This article was downloaded by: [Swinburne University of Technology]On: 24 August 2014, At: 17:51Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Psychotherapy ResearchPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tpsr20

The sound of change: A study of the psychotherapeuticprocess embodied in vocal expression. Laura Rice'sideas revisitedAlemka Tomicica, Claudio Martinezb & Mariane Krauseca Department of Psychology, Universidad de Los Andes, Santiago, Chileb Faculty of Psychology, Universidad Diego Portales, Santiago, Chilec Department of Psychology, Pontificia Universidad Catlica de Chile, Santiago, ChilePublished online: 18 Mar 2014.

To cite this article: Alemka Tomicic, Claudio Martinez & Mariane Krause (2014): The sound of change: A study of thepsychotherapeutic process embodied in vocal expression. Laura Rice's ideas revisited, Psychotherapy Research, DOI:10.1080/10503307.2014.892647

To link to this article: http://dx.doi.org/10.1080/10503307.2014.892647

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/loi/tpsr20http://www.tandfonline.com/action/showCitFormats?doi=10.1080/10503307.2014.892647http://dx.doi.org/10.1080/10503307.2014.892647http://www.tandfonline.com/page/terms-and-conditionshttp://www.tandfonline.com/page/terms-and-conditions

EMPIRICAL PAPER

The sound of change: A study of the psychotherapeutic processembodied in vocal expression. Laura Rices ideas revisited

ALEMKA TOMICIC1, CLAUDIO MARTINEZ2, & MARIANE KRAUSE3

aDepartment of Psychology, Universidad de Los Andes, Santiago, Chile; bFaculty of Psychology, Universidad Diego Portales,Santiago, Chile & cDepartment of Psychology, Pontificia Universidad Catlica de Chile, Santiago, Chile

(Received 18 October 2012; revised 12 August 2013; accepted 5 February 2014)

AbstractObjective: The purposes of this article were to compare the characteristics of the vocal quality of therapists and patients inchange and stuck episodes, and to depict patient-therapist interaction sequences of vocal properties, in order to analyzemicro-regulatory processes within the psychotherapeutic interaction. Method: Application of the Vocal Quality Patternscoding system to a study of a sample of change and stuck episodes, taken from six psychotherapies. Results: The resultsmade it possible to show that the 15 psychotherapeutic change process are embodied in the modes of vocal expression oftheir participants, and that the way in which these different modes are coordinated within the interaction makes it possible toobserve regulatory micro-sequences that participate in the therapeutic change process.

Keywords: process research; alliance

Some three decades ago, Laura Rice (d. 2004)developed a highly original process-rating measurethat assessed Client Vocal Quality (CVQ). The ideasbehind this method stressed the value of closeobservations of psychotherapy process and the cre-ation of meaningful coding categories based on thoseobservations, the importance of the style of patientsexpressions rather than their content, and the relev-ance of the way in which these non-verbal expres-sions account for the activation of recurringcognitive-affective schemas of the patients. Thesegeneral ideas, along with the CVQ rating system,were presented in three inspirational papers. In thefirst of them, Rice and Wagstaff (1967) focused onclients vocal quality and expressive style as indica-tors of psychotherapy productivity; they also definedfour subclasses of qualitatively different kinds ofvoice patterns that varied among clients andthroughout the sessions, which allowed them todifferentiate good and bad sessions in agreementwith the therapists assessment. In the second one,

Rice and Kerr (1986) depicted the developmentprocess of the Client and Therapist Voice Qualitysystem (CVQ and TVQ), the studies of reliabilityand validity of both systems and its applications inpsychotherapy process research. In the third one,Wiseman and Rice (1989) observed significanteffects regarding the influence of the therapists vocalquality over the patients vocal quality.

Following the pioneering ideas of Rice and collea-gues, we developed a new coding system of VocalQuality Patterns (VQP) applicable to psychothera-peutic dialogue, a single tool for patients and thera-pists, designed to fit the Chilean cultural contextand possibly a Hispanic one; this system was pre-sented in a previous work (see Tomicic, Martnez,Chacn, Guzmn, & Reinoso, 2011). This studypresents the systems application to the analysis ofrelevant episodes in psychotherapy. Our purposewas to observe the process of change embodied inthe expressive vocal styles of the participants, and touncover regulatory sequences between them.

Correspondence concerning this article should be addressed to Alemka Tomicic, Department of Psychology, Universidad de Los Andes,Santiago, Chile. Email: atomicic@gmail.com

Psychotherapy Research, 2014http://dx.doi.org/10.1080/10503307.2014.892647

2014 Society for Psychotherapy Research

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

mailto:atomicic@gmail.comhttp://dx.doi.org/10.1080/10503307.2014.892647

The Relevance of the Voice in HumanCommunication

The human voice has the property of giving qualityto the speech by the so-called suprasegmental orprosodic parameters (Knapp, 1988; Martinez, 2003;Nespor & Vogel, 1994). According to psychoacous-tic studies (Sapaly, 2005) the production and per-ception of voices depend on the psychophysiologicalstate of the individual, as well as prosodic features ofa particular language. Both factors allow the distin-guishing of the basic properties of intensity, pitch,and timbre of the voice (Sapaly, 2005). Thus, theperception of voices relies on the interplay betweenobjective acoustic properties and the subjectivecondition of the listeners, determined in part bytheir psychophysiological characteristics.

