other category in

10

Emilia Carlsson et al.

Table 4. Occurrences of use of each category of strategy, as a percentage of all occurrences of strategy use within each of the seven dyads

1. Response token 2. Contribution for flow 3. Topic shift 4. Open-class initiation of repair 5. Request for use of alternative or augmentative communication 6. Request for specified clarification 7. Request for verbal repetition or use of verbal modality 8. Guess/completion/suggestion 9. Elaboration/specification

that theme (category 8: guess/completion/suggestion) accounted for only 3.5% (range = 0–8%) of the occurrences of strategy use in the SA dyads and 4% (range = 0–11%) in the PD dyads. However, the PwCD in PD 1 did exhibit more explicit cases of word search, thus offering opportunities for the CP to try to provide candidate solutions. In this context, it should also be mentioned that the persons with PD often either abandoned the topic or lost track of it when trying to use circumlocutions. Finally, the remaining types of strategies (categories 4–7 within the theme of Request for clarification or modification) accounted for only 0–11% of all occurrences across the seven dyads (table 4). Regarding the total number of occurrences of strategy use, no statistically significant difference was found between SA dyads and PD dyads. There were generally more occurrences of strategy use in the SA dyads than in the PD dyads, but this difference was not statistically significant (p = 0.400). The average number of occurrences was 68.7 (range = 63–74) for the PD dyads and 129.5 (range = 58–158) for the SA dyads (table 3). Note that the ranges overlap, meaning that a dyad with a person with aphasia does not necessarily manifest more occurrences of strategy use in conjunction with signs of problems than a dyad with a person with PD— specifically, SA 4 (where the PwCD has severe aphasia) had only 58 occurrences of strategy use while PD 1 had 69 and PD 3 had 74. Use of strategies in relation to diagnosis Comparison of the SA and PD dyads regarding the use of individual strategies yielded a mixed pattern. CPs in both types of dyads tended to avoid initiating repair (i.e. they tended to use strategies in the theme of No participation in repair). However, CPs in PD dyads were most likely to use contributions for flow (category 2), which represented 48% of all occurrences in the PD dyads, while the most common strategy in the SA dyads was use of response tokens (category 1)—i.e. waiting and encouraging the PwCD to proceed without initiating repair—which accounted for 38.5% of all occurrences

PD 1

PD 2

PD 3

SA 4

SA 5

SA 6

SA 7

19% 32% 3% 7% 0% 0% 0% 11% 28%

17% 62% 0% 13% 0% 0% 0% 0% 8%

15% 51% 1% 12% 6% 0% 0% 1% 14%

12% 40% 2% 3% 3% 0% 0% 4% 36%

10% 50% 1% 4% 1% 1% 0% 3% 30%

39% 17% 1% 2% 0% 1% 11% 8% 21%

77% 17% 0.5% 0.5% 2% 0% 0% 0% 3%

in the SA dyads (figure 2). However, it should be noted that there is a great deal of variation between individual dyads. There was a small—and statistically non-significant (p = 0.057)—difference between the PD and SA dyads in the frequency of open-class initiation of repair (category 4) (figure 2 and table 4). The CPs in the PD dyads used this in 11% (range = 7–13%) of the occurrences while the CPs in the SA dyads only did so in 2% (range = 0.5–4%) of the occurrences. Another small difference— also statistically non-significant but still noticeable—was that elaboration/specification (category 8) accounted for 20% (range = 3–36%) of all occurrences in the SA dyads but only 16.5% (range = 8–28%) in the PD dyads (figure 2). Furthermore, two categories—request for use of alternative or augmentative communication (category 5) and request for verbal repetition or use of verbal modality (category 7)—were observed only among the SA dyads; the latter was actually used only in one of the SA dyads (SA 6), where the woman, who had mild-to-moderate aphasia, frequently ended up practising naming with her husband. Discussion In this section, the results will be elaborated upon. To begin with, the reliability and validity of the set of categories developed will be discussed. This will be followed by a discussion about the types of strategies used and the finding that there were no obvious differences in the use of strategies depending on whether a dyads included a person with PD or a person with stroke-induced aphasia. The qualitative analysis of the strategies used in the video-recordings resulted in a set that comprised nine different categories. This is a comparatively large number and may represent a challenge in terms of reliability (especially given that some categories were very infrequent in the material). Although the inter-rater reliability of that set of categories was generally satisfactory, some of the definitions may need to be further specified. For example, in cases where the PwCD had not

