15
clinical linguistics & phonetics, 2002 , vol. 16, no. 5, 345± 359 Transcribing discourse: interactions with Alzheimer’s disease NICOLE MU ¨ LLER† and JACQUELINE A. GUENDOUZI‡ †University of Louisiana at Lafayette ‡Southeastern Louisiana University, Hammond ( Received 5 January 2002; accepted 28 February 2002 ) Abstract This paper illustrates the use of a ‘discourse line’ in transcribing spoken interaction between a person with Alzheimer’s disease, and a visitor. Discourse is here interpreted as a metacategory, or an analytic level of interaction. We view tran- scribing as an integral part of ‘doing discourse’, and use two sub-layers of the discourse line, dedicated to speech acts and conversation analysis, respectively. The prosody and voice layer is used to show the analysis of a speaker’s use of a speci c voice quality in discourse terms. Keywords : transcribing, discourse, speech acts, conversation analysis, Alzheimer’s disease. Discourse as metacategory Qualitative research in discourse and interactional patterns has become a prominent feature of clinical linguistics and phonetics (CLP) in recent years. Particularly the study of neurogenic communication disorders has experienced a surge of research using methods originating in for example sociology, pragmatics, anthropology, conversation analysis and functional linguistics (e.g., Aphasiology , 13(4/5), 1999; Damico, Simmons-Mackie, Oelschlaeger, Elman and Armstrong 1999; Damico, Oelschlaeger and Simmons-Mackie 1999; Simmons-Mackie and Damico 1999). A focus on interactional, or discourse, data has brought with it a shift in perspective, away from communicative disorder as solely situated in the limitations of a person’s communicative impairments, towards a wider focus on communication as a joint, collaborative, contextualized achievement. There is now an established, though comparatively young, research tradition using qualitative methods in neuro- genic disorders, and the discussion has become truly multidisciplinary. While this Address correspondence to: Dr Nicole Mu ¨ller, Department of Communicative Disorders, University of Louisiana at Lafayette, PO Box 43170, Lafayette, LA 70405-3170, USA. Tel: 1 1 337 482 6870. Fax: 1 1 337 482 6195. E-mail: [email protected] Clinical Linguistics & Phonetics ISSN 0269-9206 print/ISSN 1464-5076 online © 2002 Taylor & Francis Ltd http://www.tandf.co.uk/journals DOI: 10.1080/02699200210135875 Clin Linguist Phon Downloaded from informahealthcare.com by University of North Texas on 11/11/14 For personal use only.

Transcribing discourse: interactions with Alzheimer's disease

Embed Size (px)

Citation preview

Page 1: Transcribing discourse: interactions with Alzheimer's disease

clinical linguistics & phonetics, 2002, vol. 16, no. 5, 345± 359

Transcribing discourse: interactions withAlzheimer’s disease

NICOLE MULLER† andJACQUELINE A. GUENDOUZI‡

†University of Louisiana at Lafayette‡Southeastern Louisiana University, Hammond

(Received 5 January 2002; accepted 28 February 2002 )

Abstract

This paper illustrates the use of a ‘discourse line’ in transcribing spoken interactionbetween a person with Alzheimer’s disease, and a visitor. Discourse is hereinterpreted as a metacategory, or an analytic level of interaction. We view tran-scribing as an integral part of ‘doing discourse’, and use two sub-layers of thediscourse line, dedicated to speech acts and conversation analysis, respectively.The prosody and voice layer is used to show the analysis of a speaker’s use of aspeci� c voice quality in discourse terms.

Keywords: transcribing, discourse, speech acts, conversation analysis,Alzheimer’s disease.

Discourse as metacategory

Qualitative research in discourse and interactional patterns has become a prominentfeature of clinical linguistics and phonetics (CLP) in recent years. Particularly thestudy of neurogenic communication disorders has experienced a surge of researchusing methods originating in for example sociology, pragmatics, anthropology,conversation analysis and functional linguistics (e.g., Aphasiology, 13(4/5), 1999;Damico, Simmons-Mackie, Oelschlaeger, Elman and Armstrong 1999; Damico,Oelschlaeger and Simmons-Mackie 1999; Simmons-Mackie and Damico 1999). Afocus on interactional, or discourse, data has brought with it a shift in perspective,away from communicative disorder as solely situated in the limitations of a person’scommunicative impairments, towards a wider focus on communication as a joint,collaborative, contextualized achievement. There is now an established, thoughcomparatively young, research tradition using qualitative methods in neuro-genic disorders, and the discussion has become truly multidisciplinary. While this

Address correspondence to: Dr Nicole Muller, Department of Communicative Disorders,University of Louisiana at Lafayette, PO Box 43170, Lafayette, LA 70405-3170, USA.Tel: 1 1 337 482 6870. Fax: 1 1 337 482 6195. E-mail: [email protected]

Clinical Linguistics & PhoneticsISSN 0269-9206 print/ISSN 1464-507 6 online © 2002 Taylor & Francis Ltd

http://www.tandf.co.uk/journalsDOI: 10.1080/02699200210135875

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 2: Transcribing discourse: interactions with Alzheimer's disease

N. Muller and J. A. Guendouzi346

development is a productive one, it is also potentially confusing: The availability ofmultiple methods opens up new perspectives on the complexities of human inter-action. However, each transfer of a theory or analytic methodology brings with itits own set of challenges, some methodological and analytic, some representational(see also Perkins (2000) for further discussion).

