research NO PLAGERISM

Recipient participation in conversations involving participants with fluent or non-fluent aphasia Minna Laaksoa and Sisse Godtb

aDepartment of Psychology and Speech-Language Pathology, University of Turku, Turku, Finland; bDepartment of Behavioural Sciences, University of Helsinki, Helsinki, Finland

ABSTRACT The present study compares the ways in which conversational partners manage expressive linguistic problems produced by participants with fluent vs. non-fluent aphasia. Both everyday conversations with family members and institutional conversations with speech-language thera- pists were examined. The data consisted of 110 conversational sequences in which the conversational partners addressed expressive aphasic pro- blems. Most problems of the speaker with fluent aphasia were locally restricted phonological and word-finding errors, which were immediately repaired. In contrast, the sparse expression of the speaker with non-fluent aphasia was co-constructed by conversational partners in long negotia- tion sequences to establish shared understanding. Some differences between recipient participation ineveryday and institutional conversation were found. The results emphasise the relevance of the nature of the expressive linguistic problems on participation in interaction. They also add to the clinical knowledge of handling aphasic problems in conversa- tion. This knowledge can be used for developing interaction-focused intervention.

ARTICLE HISTORY Received 1 December 2015 Accepted 4 August 2016

KEYWORDS Aphasia type; co- construction; family members; other-repair; speech-language therapists

Introduction

The objective of the present study is to compare the ways in which participants manage expressive linguistic problems related to fluent conduction aphasia and non-fluent agrammatic aphasia in institutional and everyday settings. In general, conversations involving participants with aphasia (PWAs) are characterised by frequent expressive linguistic problems and efforts to resolve the problems in extended conversational sequences (e.g. Laakso & Klippi, 1999; Milroy & Perkins, 1992). In problem-solving sequences both PWAs and their interlocutors participate in clarifying meanings and establishing shared understanding. However, collaborative partici- pation may differ between family members and speech-language therapists (SLTs) (Lindsay & Wilkinson, 1999; Laakso, 2015). Besides the difference in everyday and institutional participa- tion roles, the type of aphasia may also influence the participation of the interlocutors: fluent production with frequent linguistic errors may call upon different collaborative actions than non-fluent and linguistically sparse production.

Conversational speech shows adaptations to the underlying linguistic deficits characteristic of the type of aphasia. In fluent conduction aphasia, speech is frequently distorted by phonemic paraphasias (sound errors) and efforts to repair the problem by approximating

CONTACT Minna Laakso [email protected] Department of Psychology and Speech-Language Pathology, University of Turku, Assistentinkatu 7, Turku, FIN-20014, Finland.

CLINICAL LINGUISTICS & PHONETICS 2016, VOL. 30, NO. 10, 770–789 http://dx.doi.org/10.1080/02699206.2016.1221997

© 2016 Taylor & Francis

the right sound (Bartha & Benker, 2003; Kohn, 1984). Due to the underlying phonological deficit, word-retrieval difficulties and word searching occur. In addition, speakers with conduction aphasia may try to overcome word-finding problems with circumlocutory para- phrasing or by using easily retrievable but semantically weak lexical items (e.g. Helasvuo, Laakso, & Sorjonen, 2004). In non-fluent agrammatic aphasia speech is characterised by the omission of grammatical morphemes, incomplete and incorrect syntactic constructions, and short utterances (Menn, O’Connor, Obler, & Holland, 1995). As the underlying syntactic deficit impairs verb and sentence production, speakers with agrammatic aphasia often resort to telegraphic speech style by producing single-word utterances that are mostly uninflected nouns (Heeschen & Schegloff, 1999). One adaptation in conversation is the fronting of noun phrases to the beginning of the turn, so that the speakers overuse left dislocation in their turn construction (Beeke, Wilkinson, & Maxim, 2003). Furthermore, the meaning of the turn may emerge bit by bit by producing sequentially adjacent items without tying them together by grammatical means (Beeke, Wilkinson, & Maxim, 2007; Heeschen & Schegloff, 1999). In short, the two types of aphasia, fluent conduction aphasia and non-fluent agrammatic aphasia, have different (phonological vs. syntactic) underlying deficits that manifest in dissimilar ways in linguistic expression in conversation.

In the present study, the main focus is on the recipient-initiated next turn actions that aim to manage expressive linguistic problems produced by the PWA in a prior turn. In conversation analysis (CA), expressive problems are characterised as trouble sources that can be managed by self- or other-initiated repair. PWAs are not as efficient in self- repairing their own speech as is common in ordinary conversation (cf. Laakso, 1997: 130–140). Thus other-initiated actions by the interlocutors become more emphasised. Other-initiated repair can be accomplished in many ways, including asking questions (such as ‘what’, ‘who’, etc.), repeating the trouble source, or by offering candidate under- standings for the recipient to confirm or reject (Schegloff, 2007: 101). In ordinary everyday conversation the recipients usually do not directly correct the original speaker, as self- repair is preferred over other-correction (Schegloff et al., 1977). However, in asymmetric interactions (e.g. between adults and children) it is common that the more competent interlocutors are active in resolving the problems and do other-correct (Norrick, 1991). Also in connection with aphasia, co-participants are active and tend to provide candidate words to resolve e.g. word searching (Laakso & Klippi, 1999; Oelschlaeger & Damico, 2000). However if the PWA has non-fluent speech, the interlocutors often co-construct the sparse talk by adding meaning and extending the agrammatic elements into full utterances (Goodwin, 1995; Heeschen & Schegloff, 1999). Interlocutors also interpret embodied phenomena such gestures, gaze and facial expressions in co-constructing the meaning of aphasic utterances (e.g. Beeke et al., 2013; Goodwin, 2003; Klippi, 2015; Laakso, 2014). Besides interpretation, conversational partners may use an open-format other-initiation of repair (e.g. ‘what’ or ‘sorry’) which returns the speaking turn to the PWA (Barnes, 2016). In such cases the PWAs may have difficulties in resolving the problem. The partners may also resist participating in problem solving (e.g. Aaltonen & Laakso, 2010; Barnes & Ferguson, 2015). However, resistance is not the focus of the present study as it analyses collaborative recipient actions.