Communicative strategies used by spouses

11

Figure 2. Distribution of the use of the different categories of strategies as a percentage of all occurrences of strategy use within the PD and SA dyads: RT = Response token; CFF = Contribution for flow; ToS = Topic shift; OCR = Open-class initiation of repair; AAC = Request for use of alternative or augmentative communication; RqC = Request for specified clarification; RqRM = Request for verbal repetition or use of verbal modality; GCS = Guess/completion/suggestion; ES = Elaboration/specification.

really been able to express any information, it was sometimes difficult to decide whether the CP guessed or tried to complete the PwCD’s information (i.e. category 8, guess/completion/suggestion) or rather tried to elaborate on or to specify the information that the PwCD had expressed in an unclear way (i.e. category 9: elaboration/specification). However, having a large number of categories may, at least when the categories are satisfactorily defined, increase the validity of a set. Another factor that would seem to favour the validity of the present categories is that they evolved in a data-based process, based on the analysis of strategies used in natural interactions in dyads affected by different types of communicative disabilities. On the other hand, as the aim was to create a reliable set of categories, possible subcategories were merged during the process to reduce the number of categories included. For example, category 6 (request for specified clarification) may comprise two of the categories included in the typology of Schegloff et al. (1977), namely both the use of question words in isolation and the partial repeat of the trouble-source turn together with a question word. This could be a shortcoming of the present set of cat-

egories, given that Schegloff et al. claim that the latter is a stronger type of repair initiator than the former. In relation to CP training, increased use by CPs of partial repeats of the trouble-source turn together with a question word could indicate a positive change in that it could result in more efficient repair. Still, although the validity of the present set of categories remains to be further explored using larger groups of participants with different types of communicative disabilities, this study does show that the description provided by Schegloff et al. of different types of repair initiators in conversation is not adequate to cover the entire range of what may happen in cases of trouble in interactions where one of the interlocutors has a communicative disability. Common to all PwCDs participating in this study was that they suffered, to different degrees, from problems expressing themselves due to either articulatory or phonological problems as well as anomia and other cognitive impairments. The validity of the results is further strengthened by the fact that there was variation in the type and degree of the participants’ problems, as shown by formal test results on word fluency and comprehension, and that the CPs used different types of strategies

12 adapted to the needs and resources of each PwCD. On the other hand, another commonality between the participants in this study was that they were all about to participate in an intervention intended to develop the CPs’ ability to support communication. This may mean that the CPs included were all persons who saw their own actions as a possible resource for the respective PwCD. This may have been reflected in their use of strategies, which might have been different if they had ascribed the communication problems solely to the PwCD. The two most common strategies across all dyads belonged to the theme of No participation in repair, meaning that the spouses were quite likely to abstain from initiating or taking part in repair in cases where a contribution from the PwCD had been perceived as problematic by the researchers. The CPs in both types of dyads tended to produce either response tokens or topic-related contributions that facilitated the flow of the conversation without apparently involving any repair work in relation to the most recent sign of problems. That is, the CP often did not treat incomprehensible or incomplete information as a problem. According to previous research, this is not uncommon, especially in relation to communication disorders, where repair may be avoided in order not to expose a person’s linguistic incompetence and thus to save face (Bloch and Beeke 2008, Goffman 1955, Wilkinson et al. 2003). Indeed, Barnes and Ferguson (2014) describe several practices used by communication partners of persons with aphasia as an alternative to repair work. Here it should be kept in mind that the spouses, being more familiar with the speech characteristics and experiences of their partners and with the situational context, may often have understood rather well what the researchers perceived as sequences of incomprehensible speech or incomplete information. In other words, in several of the cases classified as belonging to the theme of No participation in repair, repair was not initiated because no repair was needed. It is of course questionable whether such sequences should be analysed as problematic at all. On the other hand, however, not attending to such cases in an analysis of interactions including persons with communicative disabilities may distort the picture of the processes involved. Avoidance of repair may help a person to save face, but it may also affect the amount of information transferred, and persons with communicative disabilities are often dependent on the cooperation of their conversation partners to make themselves understood (Milroy and Perkins 1992, Perkins 2003). Another common strategy, irrespective of diagnosis, was for the CP to take part in repair and to try to elaborate on or to specify what the PwCD had meant to say (category 9: elaboration/specification). This category would include cases where the CP sought verification of his or her understanding of the previous turn,