Some basic assumptions

In our approach to the analysis of discourse in Senile Dementia of the Alzheimer’sType (SDAT), and to transcribing as an integral part of the analysis, we are makingsome basic assumptions: (1) Communicative ‘de� cit’ or ‘disorder’, for the presentpurposes, is seen as situated in the communicative event, rather than the individualaZicted with SDAT. This de� cit needs to be operationalized in terms of real com-municative events (rather than data elicited in testing situations), more speci� callyin free conversations, as the prototypical communicative event. (2) In order tounderstand communicative events, and any potential de� cits residing therein, weneed to describe patterns. This requires adequate descriptive terminology, and cate-gorization of the events described. (3) While descriptive frameworks can be foundin various approaches to human interaction and its underlying principles, we needto keep in mind that any approach both limits and focuses the perspective taken onthe object of analysis. (4) Since available analytic approaches to human interactionand language use are implicitly situated in ‘normal’ interaction and ‘normal’ mentalfaculties,1 the context of communicative disorders may require an adaptation, ormaybe even a partial rede� nition of the categories employed. (5) Transcription anddescription are integral parts of analysis and interpretation. As discussed in the leadarticle to this issue, we view transcription as a complex process that is necessarilyand deliberately selective, though � exible and cyclical. Transcribing provides a focusand a basis for the description of interactive patterns, familiarizes the analyst withthe data, and becomes one of the vehicles through which analysis is transmitted tothe reader.

This approach to ‘analysis through transcribing’ also led us to rede� ne ourconcept of discourse as an analytic category: For the purposes of our work, we view‘discourse’ as a metacategory, or a conglomeration of descriptive subsystems whichcontribute to an analytic picture of an interaction. Put more simply, one could saythat in our de� nition, people talk and interact, analysts do discourse.2 A multi-layered transcript thus serves as an attempt at anchoring the analyst’s process ofdiscovery and categorization in a graphic mode.

Analysis through transcribing

The data

The data we are analysing is a tape-recording of a free conversation between twoparticipants:3 F is a female in her early 80s, diagnosed with SDAT several yearsprior to the conversation. She has also experienced non-speci� ed stroke damage inthe past, and has a signi� cant, long-standing hearing impairment. R, the researcher,is a female in her early 40s who visits F on a weekly basis at the residential carefacility where F has been a resident for several years. The relationship that R hasto F has developed over an 8 month period from that of a complete newcomer to

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 3: Transcribing discourse: interactions with Alzheimer's disease

T ranscribing discourse 347

F’s limited social circle (consisting largely of her husband, staV at the care facility,other residents, and very occasional visitors from her family), to that of a regular,and expected visitor. The conversation was recorded using a Sony WalkmanProfessional with a free-standing microphone, in F’s bedroom at the care facility.

Transcription and presentation

The transcript presented here is a collaborative eVort between the two authors ofthis paper. We chose to arrange the information captured in transcribing in threelayers: the ‘base’ line, or orthographic layer, marked ‘O’, the prosody and voicelayer, marked ‘P’, and a discourse layer, marked ‘D’. A full key to the transcriptionconventions employed can be found towards the end of this paper; not all symbolsused are discussed here in full. In order to keep the length of the paper withinreasonable limits, we are including extracts from the conversation here, rather thanthe full, multilayered transcript in its entirety.4

Prosody and voice

Voice quality: interactions with conversation managementThe prosody and voice layer (see also Ball, 2002) is here used to indicate markeddiVerences in speech rate and volume relative to the surrounding speech, as in lines19, 32, 43 or 57, as well as voice quality. Basic prosodic markers are integrated intothe orthographic layer, as is customary in discourse studies (see transcription key).Throughout, F’s voice quality has a distinct creaky quality, which is not marked inthe transcript, as there are no prominent � uctuations in this. Speci� cally markedare whisper (‘W’; line 49), and the voice quality associated with ‘silent’ laughter(‘LV’ or ‘laughing voice’, e.g., lines 28, 64, 71). This latter voice quality is distingu-ished in the transcript from laughter per se, see e.g., lines 6, 28. We do not attemptan orthographic representation of the sounds of laughing,5 but maintain a binarydistinction between laughter, and LV. The former could typically be described as abrief laugh or chuckle. The role of LV in the conversation is more fully discussedbelow.

Orthographic transcription and intelligibility

The orthographic layer6 includes information concerning speech intelligibility. In thestudy of ‘normal’ (i.e., ‘non-disordered’) discourse, unintelligible speech tends to bea side issue, often marked in the transcript by the use of parentheses to indicateeither completely unintelligible speech (empty parentheses) , or a transcriber’s bestguess (see Damico and Simmons-Mackie, 2002; Atkinson and Heritage 1984[1999]:246; Gumperz, 1992: 248). Unintelligible stretches of speech tend not to become afocus of analysis. However, in the study of SDAT (and other communicationdisorders), speech intelligibility is an issue in itself, and therefore the classi� cationof degrees of intelligibility becomes rather more important, if not unproblematic interms of measurement. In our method of transcribing, we do not attempt to imposean objective measure of intelligibility; a � ve-fold, non-scalar classi� cation arises outof the transcription process itself: (1) Fully intelligible speech is transcribed ortho-graphically. Stretches of speech where the transcribers are not sure of their interpreta-tions are enclosed in parentheses.7 (2) Within parentheses, a best guess at words

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 4: Transcribing discourse: interactions with Alzheimer's disease

N. Muller and J. A. Guendouzi348

and phrases is transcribed orthographically , in other words, the transcribers feltcon� dent enough to attempt a semantic interpretation of the speech. (3) Where thiswas not possible, sometimes a segmental interpretation is possible; this is indicatedby the use of a broad phonetic transcription. (4) Sometimes, the transcribers cannotmake con� dent judgements as regards speech segments, but are able to count thenumber of syllables produced, which are indicated by the use of one ‘x’ for eachsyllable, and ‘X’ for a syllable produced with noticeably greater prominence thanthe surrounding ones (see Gumperz, 1992, for a similar convention) . (5) Where evena syllable count was not possible, unintelligible speech was timed. Lines 35, 40, 44and 56 in Extract 1 below illustrate the use of these conventions.