In collaborative participation, some differences between everyday and institutional conversational partners have been found. When compared to encounters with every- day interlocutors, the interaction with representatives of a medical institution (e.g. an

CLINICAL LINGUISTICS & PHONETICS 771

SLT in health care) is usually more goal oriented and has special constraints as to what are the appropriate conversational actions (for medical interaction, see e.g. Heritage & Maynard, 2006). These constraints may result in differences in an SLT’s actions in conversation with a PWA as compared to a conversation between everyday conversation partners. Indeed, differing repair patterns in spouse vs. SLT conversa- tions of PWAs have been found: The spouses tend to repeatedly correct the PWAs’ production errors even after the meaning has been recovered, whereas the SLTs avoid correction (Lindsay & Wilkinson, 1999). In addition, significant others are found to offer collaborative completions to word searches, whereas SLTs tend to ask questions or offer candidate understandings longer than one word (Laakso, 2015). However, thus far, very few studies have compared recipient-initiated repair activities in institu- tional and home conversations of the same PWAs. Furthermore, to our knowledge, the influence of different types of aphasias on recipient participation has not been compared.

The purpose of the present multiple-case study is to compare the relevance of the aphasia type (fluent vs. non-fluent) to the non-aphasic interlocutors’ management of expressive linguistic problems in conversation. The research focus is on recipient-initiated actions. Everyday vs. institutional conversational partners are also compared. The research questions are as follows:

● What kind of expressive linguistic problems and problem-handling sequences man- ifest in conversations involving PWAs with fluent vs. non-fluent aphasia?

● Are the recipient-initiated problem-handling practices of the non-aphasic interlocu- tors different in conversations with:

● participants with fluent vs. non-fluent aphasia? ● family members vs. SLTs?

Data and method

Conversations and participants

The data consisted of four videotaped conversations (see Table 1). In these conversations there were two kinds of conversational partners, family members and SLTs, in interaction with participants with two different kinds of aphasia, Lauri (L) with conduction aphasia and Kalevi (K) with agrammatic aphasia. L conversed at home both with his young adult

Table 1. Conversations and participants. Conversation Participants Location Duration

Conversation 1 L (fluent speaker), L’s home, on the living room couch 00:40:00 hours T and A (two grandsons)

Conversation 2 L (fluent speaker), L’s home, on the living room couch 00:41:16 hours SLT-L

Conversation 3 K (non-fluent speaker), K and E’s home at the kitchen table during afternoon coffee

00:34:11 hours

E (the wife) Conversation 4 K (non-fluent speaker), SLT-K University clinic 00:27:01 hours Total 02:22:18 hours

Note. SLT-L = Speech-language therapist working with L; SLT-K = Speech-language therapist working with K.

772 M. LAAKSO AND S. GODT

grandsons (T and A), and with his SLT (SLT-L). K interacted at home with his wife (E) and at the university clinic with his SLT (SLT-K). Pseudonyms are used and name initials of all participants have been changed to anonymise them. All participants gave in writing their informed consent to take part in the study and the principles of research ethics were carefully followed in gathering and analysing the data. The data belong to a larger research project examining the management of problems of speaking and understanding in aphasic conversation (Academy of Finland, grant no 49250).

The conversations of L lasted approximately 40 minutes each, whereas K’s conversa- tions lasted about 30 minutes each. The conversations were digitally videotaped on a MiniDV tape with an additional microphone. The PWAs and their family members chose a time and place where they usually interacted (e.g. a regular visit or an afternoon coffee) and the cameraman (researcher or research assistant) came to videotape that. The participants were asked to interact in the way they ordinarily would. The cameraman was present in the beginning and end of (and occasionally during) the videotaping when handling the camera was necessary. As both PWAs were receiving speech and language therapy via a university clinic, they were asked to converse with their SLTs during a therapy session. No specific instructions concerning the conversations were given. These sessions were videotaped by the participating SLTs.

Both PWAs were retired males with chronic aphasia (see Table 2). They both had a university degree. L (a former CEO) was a 79-year-old man who had suffered an infarc- tion of the left medial cerebral artery 5 years ago. The infarction had left him with moderately severe conduction aphasia, and he had had several periods of speech and language therapy. L’s language production was fluent but paraphasic with many phonemic and some semantic errors. He also displayed trouble with word-finding and recurrently searched for words. K (a former dentist) was a 65-year-old man who had suffered an infarction of the left medial cerebral artery 29 years ago. The infarction had left him with severe Broca’s aphasia, and he had had intensive speech and language therapy during the first year following his infarct and shorter periods after that. K’s language production was non-fluent and agrammatic consisting mainly of uninflected nouns.

In conversation, L was able to use language in various ways for initiating and respond- ing (including questions, answers, comments and narrative turns), whereas K’s commu- nicative actions were mainly responsive to the interlocutors’ actions.

Transcription and analysis

The video material was analysed as Windows Media files. First the conversations were transcribed according to conversation analytic notation (Atkinson & Heritage 1984: ix–xvi).

Table 2. Background data of the participants with aphasia.

Participant Lesion

localization Duration of

illness Diagnosis Aphasia type

(BDAE) BNT score

Severity of aphasia (BDAE)

L, male, 79 years MCA 5 years CVA Conduction/fluent 34/60 3 K, male, 65 years MCA 29 years CVA Broca/non-fluent 17/60 2

Note. MCA = medial cerebral artery, CVA = cerebrovascular accident, BDAE = Boston Diagnostic Aphasia Examination (Goodglass & Kaplan, 1983; Finnish version Laine, Niemi, Tuomainen & Koivuselkä-Sallinen, 1997); BNT = Boston Naming Test (Kaplan, Goodglass, Weintraub & Segal, 1983; Finnish version Laine, Koivuselkä-Salllinen, Hänninen & Niemi, 1997); BDAE severity scale 1 = most severe, 5 = mildest).

CLINICAL LINGUISTICS & PHONETICS 773

Along with speech, embodied actions such as hand gestures were described in small capitals within double parentheses on a separate line below the utterance they co-occurred with. For the sake of clarity, the data extracts include embodied actions only if they are relevant for the present analysis. In the data extracts (see example 1), an English translation is provided (in bold face) alongside the original transcription (in italics). When of analytic interest, the talk of each person is depicted in three lines consisting of the original Finnish, an English word- by-word gloss (see glossing symbols in the Appendix), and an English translation. In the gloss line, unintelligible word forms are marked with question marks and, if possible, targets of these word forms in curly brackets (see line 02 in example 1):

The results were obtained through CA paying special attention to repair organisation (Sacks, 1992; Schegloff et al, 1977). First, all instances where any of the speakers oriented to a section of a conversation as problematic were identified. These problem handling instances were divided into sequences by topical organisation. Out of these topical sequences a total of 72 sequences were such where the non-aphasic interlocutors, i.e. the recipients of the PWAs’ talk, indicated the previous turns of the PWAs as problematic. These were referred to as negotiation sequences. During this preliminary analysis, it was observable that the negotiation sequences in conversations involving the two men were different. The conversations of the fluent speaker allowed for a turn-by-turn analysis of repair organisation while in the conversations of the non-fluent speaker separate repair sequences were not easily identifiable. Thus, the initial 27 negotiation sequences of the fluent speaker were further analysed for local repair phenomena, and altogether 65 repair sequences of two or more turns during which the problem was dealt with were identified (see Table 3). In connection with non-fluent aphasia, problem handling manifested in almost all conversational turns of the participants without clearly identifiable repair sequences, and could only be investigated at a broader topical sequential level. Thus, forcing the 45 topical negotiation sequences of the non-fluent speaker into more concise repair sequences was not appropriate. The sequences analysed accordingly consisted of 110 problem-handling sequences in total (see Table 3).