Emilia Carlsson et al. which in the typology of Schegloff et al. (1977) represents the strongest type of repair initiator. Category 8: guess/completion/suggestion was used less often, which may seem surprising given that guess-and-completion strategies are often described as commonly used in relation to aphasia (Ferguson 1993, Lubinski et al. 1980, Oelshlaeger and Damico 2000). However, it should be noted that the present study, to enable a more elaborate analysis of strategies, made a distinction between two categories (8 and 9) within the theme of Providing candidate solutions which was not made in the abovementioned studies. What is more, as noted above, it was not always easy to decide whether an occurrence belonged to category 8 or 9. Together, these two categories accounted for over 20% of cases in both PD and SA dyads. The use of strategies belonging to the theme of Request for clarification or modification was not that common in this material, meaning that more data showing use of these strategies are needed for any general conclusions to be drawn. However, the infrequent use of these strategies may actually be seen as a sign of the adaptations made by the CP in interactions affected by communicative disabilities. We are not aware of any studies comparing the use of requests for clarification as between typical conversational interaction and interaction involving persons with communicative disabilities. However, it is a reasonable hypothesis that this strategy may be more common in cases where the person requesting clarification expects the other person to be able to perform the repair quite easily. Where this is not so, the person needing clarification may instead choose to adapt to the needs of the other person and cooperate by providing a guess or alternatives. When it comes to differences in the use of strategies due to the type of diagnosis causing the communicative disability, the results are complex. Some of the weak tendencies towards differences seen between the PD and SA dyads may be related to the small number of dyads included in this study. Other factors affecting the results may be the differences across dyads in the severity of symptoms, age and other individual factors. The SA dyads generally tended to have more occurrences of strategy use than the PD dyads, although this was not always the case. Nor was the stage of PD or the severity of the aphasia always related, in an obvious way, to the number of occurrences of strategy use. Regarding the frequency of individual types of strategies, there was no statistically significant difference between the dyads that could be related to the type of diagnosis. The CPs in both PD and SA dyads often avoided initiating repairs or encouraged the PwCDs to perform the repair themselves. The frequent tendency for CPs in PD dyads to produce a comment for flow, without taking part in repair, may not be fully explained by the CPs having understood the

Communicative strategies used by spouses speech better than the researchers performing the analysis. In fact, contributions for flow (category 2) were also common in the two SA dyads where the persons with aphasia had non-fluent speech and were often not able to make themselves understood. The urge to save face by avoiding cumbersome repair may be a good reason to produce only a comment enabling the conversation to proceed. The CPs in the PD dyads tended to use openclass initiation of repair (category 4) more often than the CPs in the SA dyads. In line with the reasoning above, this could be related to the fact that in the case of the persons with PD, the trouble was often caused by motor speech problems and they were often able to make the repair themselves by repeating or rephrasing their messages. With the exception of SA 7, the use of strategies belonging to the theme of Providing candidate solutions (categories 8 and 9) was quite common in the SA dyads, but it was equally common in PD 1—where the PwCD did exhibit more explicit occurrences of word search, thus offering opportunities for the CP to try to provide candidate solutions. The purpose of this study was to develop a reliable set of categories to describe the strategies used by conversation partners of persons with communication disorders due to stroke-induced aphasia or to PD. Such a set could be used, both in clinical practice and in future research, to assess the effectiveness of different types of strategies in specific dyads as well as in relation to different diagnoses. It remains a task for future research to explore possible patterns of sequential order in the use of these different strategies, as described in the typology of Schegloff et al. (1977).