Classifying intelligibility in transcribing brings with it a paradox that is hard toresolve, namely that of intelligibility as a perceptual and interactive process duringthe conversation, versus intelligibility arising out of repeated hearings of a recording.The two obviously do not map neatly onto one another, and we have to keep inmind that the diVerent categories of intelligibility that arose out of transcribingrepresent the transcribers’, rather than the interactants’, perspective. As a participantin an interaction, one does not count syllables, nor time stretches of speech onecannot understand. However, this discrepancy between perception-in-interactionversus perception-in-analysi s is a factor in all categories imposed during transcribing.

Doing Discourse: Managing reduced intelligibility and comprehensibilityF’s talk exhibits marked � uctuations in intelligibility (and talk that is intelligible interms of words, sentence structure etc., is not necessarily comprehensible in allcases). In fact, a large part of the transcribing eVort goes into managing these� uctuations. Even though the classi� cation imposed is a construct of transcribing,rather than interacting, reduced intelligibility also places a higher burden of problemsolving on an interactant (assuming that she wishes the interaction to continue).Taken together with the position of one of the authors as participant-transcriber -analyst, and the other one as transcriber-analyst-monitor , this presents a usefulpoint of attack for analysis: are there patterns in the � uctuations we categorize?Furthermore, are there patterns in R’s contributions that relate to decreased intelli-gibility on the part of F? The analysis, with decreased intelligibility as a point ofdeparture, is incorporated in sub-strata of the discourse layer of the transcript. Wecode utterances for Speech Acts (sub-stratum: D-SA), and Conversation Analysis(D-CA). Two basic questions underlie this coding: SA analysis asks, ‘What is theinterlocutor doing with the utterance?’ CA asks, ‘What is the utterance doing at thispoint in the conversation?’ SA analysis, of course, implies that it is possible toreconstruct an interlocutor’s intention from her output, which can be tricky, evenmore so where intelligibility is impaired.

SAs are assigned to the following categories:8 Directive (D): speaker intends todirect hearer’s behaviour. Within this category, we specify a subcategory continuer(D-Cont): speaker directs the addressee to continue speaking. An expressive (E) isan expression of the speaker’s state of mind or emotional state, or attitude towardsthe world. A question is a request for information. Within the category question, wedistinguish further the subcategories question-con� rmation (Q-Cf ), a request forcon� rmation of the speaker’s understanding of a state of aVairs, and a question-cue(Q-Cu), a hybrid category, of a self-cue (i.e., a ‘jogging’ of the speaker’s ownmemory), which becomes a request for information when the self-cue fails. A

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 5: Transcribing discourse: interactions with Alzheimer's disease

T ranscribing discourse 349

representative (R) makes a statement about the world. Within all categories, repeti-tions or close paraphrases are marked as recapitulations (recap). Two dangers in SAanalysis are arbitrariness in assigning SAs to utterances (see e.g., Levinson, 1980),and a proliferation of labels and distinctions (see e.g., Ferguson, 2000), both ofwhich may obscure possible patterns in the data, the former by creating patterns ina transcript where there were none in the interaction, the latter by insisting onexcessive detail. We have attempted to keep the subcategories to a minimum. Thedistinctions drawn arose, again, out of the process of analysis, rather than beingsuperimposed a priori. This still means, of course, that what distinctions there arerepresent a construct of analysis, rather than inherent features of the interaction,but as long as it is acknowledged that our SA substratum represents the transcribers’reconstruction of the interlocutors’ act,9 it becomes a useful tool for anchoring partof the analysis, particularly in contrast with the CA substratum, which attempts toclassify the transcribers’ construction of conversational structure.

The CA substratum focuses on two areas of conversation management: the � rstis trouble and repair, i.e., indications that the conversation needs to be realigned inorder to continue, and what is done to realign it. The categories used are troubleindicating behaviour (TIB), and repair (Rep). We also code certain turn-yielding cues(TYC), i.e., indicators that a turn is coming to an end (although it may not appearto be � nished in terms of, for example, syntax or content). The second focus is topicmanagement. We code topic shifts (TS), and the subcategories related (Rel ) andretrievable (Ret) topic. The former oVers a clear content-based link to the immedi-ately preceding topic, whereas the latter is retrievable from the context of theinteraction, even though an explicit link to the preceding topic does not exist. Atopic expansion (TExp) adds to a previously introduced topic, whereas a topic uptake(TU ) may only reiterate information already given. Topic establishment (TEst) refersto an utterance which re-introduces a topic that was not taken up in the precedingutterance, following a topic shift in the utterance before. Our line-by-line discussionof the discourse layer, its substrata, and how they interact with the other layers willfocus on two extracts from the conversation.