Results

The nature of expressive linguistic problems in conversations of participants with fluent vs. non-fluent aphasia

The trouble sources of the speaker with fluent aphasia were identified either as word search (see example 2; cf. also Goodwin & Goodwin 1986) or as a problem of general intelligibility (see example 3, cf. also Damico et al., 2008), when a clear word search was not present but

(1) Example of the transcription, glossing and translation of data extracts.

01 L: onko se- tuo- (.) hh .mt Ville kertonu is-Q he that 1nameM tell-PPC has he- that- (.) hh .mt Ville told you

02 minkälaista se on (0.5) amer- amerissa- (-) what.sort.of-PAR it is {army}-INE what it is like (0.5) in the a- ar- (-)

774 M. LAAKSO AND S. GODT

something in the turn was unintelligible enough to stop the flow of conversation, and led to a subsequent repair initiation in the next turn by the interlocutor. In the following example (2), a word search occurs (lines 01–05). The fluent speaker L is talking with his two grandsons (A and T) about global politics. On line 01, L starts a new topic and ends up trying to produce a place name in Africa:

L’s utterance on lines 01–05 includes many search markers such as pauses (three of them extending to one second, see lines 02, 04, 05), comments on his own knowledge or ability to find the word: ‘I don’t know what it is’, (line 01) and attempts at approximating the phonemic structure of the target words (lines 02–05). On line 06 his grandson A offers his suggestion for the searched word, Eritrea. It is noteworthy that A’s suggestion phonemi- cally resembles L’s last attempt at the word. Due to the nature of his aphasia L is able to produce expression referring to a place (‘there’, line 03) and phonemic approximations of the place name (lines 04–05), which may make it easier for the recipient to offer a

Table 3. Analysed problem handling (repair and negotiation) sequences. Conversation Sequences

Conversation 1 29 repair sequences (L, grandsons)

Conversation 2 36 repair sequences (L, SLT)

Conversation 3 29 negotiation sequences (K, wife, cameraman)

Conversation 4 16 negotiation sequences (K, SLT)

Total 110

(2) Word search. (The possible targets of erroneous or cut-off word forms are in curly brackets in the gloss line. If no target can be traced the word is marked with a question mark on the gloss line.)

01 L: Mmm se:(.) en tiiä- (.) en tiiä hh mikä se on it NEG-1 know-INF NEG-1 know-INF what it is

Mmm it:(.)I don’t know-(.)I don’t know hh what it is

02 e- e- e: e::len kaks (a: h eutrah (1.3) {eilen=yesterday} two ? e- e- e: {yesterday} two (a: h eutrah (1.3)

03 epron hh)(0.7).HH sielä hh ö: (0.5).mt lasa hh ? there ?

epron hh)(0.7).HH there hh uh: (0.5).mt lasa hh

04 lasarin hhh (1.0).hhh apessiina (.) apessiina hh ? {Abessiinia=Abyssinia} {Abyssinia}

lasarin hhh (1.0).hhh {Abyssinia} (.){Abyssinia}hh

05 (1.0) vedutrea hh {Eritrea}

(1.0) {vedutrea} hh

06 A: Eritrea niinkö. Eritrea PRT-Q (appr. you mean) you mean Eritrea.

07 L: Nii. PRT Right.

CLINICAL LINGUISTICS & PHONETICS 775

candidate interpretation. Grandson’s candidate is formulated as a polar question with the inferencing particle niin (its meaning here as an independent clausal unit corresponds to ‘right’) and the question clitic -kö, which is then confirmed as correct by L.

Example (3) illustrates the other characteristic trouble source by speaker L, a problem of general intelligibility (see lines 04–05). L is talking with his SLT (SLT-L) about the heating system in his house. In the beginning of the sequence, SLT-L asks which heating system the house has (line 01):

On line 02, L responds without problems that his house has district heating, a heating system in which a centralised plant using gas or oil delivers heating for a whole town suburb. From line 04 onwards L begins to expand his original response but is not able to continue his utterance to an understandable completion. Instead, after describing the district heating as ‘it burns somewhere’, he only produces some cut-off attempts at words (line 05). In overlap with L’s attempts the SLT-L starts to process the trouble by offering her candidate understanding (line 06). Using the particle niin (here in turn-initial position appr. ‘I see’), SLT-L’s candidate ‘somewhere far away’ is formulated as inferen- cing from L’s prior talk about district heating (‘far-heat’, line 02) rather than on the phonemic structure of L’s cut-off attempts on line 05. It is also produced as a completion to fit L’s utterance: ‘it burns somewhere’ – ‘somewhere far away’. However, as the exact nature of the expression L was striving at remains unintelligible, he only acknowledges SLT-L’s candidate resolution with ‘mmm’ (line 07).

In sum, the expressive problems in the turns of the speaker with fluent conduction aphasia were regularly phonemically distorted words that were cut off to search for a better approximate. In some cases, the cut-off attempts provided information for the recipient to interpret the word and offer a specific candidate (example 1). In the other cases, the recipients produced more general interpretations based on L’s prior talk (example 2). L’s expressive problems were thus specific occurrences amongst otherwise understandable conversation that could be treated with specific recipient-initiated repair

(3) Problem of general intelligibility.

01 SLT-L: Onko tässä niinku öljy vai sähkölämmitys tässä- is-Q here-INE PRT(like)oil or electric heating here-INE Is it like oil or electric heating system in here-

02 L: Se on kaukolämpö. it is disctrict.heating (literally: ‘far-heat’) It is district heating.

03 SLT-L: Kaukolämpö [joo. District heating [yeah.

04 L: [Nii.(0.8) se palaa(2.0)jossaki PRT it burn-3 somewhere [Right.(0.8) it burns(2.0)somewhere

05 (tapah- (.) [ve-) {tapahtua=happen} ? (hap- (.) [ve-)

06 SLT-L: [Nii jossain kaukana. PRT somewhere far.away [I see somewhere far away.

07 L: Mmm.

776 M. LAAKSO AND S. GODT

practices. The participation structure of the repair sequences in L’s conversations mani- fested as follows:

(1) There is an identifiable problem in the turn of the fluent aphasic speaker. (2) The recipient offers a candidate resolution in the next turn. (3) The PWA confirms/rejects the repair.