Conclusions According to the results of this study, the distribution of the types of strategies used by the CPs seems to depend more on individual factors in the different dyads, such as the ability to use language, than on whether the communicative disability is due to stroke-induced aphasia or to PD at later stages. Although the set of categories needs to be trialled with larger groups of participants before it can be regarded as a valid system for the categorization of strategies used by CPs, the present findings suggests it could be developed into a useful and reliable instrument to be used for assessment before and after CP training.

Acknowledgements Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

13 References BARNES, S. and FERGUSON, A., 2014, Conversation partner responses to problematic talk produced by people with aphasia: some alternatives to initiating, completing, or pursuing repair. Aphasiology. doi: 10.1080/02687038.2013.874547 ¨ RNRAM, C., HARTELIUS, L., LAAKSO, K. and JOHNELS BERG, E., BJO B., 2003, High-level language difficulties in Parkinson’s disease. Clinical Linguistics and Phonetics, 17, 63–80. BLOCH, S., 2005, Co-constructing meaning in dysarthria: word and letter repetition in the construction of turns. In K. Richards and P. Seedhouse (eds), Applying Conversation Analysis (Basingstoke: Palgrave Macmillan), pp. 38–55. BLOCH, S. and BEEKE, S., 2008, Co-constructed talk in conversations of people with dysarthria and aphasia. Clinical Linguistics and Phonetics, 22, 974–990. BLOCH, S. and WILKINSON, R., 2009, Acquired dysarthria in conversation: identifying sources of understandability problems. International Journal of Language and Communication Disorders, 44(5), 769–783. BLOCH, S. and WILKINSON, R., 2011, Acquired dysarthria in conversation: methods of resolving understandability problems. International Journal of Language and Communication Disorders, 46(5), 510–523. DE RENZI, E. and VIGNOLO, L. A., 1962, The token test: a sensitive test to detect receptive disturbances in aphasia. Brain, 85, 665–678. DREW, P., 1997, ‘Open’ class repair initiators in response to sequential sources of troubles in conversation. Journal of Pragmatics, 28, 69–101. FERGUSON, A., 1993, Conversational repair of word-finding difficulty. In M. L. Lemme (ed.), Clinical Aphasiology (Austin, TX: PRO-ED), vol. 21, pp. 299–310. FORSGREN, E., ANTONSSON, M. and SALDERT, C., 2013, Training conversation partners of persons with communication disorders related to Parkinson’s disease—a protocol and a pilot study. Logopedics Phoniatrics Vocology, 38, 82–90. GOFFMAN, E., 1955, On face-work: an analysis of ritual elements of social interaction. Psychiatry: Journal for the Study of Interpersonal Processes, 18, 213–231. GOODGLASS, H. and WINGFIELD, A., 1997, Anomia: Neuroanatomical and Cognitive Correlates (San Diego, CA: Academic Press). GOODWIN, C., 1995, Co-constructing meaning in conversations with an aphasic man. Research on Language and Social Interaction, 28, 233–260. GRICE, H. P., 1975, Logic and conversation. In P. Cole and J. Morgan (eds), Syntax and Semantics 3: Speech Acts (New York, NY: Academic Press), pp. 41–58. GRIFFITHS, S., 2013, Multiple repair sequences in everyday conversations involving people with Parkinson’s disease. Paper presented at the ‘Atypical Interaction—Conversation Analysis and Communication Impairments’ conference, University of Sheffield, Sheffield, UK, June 2013. GRIFFITHS, S., BARNES, R., BRITTEN, N. and WILKINSON, R., 2011, Investigating interactional competencies in Parkinson’s disease: the potential benefits of a conversation analytic approach. International Journal of Language and Communication Disorders, 46, 397–509. GRIFFITHS, S., BARNES, R., BRITTEN, N. and WILKINSON, R., 2012, Potential causes and consequences of overlap in talk between speakers with Parkinson’s disease and their familiar conversation partners. Seminars in Speech and Language, 33(1), 27–43. HARTELIUS, L., LINDBERG, J., PETERSSON, L. and SALDERT, C., 2011, Perceived changes in communicative interaction in atypical Parkinsonism. ISRN Neurology, 2011, 256406.