Extract 1: Lines 19 to 62

A � rst glance at the Speech Act substratum (D-SA) of the discourse layer showsthat it is de� nitely R who ‘asks the questions’ in this conversation. R’s question inline 19 is the second repetition of a previous question, now presented with a distinctreduction in speechrate (marked lento on the prosody and voice layer). F respondswith an evaluation followed by a statement which deteriorates towards the end ( line20), and this in turn prompts a further question from R (line 21), to which Fresponds, after a pause with a minimal representative (‘no’), followed by an evalu-ation of her own situation. We interpret R’s minimal (and latched) utterance in line23 as a continuer, i.e., an attempt on R’s part to keep F talking. The question inline 19 can be seen as a repair move and attempted topic establishment, in that itpermits the realignment of the conversation by F in line twenty, a topic expansion.Line 21 expands the ‘Christmas’ topic further, as does 22, which however providesanother potential dead end for the conversation: An evaluation such as ‘we’re justleft here dead’ invites disagreement, which would imply that the speaker’s judgementis impaired, or else agreement to a less than enviable situation. Instead, the minimalturn in line 23 avoids either option, and hands the turn back to F.

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 6: Transcribing discourse: interactions with Alzheimer's disease

N. Muller and J. A. Guendouzi350

F’s attempted expressive statement ( line 24 ) turns into a question-cue, to whichR feeds back a latched, minimal continuer ( line 26), and, after an unintelligiblecontinuation on the part of F ( line 27), R provides an attempted con� rmation ( line28) of her own understanding in the shape of a partial repetition of the last intelligiblepart of F’s preceding utterance. This is followed by R’s statement about herself (‘mybrain is going’), preceded by a brief laugh and accompanied by LV (discussed inmore detail below). As regards topic management, F’s utterance in line 24 is anattempted, but unsuccessful, topic shift, followed by an attempted topic uptake in28 and a momentary shift to a related topic (‘my brain’). F’s attempted statementabout herself ( line 29), which again deteriorates as it progresses, appears to be anattempted topic establishment. The deterioration prompts the same strategy from Ras in 28, a con� rmation ( line 31), which serves again as an attempted topic uptake,preparing the way for further expansion on the part of F. We see the same strategyresurfacing in line 36, where R’s con� rmation (repeated) represents a recapitulationof the last fully intelligible part of F’s utterance, without adding further information.Thus the utterance, again, serves to establish a topic, leaving it up to F’s nextcontribution to expand further.

F’s next utterance ( line 40–42) twice shows a pattern we also � nd in otherconversations between F and R, where an attempted representative statement deteri-orates and is followed by what appears to be a formulaic expressive statement, ‘andhe’s good to me’, and ‘he’s marvellous to me’. Since the attempted representativealso introduces a new potential referent for ‘he’ (‘my eldest brother’, in addition to‘my husband’, cf. line 32), R invites F to disambiguate , using an elliptic polarquestion ( line 43), to which F responds in line 44, following on with a representativethat may very well be a reformulation of the attempt in lines 40–42. A furtherattempted representative ( line 47), prompted by R’s minimal continuer ( line 46 ),again deteriorates. As regards topic management, trouble and repair, again we seea pattern that recurs frequently in R’s interactions with F. A topic is establishedover several turns, where F’s turns supply topic expansions and (sometimesattempted) shifts to related topics. These attempts typically introduce new troublespots, whereas R’s turns facilitate topic establishment, and repair, and thus thecontinuation of the conversation, without contributing further topic developments(and thus new information that would increase the processing load for F ).

R changes strategy in line 48 and begins a series of questions in attempt to elicitfurther information about ‘the brother’, which is, however, not successful. F’swhispered ‘my brother, my brother’, following a four second pause, could be inter-preted, in SA terms, as a placeholder, indicating that she intends to comply withR’s request for information. However the following ten second pause and � veunintelligible syllables prompt R’s next question, which is again followed by aplaceholder, and a pause, which gives way to an expressive ‘oh dear’, indicating thatF is unable to comply with the request. R modi� es her questions, from openquestions to attempts to elicit yes/no responses in lines 52 and 54 (combined with acon� rmation) , which at � rst appears successful, although F’s responses, ‘no?’, withtheir rise in intonation, might equally well be interpreted as a continuer (‘keepasking’) or as an expression of uncertainty. In topic management terms, R’s questionscan be seen as attempted topic expansions on ‘the brother’, which are, however, nottaken up in F’s utterances, each of which indicates trouble in that it signals the needfor a further contribution from R for the conversation to continue. F’s utterance inline 55 appears to shift away again from ‘the brother’ (though the antecedent of‘they’ is unclear; ‘brother’ and ‘husband’? A generic ‘they’?), to herself.

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 7: Transcribing discourse: interactions with Alzheimer's disease

T ranscribing discourse 351

R’s (true or pretended) con� rmation of understanding ( line 57) is followed byanother question, displaying the strategy encountered earlier whereby a piece ofinformation contributed by F is recast into a question, i.e. a request for F toelaborate. Thus the utterance serves as an attempt to repair and take up a newtopic. The following 10 second pause shows that the strategy is of limited success.F’s response, of minimal intelligibility, con� rms the topic (F herself, ‘I get ...’), andappears to be an attempted expansion of the topic.

R’s next question ( line 59), displays a new strategy. It shifts the topic to thehere and now, to an observable phenomenon (i.e., a topic that is retrievable fromthe context) ; again we see a step-wise establishment and eventual uptake andexpansion of the topic, via a trouble indicating behaviour by F ( line 60), and anexpansion by R ( line 61). F’s question-continuer (followed by four unintelligiblesyllables) invites further speci� cation from R, in the shape of two representativestatements, and a question which in turn prompts a yes/no answer. To this Fresponds with a con� rmation of the presupposition in R’s earlier question (‘Whathave you done to your leg’—i.e., ‘you’ve done something to it’), by reusing thephrasing supplied by R (‘did you bump it’), and expanding on the information thussupplied (‘wi’ bein’ pushed’). It is arguable that R’s question ( line 61) provides vitalscaVolding that results in one of the � rst instances in our sample where F producesa completely intelligible, and syntactically complete and complex utterance.