On the contrary, in the conversations of the non-fluent speaker almost all speaking turns of the PWA were problematic due to his limited linguistic expression. Most of K’s turns could be characterised as minimal expression which led to a telegraphic construction of a turn (see example 4, lines 04–05). In this example the non-fluent speaker K is talking with his wife (E) about his hobby, photography:

In example (4), the wife introduces a new topic to the conversation by asking a question (line 01). The non-fluent speaker K responds to the question minimally with an affirmative particle joo ‘yes’, and does not develop the topic any further (line 02). His turn thus provides the most minimal possible answer to E’s question. On line 3, the wife then develops the topic by asking a follow-up question. To this question K responds with a more extended answer that he constructs in telegraphic style (lines 04–05; cf. Beeke et al., 2003; Heeschen & Schegloff 1999). He starts with a clausal construction joo mutta toi (‘yes but that’), cuts it off and pauses for almost 5 seconds. After the pause he produces isolated words: the elements of K’s turn are adjacent, separated with pauses, but do not form a clear clausal structure. He frequently uses the pronoun toi ‘that’ for projecting that there is more to come, as toi is used as a modifier to a noun (cf. Helasvuo et al., 2004). Most of the words he uses are not inflected, and finally he is able to say his main point, ‘25 pictures’, although without a clear sentence structure its meaning is hard to interpret. However,

(4) Minimal expression and telegraphic construction of a turn.

01 E: Onks valokuvia viel tekemättä. is-Q photograph-PL-PAR still do-INF Are there still photographs to develop.

02 K: Joo. Yes.

03 E: Otit sä sielt Matin synttäreilt kuvia. Take-2-PST you there-ABL 1nameF-GEN birthday-ABL picture-PL-PAR Did you take pictures at Matti’s birthday party.

04 K: Joo mutta toi- (4.7)aa vaitta toi- (2.8) iso toi- PRT but that ? that big that Yes but that-(4.7.) ah vaitta that-(2.8) big that-

05 (3.0)kaksikymmentä (.) viisi (1.5) kuvaa. twenty five picture-PAR

(3.0)twenty (.) five (1.5) pictures.

06 E: Jäi ottamatta. leave-PST take-INF Were not taken.

07 K: Ei (1.7) mutta toi, no but that No (1.7) but that,

08 (8.4)

CLINICAL LINGUISTICS & PHONETICS 777

E offers a completion to K’s telegraphic construction (line 06). In this way she co-constructs K’s utterance into a full sentence: ‘Yes but twenty five pictures were not taken’. K, however, rejects her interpretation (line 07), and the negotiation continues.

In sum, the expressive linguistic problems of the PWA with non-fluent aphasia were more extensive affecting most of his speaking turns. As a consequence, the participation structure of the negotiation sequences in these conversations manifested as follows:

(1) The non-aphasic speaker presents a question or new topic and a response is expected; the turns are speech eliciting in nature.

(2) The speaker with non-fluent aphasia produces a response which results in trouble because of sparse expression.

(3) The non-aphasic speaker co-constructs the utterance supplying a fuller grammatical structure; the turns are interpretative in nature.

(4) The negotiation continues until mutual understanding is satisfactorily reached or attempts thereof are renounced.

In the following, we will illustrate further these different participatory structures within the conversations of fluent vs. non-fluent participants. We will focus on the methods employed by the recipients to address the trouble emerging from expressive linguistic difficulties. First, we will analyse and compare the family members’ and SLT’s next turn repair actions in the conversa- tions of the participant with fluent aphasia. Second, we will look at the family member’s and SLT’s actions in the conversations of the participant with non-fluent aphasia.

Recipient actions in conversations of the participant with fluent aphasia

As shown earlier, the trouble sources of the fluent speaker proved to be more local. As a consequence, they were solved by relatively uncomplicated next turn repair actions. Example (5) illustrates a paraphasic word search, the most common expressive problem in the con- versation between the fluent speaker L and his two grandsons (A and T) (see lines 01–04). T responds to the search with a word candidate (line 05; for a similar practice see e.g. Oelschlaeger & Damico, 2000):

(5) Offering a word candidate.

01 L: Nii. (2.4) .mt joo hh .mt (3.2).mt hh (on hh nä hh) right yeah is ? Right. (2.4) .mt yeah hh .mt (3.2).mt hh (has hh nä hh)

02 (1.5) .mt (1.8) onko se- tuo- (.) hh .mt Ville is-Qhethat 1nameM (1.5) .mt (1.8) has he- that- (.) hh .mt Ville

03 kertonu minkälaista se on (0.5) amer- amerissa- (-) tell-PPC what.sort.of-PAR it is {army}-INE told you what it is like (0.5) in the a- ar- (-)

04 amer[(---) {armeija=army} ar [(---)

778 M. LAAKSO AND S. GODT

In example (5), the fluent speaker L asks a question about a mutual relative (lines 01–04). On lines 03–04 he runs into word finding difficulty in completing his turn and cannot reach the correct form of the word. However, at the end of a grammatical construction his word search is suggestive enough for T, one of the grandsons, to come up with a word candidate, and so on line 05, T joins the search in overlap by offering the word to complete L’s turn. The close relationship between the participants and the family mem- ber’s shared knowledge on the topic may also make it easier to offer a word candidate. The word candidate is approved by L who repeats it (line 06). This example works to exemplify how a linguistic difficulty, a word search, can quite directly and immediately be fixed in order to move forward in the conversation much like what would happen in an ordinary conversation between non-aphasic speakers (cf. Goodwin & Goodwin, 1986).

Besides word candidates, direct other-corrections were common in fluent conversations (see example 6). Here L and his two grandsons (T and A) are talking about the boys’ prospective matriculation examinations. An unmodulated other-correction typical of everyday interaction (cf. Haakana & Kurhila, 2009) occurs on line 03:

In commenting how the boys must prepare themselves for the exams, L unsuccessfully approximates the word prepata ‘prime’ with a cut-off attempt preta- and a further form peretal (line 01). L is already beginning to move on in making his comment (line 02) when A, the other grandson, corrects (line 03) the paraphasic form in overlap with L who is finishing his turn. Although A’s turn is a direct other-correction, it is not treated as

05→ T: [Armeijassa.= army-INE

[In the army.=

06 L: =Armeijassa. army-INE =In the army.

(6) Other-correcting.

01 L: Teidän pitää- (1.1) nyt sitte- (.) preta- peretal you-PL-GEN must now then {prime} {prime} So you need to-(1.1) now- then- (.) pri- pirem

02 eng[lang- English for Engl-

03 → A: [Prepata. prime-INF

[prime.