14 HARTELIUS, L. and SVENSSON, P., 1994, Speech and swallowing symptoms associated with Parkinson’s disease and multiple sclerosis: a survey. Folia Phoniatrica, 46, 9–17. HOEHN, M. M. and YAHR, M. D., 1967, Parkinsonism: onset, progression and mortality. Neurology, 17, 427–442. KAGAN, A., WINCKEL, J., BLACK, S., DUCHAN, J. F., SIMMONSMACKIE, N. and SQUARE, P., 2004, A set of observational measures for rating support and participation in conversation between adults with aphasia and their conversation partners. Topics in Stroke Rehabilitation, 11, 67–83. LAAKSO, M. and KLIPPI, A., 1999, A closer look at the ‘hint and guess’ sequences in aphasic conversation. Aphasiology, 13, 345–363. LANDIS, J. R. and KOCH, G. G., 1977, The measurement of observer agreement for categorical data. Biometrics, 33, 159–174. ¨ , E. and WERNER, C., 1995, A-ning, Neurolingvistisk LINDSTROM Afasiunders¨okning (Stockholm: Ersta utbildningsinstitut). LUBINSKI, R., DUCHAN, J. and WEITZNER-LIN, B., 1980, Analysis of breakdowns and repairs in aphasic adult communication. In R. H. Brookshire (ed.), Clinical Aphasiology: Conference Proceedings (Minneapolis, MN: BRK), pp. 111–116. MILLER, N., NOBLE, E. and BURN, D., 2006, Life with communication changes in Parkinson’s disease. Age and Ageing, 35, 235–239. MILROY, L. and PERKINS, L., 1992, Repair strategies in aphasic discourse: towards a collaborative model. Clinical Linguistics and Phonetics, 6, 27–40. OELSHLAEGER, M. L., 1999, Participation of a conversation partner in the word searches of a person with aphasia. American Journal of Speech–Language Pathology, 8, 62–71. OELSHLAEGER, M. L. and DAMICO, J. S., 2000, Partnership in conversation: a study of word search strategies. Journal of Communication Disorders, 33, 205–225. ORANGE, J. B., LUBINSKI, R. D. and HIGGINBOTHAM, D. J., 1996, Conversational repair by individuals with dementia of the

Emilia Carlsson et al. Alzheimer’s type. Journal of Speech and Hearing Research, 39, 881–895. PERKINS, L., 2003, Negotiating repair in aphasic conversation. In C. Goodwin (ed.), Conversation and Brain Damage (New York, NY: Oxford University Press), pp. 147–162. SALDERT, C., FERM, U. and BLOCH, S., 2014, Semantic trouble sources and their repair in conversations affected by Parkinson’s disease. International Journal of Language and Communication Disorders (in press). SCHEGLOFF, E. A., 1982, Discourse as an interactional achievement: some uses of ‘uh huh’ and other things that come between sentences. In D. Tannen (ed.), Analyzing discourse: Text and Talk (Washington, DC: Georgetown University Press), pp. 71–73. SCHEGLOFF, E. A., JEFFERSON, G. and SACKS, H., 1977, The preference for self-correction in the organization of repair in conversation. Language, 53, 361–382. SIDNELL, J., 2010, Conversation Analysis. An Introduction (Oxford: Wiley-Blackwell). SIMMONS-MACKIE, N. and KAGAN, A., 1999, Communication strategies used by ‘good’ versus ‘poor’ speaking partners of individuals with aphasia. Aphasiology, 13, 807–820. SIMMONS-MACKIE, N., RAYMER, A., ARMSTRONG, E., HOLLAND, A. and CHERNEY, L. R., 2010, Communication partner training in aphasia: a systematic review. Archives of Physical Medicine and Rehabilitation, 91, 1814–1837. THORNE, S, REIMER KIRKHAM, S. and O’FLYNN-MAGEE, K., 2004, The analytic challenge in interpretive description. International Journal of Qualitative Methods, 3(1), 1–21. WILKINSON, R., BEEKE, S. and MAXIM, J., 2003, Adapting to conversation. In C. Goodwin (ed.) Conversation and Brain Damage (New York, NY: Oxford University Press), pp. 59–89.