Extract 2: Lines 148 to 166

Our second extract shows, even at � rst sight, some important diVerences to theearlier one: F’s utterances are generally more intelligible (and comprehensible) , andthey are syntactically more complex and complete. They also display what wouldappear to be a strategic use of the ‘laughing voice’ (LV). An increase in intelligibilityover the course of a conversation is a feature we have observed regularly in ourdata (and documented in our transcripts) , and although the phenomenon awaitsfurther, and more detailed, analysis, we have identi� ed several potentially contribut-ing factors. F appears to experience a ‘warm-up’ eVect, in that as she settles into aconversation, the process of conversing becomes easier for her, which manifests itselfin utterances that are more well-formed as regards both language and speech. As itcan be argued that intelligibility-in-interaction is a joint achievement, a similar warm-up eVect could be posited for R, in that her ‘tuning in’ to the conversation mightprompt utterances that are more supportive of both F’s and her own understandingof the conversation. In this particular conversation, a further factor is that F isgiven something to drink, which makes speaking easier (as well as, of course,improving her general well-being). As regards the transcriber’s perspective on intelli-gibility, a similar ‘warm-up’ eVect seems at � rst sight likely. However, such an eVectshould also be achieved by multiple listenings to the recording, which is not the case.

Our extract shows four instances of F’s use of LV that follow a common pattern( lines 153, 157, 160, 166). LV does not span the whole utterance, but only accompan-ies the � nal part, in one case ( line 157), also with a deterioration of that utterancein intelligibility. In all utterances, LV accompanies F’s indicating that informationshe wishes to impart (or maybe rather: evaluate, line 160) is unavailable to her; inother words, an admission of a de� ciency. F’s response ( line 153) to R’s request forinformation ‘do you like to go shopping’ supplies information that is slightly oVtarget: the request targets the present; F’s representative is about the past (‘I used

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 8: Transcribing discourse: interactions with Alzheimer's disease

N. Muller and J. A. Guendouzi352

to do’), followed by an admission that F is unable to fully supply the informationasked for (‘I don’t know now’), the � nal word (‘now’) of which is accompanied byLV. A similar pattern occurs in 157, where the information requested by F ( line156) is not forthcoming and F’s further contribution, following a two-second pause,explains why and again contains LV. F’s response to R’s oVering a speci� c piece ofinformation as a possible ful� lment to her own request for information in line 156,in the shape of a question-con� rmation, may express uncertainty (‘yes I like (there)’),as it is followed by another representative echoing her earlier explanation (‘but it’sso so long ago’), and a further explanation which in turn rephrases her explanationin line 153, again with the last word accompanied by LV. In line 166, F’s representat-ive (‘I don’t know where it is’), signalling that she does not have all the informationavailable in order to follow R’s preceding utterances, again shows LV accompanyingthe end of the utterance.

In terms of speech acts and their eVect on an interlocutor, LV may be used as amitigating strategy, an implicit apology. Is this is the case, LV indicates that F isaware that she is not ful� lling the expectations set up by R’s speech acts, and isattempting to compensate for this. In terms of politeness, one could therefore speakof a face-saving strategy, both as regards F’s own and R’s face in the collaborativeenterprise of a conversation. The fact that LV accompanies the end of an utterancealso invites its interpretation in terms of conversation management, namely as acombined trouble indicating behaviour and turn yielding cue. The trouble in questionis not so much occasioned by a lack of the current speaker’s understanding of wherethe conversation is going, but rather by the need for the other speaker to take overagain in order for the conversation to continue.

Questions and observations

After this presentation of some of our data, some important questions arise withregard to the role of transcribing in the actual analysis.

What are the actual, tangible bene� ts, in terms of analysis andpresentation, of a multi-layered approach to transcribing discourse?

From the point of view of the transcriber-analyst , the most important (though time-consuming) , bene� t of this approach is that it makes a cyclical, interactive approachto the data necessary, in which a recording is revisited several times from diVerentperspectives. One might argue that it is perfectly possible to do this without actuallyintegrating the results of, for example, a Conversation Analytic approach, or ananalysis of speech acts, in a transcript. However, a second bene� t comes withattempting to pin down one’s analysis on paper (or screen), and that is the circum-stance that one is forced to make decisions of classi� cation every step of the way,and even if it is the decision to admit to insecurity concerning a label or category(witness the numerous instances of ‘?’ preceding labels in the extracts below).Multiple instances of insecurity will lead to yet another cycle of analysis (or morethan one), possible modi� cations of the categorizations used, or the inclusion ofother criteria for analysis. This is essentially a learning or discovery process, andthe transcript always remains a record of work in progress. From the point of viewof the reader, multiple layers, and the labels and de� nitions that go with them at� rst sight make the transcript hard to manage. However, including the shorthand

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 9: Transcribing discourse: interactions with Alzheimer's disease

T ranscribing discourse 353

version of our analysis in our transcript should make our process of analysis moretransparent, and thus, obviously, also more open to peer assessment (in as far asthis is possible without direct access to the recordings themselves). The multiplelayers should also serve as a reminder to readers that what is in front of their eyesis not ‘the data’, but rather what we have made of the data, in other words, ourcategorization and shaping of the data should become more visible.