04 L: Nii. PRT Right.

05 A: Joo. PRT Yes.

06 T: Ni:i. PRT Right.

CLINICAL LINGUISTICS & PHONETICS 779

problematic by L who confirms it (line 04). The other-correction makes up a minimal insertion sequence (lines 03–04), after which the conversation returns to its progression and L’s comment on lines 01–02 receives its response from both boys (lines 05-06).

The above-mentioned direct problem-handling methods (i.e. word candidates and other-corrections) were characteristic next turn repairs in the conversation between L with fluent aphasia and his grandsons. These frequent other-repairs did not appear to be considered problematic although there is a preference for self-repair in conversation (Schegloff et al., 1977). On the contrary, direct other-repair resolved the trouble effectively and permitted the participants to develop the topic instead of focusing on repair.

However, the participation of the SLT differed from the family members to some extent. In example (7) L and his SLT are talking about the funny name of L’s residential area. Instead of directly correcting, the SLT offers a display of her understanding as a full sentence (lines 05–06):

On lines 01‒03, L comments that if he were better able to talk, he would do something regarding the name. L’s turn includes pauses and attempts at self-repair and the end of the turn (line 03) is hard to understand. The SLT does not readily respond (line 04), after which she first acknowledges L’s turn and then offers her formulation (lines 05–06) of the gist of L’s meaning for L to confirm or reject, resembling rephrasing formulations used in psychotherapy (cf. Weiste & Peräkylä, 2013) which offer the therapist’s version of the client’s description. The SLT’s formulation is produced as a summarising sentence, not as a single word as were the word candidates of the grandsons. On line 07, L indicates that the SLT’s formulation is correct by approving it. In contrast to the direct methods of the two grandsons, the SLT’s approach seems more indirect and subtle. She does not interrupt the turn of L by offering immediate corrections (instead she responds after a pause). Giving time and not instantly assuming the floor may reflect the rehabilitative elements of speech-language therapy. Not taking the floor at the first possible place seems to

(7) Offering candidate understanding.

01 L: °juu,° (1.3) jos minä olisin paremmin (0.6) puhu- yeah if I are-1-CON good-COMP speak °yeah,° (1.3) if I was better (0.6) at speak-

02 puhumaan niin (.) menisin- menisin sinne joskus (.) speak-INF so go-1-CON go-1-CON there sometime speaking (.)I would go- I would go there sometime (.)

03 heh heh (1.2) vähä (käy käy-) (.) Pönttölä. a little fit fit place-name

heh heh (1.2) a little (fit fit-) (.) Pönttölä.

04 (1.3)

05 → SLT-L: Mmm. (.)pyrkisit vaikuttammaan että try-2-CON influence-INF that

Mmm. (.)you would try to influence on

06 → muutetaan nimi. change-PAS DEF name changing the name.

07 L: Niin? (0.8) PRT Yeah? (0.8)

780 M. LAAKSO AND S. GODT

encourage the speaker with aphasia to resolve the trouble on his own. In this way she allows the preferred self-repair to take place. In therapy other-correction may be experi- enced as dispreferred. The SLT’s multi-word response also models sentence-level expres- sion for the aphasic speaker.

Recipient actions in conversations of the participant with non-fluent aphasia

In contrast to the easily identified, isolated trouble sources in fluent conduction aphasia and the immediate next turn resolution thereof, in non-fluent aphasia, trouble permeates the conversation, leaving the non-aphasic speaker responsible for co-constructing a mean- ing for the turns. In example (8), K with non-fluent aphasia is talking with his wife (E) about their plans for the day in a form of a ‘hint-and-guess’ sequence where questions, interpretations and guesses by the non-aphasic participant alternate with the answers and hints by the PWA (cf. Laakso & Klippi, 1999). The wife starts the topic by informing K what she is going to do in the evening (lines 01–02):

(8) Co-constructing ‘hint-and-guess’ sequence.

01 E: No kuule (1.2) mä lähden nyt sitte tänä iltana PRT hear-2-IMP I go-1 now then this-ESS evening-ESS Well you (1.2) I’m going to go draw in pencil

02 piirtämään. draw-INF tonight.

03 K: Jaaha? PRT I see?

04 E: Mmmhm?

05 K: Minä toi khhh kauppa toi (.) leh- toi lehti? I that store that {magazine} that magazine I that khhh store that (.) mag- that magazine?

06 E: Nii? So?

07 K: Ja ((coughs)) kir- kirje (.) kirja.

and {letter} letter book And ((coughs)) let- letter (.) book.

08 → E: (Oo sun) mitä Tekniikan maailman vai? you-GEN what magazine.name-GEN or

(Oh your) World of Technology or which one?

09 K: Joo. Yes. ((K LOOKS AWAY FROM E))

10 → E: Vaiko Tietokonelehti. Or-Q magazine.name Or Computer Magazine.

11 K: £Joo se toih-£ yeah it that

£Yes this that-£

CLINICAL LINGUISTICS & PHONETICS 781

The new topic the wife introduces is met with K’s short response (line 03), and the wife encourages K to expand with a minimal interrogative Mmmhm? (line 04). As a result, K produces a telegraphic-style utterance ‘I – store – magazine’ (line 05). In this slot, his utterance is interpretable as commenting what he himself is going to do later. Again the wife invites K to continue with an inquiring minimal response (line 06). K continues by adding elements ‘and letter book’ (line 07) which do not clarify the meaning of his prior turn. The wife proceeds to offer a name of a magazine, and a ‘hint and guess’ sequence begins (line 08). The wife’s interrogative guess-type turns alternate with K’s hint-type answers until line 16 where K utters a crucial hint ‘America’ which is followed by the wife’s final candidate, Computer Magazine (line 17). The wife’s single word offers resemble the word candidates the grandsons used to resolve word searches in the fluent data. The sequence then ends with acknowledging closing turns (lines 18–19). There is an interesting detail on lines 08–09 where K seems to accept the wife’s first word candidate as correct. On closer inspection K’s turn on line 09 is hesitant in tone and he does not direct his gaze to his wife to confirm his verbal message. In the light of this conflicting information the wife, perhaps having adapted to the aphasic difficulties manifested in their conversations, is able to see that the current word candidate may not have been correct and proceeds to offer alternatives that seem to be based on her prior knowledge on magazines her husband usually buys. Another potential display of the couple’s adaptation to the husband’s disability is how the wife chooses to begin negotiating the meaning of K’s turns on lines 05 and 07. She ignores the letter and book K has mentioned but instead targets the name of the magazine which she treats as the most crucial bit of information regarding the topic. Her action resembles the behaviour of the two grandsons in the fluent speaker data in that it seems to aim at as effortless solution as possible in order to further the topic of conversation. Interpreting K’s turns in their entirety might have led into a prolongation of the negotiation.