How practical is a multi-layered approach with large amounts of data?

The practicalities of transcribing, and ‘doing discourse’ with large amounts of data fallinto two basic categories: demands of time spent transcribing and analysing (see alsoHoward and Heselwood, 2002), and demands of space taken up by the transcript. Thetime spent transcribing is certainly a (sometimes costly) factor in any data-intensivestudy of discourse, whether one chooses to ‘only’ do a detailed orthographic transcript,or also to transcribe one’s analysis. In the latter case, one is, in our own experience,more likely to spend more time on doubtful instances, passages where assigning labelsbecomes diYcult. However, this also has an added learning bene� t. Space taken upby a transcript, and the presentation of data is, as discussed in the lead article, chie� ya problem for publication. We have opted for a compromise in presenting two extracts,while oVering access to the whole transcript to interested colleagues on a personalbasis. In future, other possibilities, such as the simultaneous publication of full tran-scripts in (copyrighted!) electronic archives may arise.

And � nally: what’s new here?

As brie� y outlined above, there is a lot of recent and current work in clinicaldiscourse studies. Any qualitative approach to communication disorders shouldentail detailed analysis of primary data. Thus why should we advocate a multi-layered approach to doing discourse? We cannot, and do not wish to, pretend tohave found the philosophers’ stone of transcribing and presenting data. What wewould like to do is invite our readers to reconsider some of the processes involvedin analysing interactions: limitations, selections, categorizations, imposition andadaptation of existing frameworks, to reiterate only a few. Integrating these into atranscript, shaping the discourse from the data, is a relatively simple way of increasingthe transparency of one’s analysis, of making the analysts’ input more visible andaccessible to readers (and indeed the analysts themselves).

Extract 1

P {lento }19 O R did you have a good Christmas.

D-SA Q-recapD-CA Rep, ATEst

20 O F well (.) alright, didn’t (x) do anything, (xx) 5D-SA E RD-CA ?TU TExp

21 O R 5 did you get any presents?D-SA QD-CA TExp

(3)

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 10: Transcribing discourse: interactions with Alzheimer's disease

N. Muller and J. A. Guendouzi354

22 O F no:: (we’re just left here dead) 5D-SA R ED-CA ?TExp

23 O R 5 mhm 5D-SA D-ContD-CA MTn

24 O F 5 ¯ oh:: wha- what I want is to (3.5) I don’t know (.) what you call it (1) I callD-SA E Q-CuD-CA TS

25 O it when (something changes or) your own brain, 5

26 O R 5 ¯ [o:h*D-SA ?D-ContD-CA MTn

27 O F 5 [(xxx*xxx xx) 5

P {LV }28 O R 5 your brain. (( light laugh)) my brain. my brain is going,

D-SA ?R-Cf RD-CA ?ATU TIB TS-Rel

(3.5)29 O F but eh, (2.5) wh- when I came in here (1.5) (but I knew xxxx to this xxx

D-SA R ?RD-CA ?ATEst/?TS-Rel

30 O an it x many [(xxx)*

31 O R [your memory,*D-SA Q-CfD-CA ATU, RepP {pianissimo }

32 O F hm (.) but they er (xx naI@) well my husband he didn’t want that (.) he didn’tD-SA R RD-CA ?TExp/?TS-Rel

33 O want me to have any treatment at all, 5

34 O 5 ¯ o::hD-SA D-ContD-CA MTn

35 O (xxxx b@t) (.) and it was ( just) he hhh [(xx),*D-SA RD-CA ?ATexp

36 O R [he didn’t* want you to have treatment, noD-SA R-CfD-CA ATU, Rep

37 O R (.) he didn’t want you to have treatment.

38 O F (that’s it) (2) he wants me to be just kept here.D-SA R-Cf RD-CA Rep TE

39 O R mhm,D-SA D-ContD-CA ?TIB

40 O F not (this xx xxxxx 1sec) and he’s good to me (1) @m@ my eldest brother (.)D-SA ?R E RD-CA ?TExp ?TExp ?TExp/?TS

41 O (thought) what he wanted to be (.) kept here (.) and (@t@ @m@wEI@m @m@)

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 11: Transcribing discourse: interactions with Alzheimer's disease

T ranscribing discourse 355

42 O (xxx) he- he’s marvellous (.) to me (x)D-SA ED-CA ?TExp

P {lento }43 O R Mick or your brother (.) Mick or your brother.

D-SA Q Q-recapD-CA ?ATU, TIB-Rep

44 O my brother. (.) (ÂIks@ne:) ((sniVs)) (.) but @ my brother would like me toD-SA R-Cf RD-CA TExp

45 O come out (xx waiting me) jus bein (.) bein kept well an everything.

46 O R m:::,D-SA D-ContD-CA MTn

47 O F and my own (b@l brother xxxx (.) xxxxx)D-SA ?RD-CA TExp

(0.5)48 O R what’s your brother’s name?

D-SA QD-CA ?TS-Rel/?ATEst

(4)P {pianissimo; W }

49 O F (my brother my brother) (10 ) (xxxxx) 5D-SA ? ‘placeholder’D-CA ?ATExp

50 O R 5 where does your brother live,D-SA QD-CA ?TS-Rel/?ATExp

P {pianiss.}51 O F a: :h (3) oh dear

D-SA ED-CA TIB

52 O R Yorkshire?D-SA QD-CA Rep, ATExp

53 O F no?D-SA ?Q-ContD-CA TIB

P {allegro }54 O R oh. doesn’t live in Yorkshire. (2) does he live in CardiV,

D-SA R-Cf QD-CA ?Rep TExp

55 O F no? (2) ( XX there,) (3) oh (it had been Xx away with me (.) an theyD-SA ?Q-Cont RD-SA TIB, ?TYC ?ATS-Ret

56 O want me to (xxXXxx tUk,)

P {pianiss.} {forte }57 O R a:h ri:ght. who took it away.