The dissimilarity in everyday vs. institutional recipient participation was not as explicit in the non-fluent speaker conversations as in the fluent data. In the non-fluent data, the SLT co- constructs the talk of K in a way that resembles the wife’s actions in the home conversation. However, whereas the wife initiates conversational topics by commenting, the SLT tries to elicit

12 → E: £Vai molemmat.£ £Or both.£

13 K: £Yksi.£ £One.£

14 → E: Kumpi. which Which one.

15 (1.2)

16 K: Se (.) ((coughs)) Amerikka toi, it America that It (.) ((coughs)) America that,

17 → E: Tietokonelehti. magazine name Computer Magazine.

18 K: Mmm,

19 E: Joo. (1.5) Yeah. (1.5)

782 M. LAAKSO AND S. GODT

and prompt conversation by asking questions, which is a practice typically employed by SLTs (cf. Silvast, 1991). In example (9), the non-fluent speaker K and SLT-K are talking about K’s hobbies:

(9) Speech elicitation, prompting and co-constructing a ‘hint-and-guess’ sequence.

01 SLT-K: Mt entäs sitte siitä- (.) sä har- sanoit että sä what about then it-ELA you say-2-PST that you What about that then- (.) you take- you said that you

02 harrastat sitä valokuvausta. (.) take-2.an.interest.in it-PAR photographing-PAR take an interest in photographing. (.)

03 kerro siitä. (1.8) sul on kameroita. (.) tell-IMP it-ELA you-ADE is camera-PL-PAR tell me about that. (1.8) you’ve got cameras. (.)

04 montakin. many-CLI many too.

05 K: Joo. PRT Yes.

06 (2.5)

07 SLT-K: Minkälaisia. what.kind-PL-PAR What kind.

08 (2.1)

09 K: Vanha toi- (1.9) kaksikymmentä- ei kolmekymmentä-

Old that twenty no thirty Old that- (1.9) twenty- no thirty-

10 ei: neljäkymmentä vuotta toi- (1.6) heh vanhat. no forty year-PAR that old-PL No: forty years that- (1.6) heh old.

11 → SLT-K: Oot sitäkin harrastanu are-2 it-PAR-CLI take.an.interest-PPC You have taken an interest in that

12 → nelkyt vuotta. forty year-PAR fo forty years.

13 K: Joo. PRT Yes.

14 → SLT-K: Valokuvausta [myöskin. photographing-PAR too-CLI Photographing [as well.

15 K: [Joo. (.) joo. PRT PRT [Yes. (.) yes.

16 → SLT-K: Onko sulla kameraa siltä ajalta. is-Q you-ADE camera-PAR it-ABL time-ABL Do you have a camera from that period.

17 K: Joo. PRT Yes.

CLINICAL LINGUISTICS & PHONETICS 783

In example (9) on lines 01–04, the SLT introduces a new topic, K’s hobby of photography, into the conversation. The SLT’s question ‘do you have many cameras’ is met with a minimal affirmation by K (line 05). The conversation does not progress and the SLT follows up with another question ‘what kind of cameras’ (line 07), which prompts K to a telegraphic expression ‘old – twenty – thirty – forty – years old’ (lines 09–10). From lines 11–12, the SLT interprets and co-constructs the meaning of K’s turn. As K confirms, the SLT goes on by asking further questions to which K responds (lines 13–21). On line 22, the SLT produces one final question in the topical sequence, K responds minimally and the topic is closed by weak mutual agreement (lines 24–25). In the sequence there are SLT’s turns that demonstrate speech elicitation which may serve a specific speech therapy purpose but as a whole the sequence makes up a ‘hint-and- guess’ sequence where K’s hint/answer turns follow the SLT’s guess/question turns much like what could be seen in the conversation between K and his wife. Similarly as with fluent data, the SLT’s questions are formulated as full sentences, not as single word candidates. However, the SLT’s lack of shared background knowledge may affect her co-constructive turns.

Summary of the results

Due to the differences in linguistic deficits the overall structures of the fluent and non-fluent aphasic conversations turned out to be distinct. This difference was reflected both in the way aphasic expressive trouble appeared in these conversations and how it was managed. The most significant difference in the participation of the interlocutors proved to be that specific repair phenomena were present in the fluent conversations while the non-fluent conversations were co- constructive in nature. In the fluent conversations the non-aphasic participants were able to address specific, less obtrusive trouble sources, most typically word searches and phonemic paraphasias, quite directly and most of the time successfully with little interruption to the flow of

18 → SLT-K: Se- (.) nelkyt vuotta vanha. it forty year-PAR old That- (is)(.) forty years old.

19 K: Joo. (.) lapset ja toi- (.) ja toi- (.) joo.

PRT child-PL and that and that yeah Yes. (.)children and that- (.)and that-(.) yes.

20 → SLT-K: Et sä et itse käytä enää sitä. that you NEG-2 self use-PAS anymore it-PAR So you don’t use it yourself anymore.

21 K: Ei. (.) ei se toi- no no it that No. (.) no it that-

22 → SLT-K: Mut onks se kuitenkin käyttökelpoinen periaatteessa. but is-Q it anyway usable principle-INE But it still works in principle.

23 K: Joo? PRT Yes?

24 SLT-K: [Mmm?

25 K: [°Joo.° PRT

[°Yes.°

784 M. LAAKSO AND S. GODT

the conversation. In the non-fluent conversations the non-aphasic participants continuously interpreted and co-constructed the turns of the participant with non-fluent aphasia. The primary action in the non-fluent conversations thus became the turn-by-turn co-construction of mean- ing in an attempt to establish some level of mutual understanding.

In the data of the fluent aphasic speaker, the SLT and the grandsons used differing practices in handling the emerging trouble. The most direct single-word methods of repair (such as other- corrections and single word candidates) were used more by family members than by the SLT. In contrast, the practices used by the SLT were less direct and in the form of complete sentences (such as candidate understandings of the prior turn and summaries of a longer sequence) instead of single-word turns. In the non-fluent data both the wife and the SLT worked towards shared understanding by co-constructing ‘hint-and-guess’ sequences. The co-construction of meaning seemed to vary little with regard to the interlocutor. However, with the SLT speech elicitation activities seemed more prominent and the SLT formed her guesses and questions as complete sentences.

In sum, the results of the present study suggest two kinds of conclusions. First, the type of aphasia seems to have a significant impact on the emerging participation structure of the conversation and on the practices the recipients employ to manage the problems (i.e. repair vs. co-construction). Second, the participation of the family members and the SLTs seems to be dissimilar in certain ways.

Discussion

It has been noted in previous research that the deteriorated linguistic competence of speakers with aphasia renders aphasic conversation asymmetric which in turn gives rise to specific interactional phenomena. Such typical features include the emphasised role of non-aphasic speakers in the management of trouble in communication (see e.g. Bloch & Beeke, 2008; Heeschen & Schegloff, 1999; Oelschlaeger & Damico, 2000; Milroy & Perkins, 1992) and the prolongation of the repair sequences that appear in order to solve such trouble (see e.g. Laakso & Klippi, 1999; Lindsay & Wilkinson, 1999; Laakso, 2003). The results of this multiple-case study support the findings of the emphasised role of the non-aphasic interlocutors. However, our comparative approach revealed some differences in recipient participation depending on the type of aphasia: the principal participatory approach of handling trouble in the fluent speaker data was repair while in the non-fluent data it was co-construction.

Contrary to previous studies of aphasic conversation (e.g. Laakso & Klippi, 1999), prolonged sequential repair was not found in connection with chronic fluent conduction aphasia. In particular in the conversations with family members, the expressive aphasic problems were handled immediately and rather effortlessly in the next turn following the problem. On the other hand, the prolongation of the negotiation sequences and the role of the non-aphasic speaker were especially prominent in the non-fluent data. Thus, the findings of this study concur with previous studies that have investigated severe (e.g. Goodwin, 1995) and non-fluent agrammatic (e.g. Heeschen & Schegloff, 1999) aphasias. In the non-fluent data, initiating conversation in the first place required the active participation of the non-aphasic speaker. However, with the support of the interlocutor the non-fluent speaker was able to contribute to the conversation at hand. The non-aphasic participants viewed the speaker with non-fluent aphasia as a competent conversationalist and for the most part truly strived to interpret, suggest and create meaning for his turns. The same has been observed by Goodwin (1995). The way

CLINICAL LINGUISTICS & PHONETICS 785

Heeschen & Schegloff (1999) describe the aphasic telegraphic style mobilizing the non-aphasic speaker to act as an active collaborator and interpreter in re-forming the aphasic telegraphic utterances was also evidenced by the current study.

The current data may reinforce the conclusions that the nature of participation between family members and SLTs is different (see e.g. Lindsay & Wilkinson, 1999; Perkins, 1995). The present observations are analogous also with Laakso (2015) in that—just like in everyday conversations in general—the essential aspiration in the family conversation is to promote the topical flow of the conversation by managing the trouble as efficiently as possible. Instead the participation of the SLTs seems less efficient. The SLTs’ lack of shared knowledge on the PWAs’ life as compared with the relatives may also influence the accuracy of their interpretive actions. The SLTs’ indirect sentence-level practices resemble those found by Lindsay and Wilkinson (1999). In fluent speaker data, the family members use more direct methods than the SLT to deal with the trouble, much like the prompt word candidates described by Oelschlaeger & Damico (2000). Also noteworthy is that direct other-correction is not dispreferred, contrary to the prior findings of Perkins (1995) and Wilkinson (1995) on English aphasic conversation.

Limitations, further research, and clinical implications

The results suggest that there may be different conversational participatory structures where the type of aphasia may be one of the dividing factors. It is nonetheless best assumed that the variation is substantial even within one aphasia type which is why a dichotomy merely based on aphasia fluency should probably be considered cautiously (cf. Helasvuo, Klippi, & Laakso, 2001). The differences in the chronicity and severity of aphasia in our data may play a role and deserve further study. In our study, due to the nature of his conduction aphasia, the fluent speaker was less severely impaired in his expression than the speaker with non- fluent agrammatic aphasia. If possible, future studies could examine conversations involving participants with similarly severe but qualitatively different aphasias. Our findings also suggest that further research directed at everyday conversations along with the institutional settings is essential. Varying settings and co-participants may lead to different displays of the communicative competence of the PWA. Thus, examining only one setting may give a limited view on the possibilities for participation the PWA has.

The findings imply that clinical professionals should increasingly use videotaped everyday interactions as well as therapy data in evaluating the language and communication of people with aphasia. Systematic study of recipient-initiated repair actions in naturally occurring interaction gives research-based knowledge that can be used in training conversation partners, which has become an increasingly common approach in aphasia rehabilitation (for a review of such training approaches, see e.g. Simmons-Mackie, Savage, & Worrall, 2014; Turner & Whitworth, 2006; Wilkinson, 2010). Future studies should explore the use of next turn repair practices in clinical partner training. The differences in repair practices could be taken into consideration in partner training: one might want to train more immediate repair practices in connection with fluent conduction aphasias, whereas in connection with non-fluent agrammatic aphasia the approach might be more on co-con- struction, elicitation and assisting augmentative and alternate communication methods,

786 M. LAAKSO AND S. GODT

such as pictorial support. With detailed case studies we can accumulatively form a larger, systematic database within which training outcomes can be observed.

Declaration of interest

The authors declare no conflicts of interest.

References

Aaltonen, T. & Laakso, M. (2010). Halting aphasic interaction: Creation of intersubjectivity and spousal relationship in situ. Communication & Medicine, 7, 115–126.

Atkinson, J. M. & Heritage J (Eds.). (1984). Structures of social action: Studies in conversation analysis. Cambridge: Cambridge University Press.

Barnes, S. (2016). Aphasia and open format other-initiation of repair: Solving complex trouble in conversation. Research on Language and Social Interaction, 49, 111–127.

Barnes, S., & Ferguson, A. (2015). Conversation partner responses to problematic talk produced by people with aphasia: Some alternatives to initiating, completing, or pursuing repair. Aphasiology, 29, 315–336.

Bartha, L., & Benker, T. (2003). Acute conduction aphasia: An analysis of 20 cases. Brain and Language, 85, 93–108.

Beeke, S., Beckley, F., Best, W., Johnson, F., Edwards, S., & Maxim, J. (2013). Extended turn construction and test question sequences in the conversations of three speakers with agrammatic aphasia. Clinical Linguistics and Phonetics, 27(10–11), 784–804.

Beeke, S., Wilkinson, R., & Maxim, J. (2003). Exploring aphasic grammar 1: A single case analysis of conversation. Clinical Linguistics and Phonetics, 17, 81–107.

Beeke, S., Wilkinson, R., & Maxim, J. (2007). Grammar without sentence structure: A conversation analytic investigation of agrammatism. Aphasiology, 21, 256–282.

Bloch, S., & Beeke, S. (2008). Co-constructed talk in the conversations of people with dysarthria and aphasia. Clinical Linguistics and Phonetics, 22, 974–990.

Damico, J. S., Wilson, B. T., Simmons-Mackie, N. N., & Tetnowski, J. A. (2008). Overcoming unintelligibility in aphasia: The impact of non-verbal interactive strategies. Clinical Linguistics and Phonetics, 22, 775–782.