D-SA R-Cf QD-CA ?Rep ATU

(10)

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 12: Transcribing discourse: interactions with Alzheimer's disease

N. Muller and J. A. Guendouzi356

58 O F I get ( X[xxx)*D-SA RD-CA ?ATS/?ATExp

59 O R [what’ve* you done to your leg.D-SA QD-CA TS-Ret

(0.5)60 O F who? (xx) (.) ( XX)

D-SA Q-ContD-CA TIB

61 O R it’s on this side (3) no it’s this one here, (0.5) did you bump it.D-SA R R QD-CA TExp

62 O F I would’ve bumped it wi’ bein pushedD-SA RD-CA TU TExp

Extract 2

148 O F yes (.) he (would) er take me out for (.) er twenty minutes (.) round the greenD-SA R RD-CA TExp

149 O an the grass you know what you know the @patch of grass (.) the he @ he wouldD-SA R-Cf RD-CA

150 O take me out there but if I think he’s going out (.) and around (xxx on theD-SA ?E

151 O ground) or shopping (.) oh,D-SA ED-CA ?ATS

152 O R do you like going shopping?D-SA QD-CA TU/TExp

P {LV}153 O F I used to do, (1) but I don’t know now,

D-SA R ?E/?RD-CA TExp TYC

154 O R what sort of shopping do you like. clothes shopping? or [food shopping*D-SA QD-CA TExp

155 O R [I like the window* shopping yeah,D-SA RD-CA TExp

(1.5)156 O R where did you use to do your shopping?

D-SA QD-CA TExp

P {LV }157 O F we used to do it in @, (2) it’s so long (been Xx)

D-SA R ?R/?ED-CA ATU ?TIB, TYC

(2)

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 13: Transcribing discourse: interactions with Alzheimer's disease

T ranscribing discourse 357

158 O R Albany Rd ?D-SA Q-CfD-CA Rep

159 O F yes I like (there) (3) but its so so long ago:: and I wouldn’t knowD-SA E R ?E/?RD-CA TExp

P {LV }160 O where to start,

D-CA TYC(1.5)

161 O R well you get lost in the shops the[se days they’re* so big,D-SA RD-CA TS-Rel

162 O F [o:h yea:h*D-SA D-ContD-CA MTn

163 O R and Tescos, they’ve got a new Tescos?D-SA ?R-CfD-CA TExp

164 O F m::?D-SA D-ContD-CA MTn, TIB

165 O R by (0.5) by the river.D-SA RD-CA Rep

(1.5)P {LV }

166 O F I don’t know where it is 5D-SA ?ED-CA TIB TYC

Notes

1. The ‘normality’ of the interactants and their interactional patterns tends to be taken as agiven, rather than as something to be questioned. Even in ethnography or interactionalsociolinguistics, where one of the basic tenets is that a priori, one cannot dismiss any aspectof an interaction as non-relevant, there tends to be an implicit understanding that mutualunderstanding, for example, is something that happens—albeit in a way that is culturallyde� ned. This is, however, not something that can necessarily be taken for granted inneurogenic disorders such as SDAT.

2. For other discussions of possible de� nitions of discourse, see e.g., Brown and Yule (1983),or SchiVrin, Tannen and Hamilton (2001).

3. There is a third participant (N, a nurse at the care facility) who features very brie� y, albeitnot in the extracts we are presenting here.

4. Interested readers may contact the � rst author for a read-only copy of the transcript. Datapresentation in clinical discourse generally needs to steer a compromise course between,on the one hand, limitations of space, and, on the other, the demands of peer scrutiny andtransparency of one’s work.

5. See e.g., JeVerson (1979) and (1985) for orthographic and analytic categorizations oflaughter.

6. For a fuller discussion of the orthographic layer, see Damico and Simmons-Mackie (2002).7. See e.g., Atkinson and Heritage (1984) for the use of parentheses to mark ‘transcrip-

tionist doubt’.8. Our category ‘question’ is at times referred to as ‘rogative’ (e.g., Leech, 1983; see also

Peccei, 1999).

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 14: Transcribing discourse: interactions with Alzheimer's disease

N. Muller and J. A. Guendouzi358

9. And, of course, as long as both readers and analysts are conscious that our use of theterm SA analysis is at a considerable remove from classical Speech Act Theory (e.g.,Austin, 1962; Searle, 1969).

References

Aphasiology, 1999, Special Issue: Conversation Analysis. Aphasiology, 13 (4/5), 239–445.Atkinson, J. M. and Heritage, J., 1984[1999] , Transcript Notation. Structures of Social

Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press, pp.ix-xvi. Reprinted in: Aphasiology, 13, 243–249.

Austin, J., 1962, How to do Things with Words (Cambridge, MA: Harvard University Press).Ball, M. J., 2002, Transcribing disordered speech: the segmental and prosodic layers. Clinical

Linguistics and Phonetics, 16, 329–344.Brown, G. and Yule, G., 1983, Discourse Analysis (Cambridge: Cambridge University Press).Damico, J. S. and Simmons-Mackie, N., 2002, The base layer and the gaze/gesture layer of

transcription. Clinical Linguistics and Phonetics, 16, 317–327.Damico, J. S., Simmons-Mackie, N., Oelschlaeger, M., Elman, R. and Armstrong, E.,

1999, Qualitative methods in aphasia research: basic issues. Aphasiology, 13, 651–665.Damico, J. S., Oelschlaeger, M. and Simmons-Mackie, N., 1999, Qualitative methods in

aphasia research: conversation analysis. Aphasiology, 13, 667–679.Ferguson, A., 2000, Maximising Communication EVectiveness. In N. Muller (Ed.) Pragmatics

in Speech and Language Pathology. Studies in Speech Pathology and ClinicalLinguistics, vol. 7, pp. 53–88.

Howard, S. J. and Heselwood, B. C., 2002, Learning and teaching phonetic transcriptionfor clinical purposes. Clinical Linguistics and Phonetics, 16, 371–400.

Jefferson, G., 1979, A technique for inviting laughter and its subsequent acceptance declina-tion. In G. Psathas (Ed.) Everyday Language: Studies in Ethnomethodology (NewYork: Irvington), pp. 79–96.

Jefferson, G., 1985, An exercise in the transcription and analysis of laughter. In T. Van Dijk(Ed.) A Handbook of Discourse Analysis. Vol 3: Genres of Discourse (London:Academic Press).

Gumperz, J. J., 1992, Contextualization and understanding. In A. Duranti and C. Goodwin(Eds) Rethinking Context: Language as an Interactive Phenomenon (Cambridge:Cambridge University Press), pp: 229–252.

Leech, G., 1983, Principles of Pragmatics (London: Longman).Levinson, S. C., 1980, Speech Act Theory: The state of the art. Language Teaching and

Linguistics: Abstracts, 13, 5–24.Peccei, J. S., 1999, Pragmatics (London: Routledge).Perkins, M. R. (2000) The scope of pragmatic disability: A cognitive approach. In N. Muller

(Ed.) Pragmatics in Speech and Language Pathology. Studies in Speech Pathology andClinical Linguistics, vol. 7, pp. 7–28.

Schiffrin, D., Tannen, D. and Hamilton, H. E., 2001, The Handbook of Discourse Analysis(Oxford: Blackwell ).

Searle, J. (1969) Speech Acts (Cambridge: Cambridge University Press).Simmons-Mackie, N. and Damico, J. S., 1999, Qualitative methods in aphasia research:

ethnography. Aphasiology, 13, 681–687.

Appendix 1. Key to transciption symbols

P Prosody and voice qualityLV: ‘laughing’ voice quality; TV: ‘Tearful’ voice quality; W: whisper; lento:slow; allegro: fast; piano: quiet; forte: loud; { } mark beginning and end of thecharacterization.

O Orthographic transcription; ‘base line’(Friday) Transcriber’s best guess; unclear speech(xxXx) Unintelligible speech; number of syllables recognizable; x 5 less

prominent syll.; X 5 more prominent syll.

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.

Page 15: Transcribing discourse: interactions with Alzheimer's disease

T ranscribing discourse 359

((sniVs)) Nonlinguistic or paralinguistic aspects of an utterance, oraccompanying environmental noises or events

5 Latching[ Overlapping speech begins* Overlapping speech ends- Abrupt cut-oV: prolongation of a speech sound(.) Brief pause(2.5) Pause, timed in seconds? Rising intonation; not necessarily a question, ‘Continuing’ intonation (mid-level; may include a slight rise or

fall ). Falling intonation; not necessarily end of sentence( Marked (rise and ¯ fall in intonation, beyond what is marked by

‘?’ and ‘.’treatment Underlining marks a heavily stressed syllable

D Discourse line? Doubtful interpretation/ Alternative interpretationD-SA: Speech Act analysis

D[irective]: request for action on addressee’s part; D-Cont[inuer]: requestfor addressee to keep talking;E[xpressive] : expression of speaker’s state of mind or emotional state, orattitude;Q[uestion]: request for information;Q[uestion]-C[on]f [irmation]: request for con� rmation of own understand-ing;Q[uestion]-C[on]t[radiction]: rejects a previous proposition;Q[uestion]-Cu[e]: hybrid; self-cue; becomes request for information whenself-cue fails);R[epresentative] : Speaker makes a statement about the world;R[epresentative]-C[on� rmation]: con� rms previously given information;recap[itulation]: repetition or close paraphrase

D-CA: Conversation AnalysisA[ttempted ]: analyst’s interpretation that utterance function in terms ofconversation structuring/management is only imperfectly realizedM[inimal ] T[ur]n: minimal utterance that hands the turn immediately backto the previous speakerRep[air]: conversation is realigned to enable mutual understanding;T[opic] E[stablishment]: utterance reintroduces a topic that was not takenup in the preceding utteranceT[opic] Exp[ansion]: utterance adds to previously introduced topic;T[rouble] I [ndicating] B[ehavior]: utterance indicates the need for RepairT[opic] S[hift]: utterance introduces a topic; Rel[ated]: shows content-basedlink to immediately preceding topic; Ret[rievable]: topic is retrievable fromverbal or non-verbal contextT[opic] U[ptake]: utterance signals that a topic introduced in a precedingutterance is taken upT[urn] Y[ielding] C[ue]: Signal that turn is about to end

Clin

Lin

guis

t Pho

n D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Nor

th T

exas

on

11/1

1/14

For

pers

onal

use

onl

y.