Goodglass, H., & Kaplan, E. (1983). The assessment of aphasia and related disorders. Philadelphia: Lea and Febiger. Finnish version Laine, Niemi, Tuomainen and Koivuselkä-Sallinen 1997.

Goodwin, C. (1995). Co-constructing meaning in conversations with an aphasic man. Research on Language and Social Interaction, 28, 233–260.

Goodwin, C. (2003). Conversational Frameworks for the accomplishment of meaning in aphasia. In C. Goodwin (Ed.), Conversation and brain damage, (pp. 90–116). Oxford, UK: Oxford University Press.

Goodwin, M. H., & Goodwin, C. (1986). Gesture and co-participation in the activity of searching for a word. Semiotica, 62, 51–75.

Haakana, M. & Kurhila, S. (2009). Other-correction in everyday conversation: Some comparative aspects. In M. Haakana, M. Laakso, & J. Lindström (Eds.) Talk in interaction: comparative perspectives (pp. 152–179). Helsinki, Finland: Finnish Literature Society.

Heeschen, C. & Schegloff, E. A. (1999). Agrammatism, adaptation theory, conversation analysis: On the role of so-called telegraphic style in talk-in-interaction. Aphasiology, 13, 365–405.

Helasvuo, M.-L., Laakso, M., & Sorjonen, M.-L. (2004). Searching for words: Syntactic and sequen- tial construction of word search in conversations of Finnish speakers with aphasia. Research on Language and Social Interaction, 37, 1–37.

Helasvuo, M-L., Klippi, A., & Laakso, M. (2001). Grammatical structuring in Broca’s and Wernicke’s aphasia in Finnish. Journal of Neurolinguistics, 14, 231–254.

Heritage, J., & Maynard, D. (Eds.). (2006). Communication in Medicine: Interaction between primary care physicians and patients. Cambridge, UK: Cambridge University Press.

CLINICAL LINGUISTICS & PHONETICS 787

Kaplan, E., Goodglass, H., & Weintraub, S. (1983). Boston naming test. Philadelphia, PA: Lea and Febiger. Finnish version Laine, Koivuselkä-Sallinen, Hänninen and Niemi 1997.

Klippi, A. (2015). Pointing as embodied practice in aphasic interaction. Aphasiology, 29, 337–354. Kohn, S. E. (1984). The nature of the phonological disorder in conduction aphasia. Brain and

Language, 23, 97–115. Laakso, M. (1997). Self-initiated repair by fluent aphasic speakers in conversation. Studia Fennica

Linguistica 8. Helsinki, Finland: Finnish Literature Society. Laakso, M. (2003). Collaborative construction of repair in aphasic conversation: An interactive view

on the extended speaking turns of persons with Wernicke’s aphasia. In C. Goodwin (Ed.), Conversation and brain damage (pp. 163–188). New York, NY: Oxford University Press.

Laakso, M. (2014). Aphasia sufferers displays of affect in conversation. Research on Language and Social Interaction, 47, 404–425.

Laakso, M. (2015). Collaborative participation in aphasic word searching: Comparison between significant others and speech and language therapists. Aphasiology, 29(3), 269–290.

Laakso, M., & Klippi, A. (1999). A closer look at ‘hint and guess’ sequences in aphasic conversation. Aphasiology, 13, 345–363.

Lindsay, J., & Wilkinson, R. (1999). Repair sequences in aphasic talk: A comparison of aphasic speech and language therapist and aphasic-spouse conversations. Aphasiology, 13, 305–325.

Menn, L., O’Connor, M., Obler, L. K., & Holland, A. (1995). Non-fluent aphasia in a multilingual world. Amsterdam, The Netherlands: John Benjamins.

Milroy, L., & Perkins, L. (1992). Repair strategies in aphasic discourse; towards a collaborative model. Clinical Linguistics and Phonetics, 6, 27–40.

Norrick, N. (1991). On the organization of corrective exchange in conversation. Journal of Pragmatics,16, 59–83.

Oelschlaeger, M. L., & Damico J. S. (2000). Partnership in conversation: A study of word search strategies. Journal of Communication Disorders, 33, 205–225.

Perkins, L. (1995). Applying conversation analysis to aphasia: Clinical implications and analytic issues. European Journal of Disorders of Communication, 30, 372–838.

Sacks, H. (1992). Lectures on conversation I and II (ed. G. Jefferson). Oxford, UK: Basil Blackwell. Schegloff, E. A., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in organization

of repair in conversation. Language, 53, 361–382. Schegloff, E. A. (2007). Sequence organization in interaction. Cambridge, UK: Cambridge University

Press. Silvast, M. (1991) Aphasia therapy dialogues. Aphasiology, 5, 383–390. Simmons-Mackie, N., Savage, M. C., & Worrall, L. (2014). Conversation therapy for aphasia: A

qualitative review of the literature. International Journal of Language and Communication Disorders, 49, 511–526.

Sorjonen, M.-L. (2001). Responding in conversation. A study of response particles in Finnish. Amsterdam, The Netherlands: John Benjamins.

Turner, S., & Whitworth, A. (2006). Conversational partner training programmes in aphasia: A review of key themes and participants’ roles. Aphasiology, 20, 483–510.

Weiste, E., & Peräkylä, A. (2013). A comparative conversation analytic study of formulations in psychoanalysis and cognitive psychotherapy. Research on Language and Social Interaction, 46, 299–321.

Wilkinson, R. (2010). Interaction-focused intervention: A conversation analytic approach to aphasia therapy. Journal of Interactional Research in Communication Disorders, 1, 45–68.

Wilkinson, R. (1995). Aphasia: Conversation analysis of a non-fluent aphasic person. In M. Perkins & S. Howard (Eds.), Case studies in clinical linguistics (pp. 271–292). London, UK: Whurr Publishers.

788 M. LAAKSO AND S. GODT

Appendix

Principles and abbreviations used in glossing (modified from Sorjonen, 2001).

In the gloss, morphemes have been separated from the root with a hyphen.

The case endings are referred to by the following abbreviations:

Other abbreviations:

Case Abbreviation Approximate meaning

Genitive GEN possession Partitive PAR partitiveness Inessive INE ‘in’ Elative ELA ‘out of’ Adessive ADE ‘at, on’ Ablative ABL ‘from’

Abbreviation Meaning

1 First person ending 2 Second person ending 1NAME First name CLI Clitic CON Conditional COMP Comparative INF Infinitive PL Plural PRT Particle PST Past tense (imperfect) Q Interrogative clitic

CLINICAL LINGUISTICS & PHONETICS 789

Copyright of Clinical Linguistics & Phonetics is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.