According to Andersen (1999), voice parameterssuch as rhythm, tempo, resonance, control, andaccent are key factors for the interpretation ofspeech. He points out that, in general terms, non-verbal communication stands out in informationtransmission because it is the main carrier of com-munication at a relational and interpersonal level.In the authors view, vocal nonverbal contextualcodes are the most important ones, since theyreinforce or modify what is said verbally. Forexample, Campanelli, Iberni, Sarracino, Degni, andMariani (2007) have referenced the autonomoussemantic properties of nonverbal vocal communica-tion, identifying four functions of nonverbal orparalinguistic vocal cues: (a) hypercoding (nonverbalvocal parameters can clarify and specify the meaningof verbal units), (b) opposition (voice acousticcharacteristics contrast with the meaning of verbalunits and a new meaning emerges from this opposi-tion), (c) modification of the meaning conveyedthrough the verbal code (for example, the illocu-tionary force of speech acts), and (d) intensificationand deintensification of the meaning of verbal units.

In everyday communicative interactions, voiceprosody plays a key role, giving an account ofimplicit meanings and aspects occurring duringmutual communication (Campbell, 2004a, 2004b).Without recognition of the prosody accompanyingspeech content, it is almost impossible to recognizedifferences between stated contents (Sarko, Roth &Martin, 2010) as well as it being very difficult toidentify the implicit elements defining the type ofongoing interaction. Specifically, changes in thevoice allow attributions about the speakers emo-tional states, and additionally emotions affect thesonority of the messages (Bnzinger& Scherer, 2005;Malloch, 1999; Papouek, 2007, Scherer, 1982).However, this association between speech prosodyand emotions has not been described as a one on

one correspondence but as a connection betweenvocal qualities and types of emotions. It has beenobserved that the association between emotions andsound of speech is established as part of a gradualprocess of relationship consolidation and interactionbetween individuals (Gobl & Chasaide, 2003). Inturn, the ability to recognize emotions from theprosody may be affected by the presence of psycho-logical disorders in the individual. For example, in astudy of individuals with eating behavior disorders,they showed a significant deterioration in the abilityto recognize facial and vocal emotions, mostly inthe case of negative emotions (Kucharska-Pietura,Nikolaou, Masiak, & Treasure, 2004). Thus, identi-fication of the vocal qualities may be very importantfor the recognition of changes in the interlocutorsaffective states (Gobl & Chasaide, 2003), a necessarycondition for stable, meaningful, and satisfyingsocial relationships (Knapp, 1988).

The Voice in the Psychotherapeutic Dialog

In psychotherapeutic interaction, the speech acousticparameters are one of the main sources of informa-tion about the meaning of the acts performed inpatient-therapist communication (Knoblauch, 2000,2005). Even more so, the idea has been advancedthat the quality of the speakers voice may influencethe emotional state of the listener (Bachorowski &Owren, 2008; Russel, Bachorowski, & Fernandez-Dol, 2003); for instance, a voice that reflected thetherapists relaxedness and confidence could calmthe patients agitated voice and its associated emo-tions (Bady, 1985). Similarly, in a study on vocalcommunication Scherer and Bergmann (1990)found that psychotherapy participants infer andcause emotions in each other through the sounds ofspeech. Knoblauch (2000, 2005) has stated that thevocal quality of patients and therapists can be one ofthe key aspects for describing change processes inpsychotherapy. He contends that, in psychothera-peutic interaction, psychological meanings areexchanged not only through participants speech,but also through the sounds of their voices.

In an exploratory study, Tomicic, Bauer, Martnez,Reinoso, and Guzmn (2009) researched the import-ance that psychotherapists ascribe to the nonverbalaspects of their own voices and of their patients in thepsychotherapeutic process. One of the studys salientresults was that the therapists interviewed not onlyacknowledged generic voice use as a non-verbal orpara-verbal tool, but also its specific functions accord-ing to the identification of diverse characteristics of theprocess and of psychotherapeutic space. The atten-tion paid to voice quality allowed them to differentiateemotions and improve their understanding of their

2 A. Tomicic et al.

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

patients; in addition, they also stated that their voicewas a tool to produce changes (e.g., transformation ofthe sessions emotional climate, use of the voice as achange marker during the psychotherapeutic process,and use of the voice for specific interventions).

In a second qualitative study Bauer et al. (2010)explored the importance that patients assign to thenon-verbal aspects of their own voices and those oftheir therapists during their psychotherapies. Specif-ically, the patients were asked whether therapistsrespond to the patients voice and how did they doso; conversely, they were asked whether theyresponded to the therapists voice and in whatmanner. The results showed that interviewed patientsdescribe voice tone in a differentiated way and thatthey regulate affects and atmospheres through theirown voice. They experience their therapists voicetone fairly consciously and are able to indicate theideal therapists voice. They also said they heartherapeutic intentions from the voice tone of theirtherapists and perceive the regulation of the thera-peutic relation through vocal and rhythmic elementslike melody, intensity, and tempo.

The Study of the Voice in PsychotherapeuticInteraction

In the field of psychotherapy, few studies about thevoice have been conducted, and existing ones havefocused on the analysis of vocal expression of emo-tions (see Bachorowski & Owren, 1995, 2008, 2009;Moneta, Penna, Loyola, Buchheim, &Kchele, 2008;Scherer, 1982). That is, they have focused on whatCampbell (2007) has called speaker-centered speecha line of study which assumes that different emo-tional states are linked to different lexical choices,speech styles, and phrasingswhile neglectingcommunication-centered research. This line of studyassumes the existence of a prosodic modulation ofspeech in interaction, i.e., those speakers emotionalstates may vary due to changes in the state of therelationship with their conversational partners;furthermore this intentionality of speakers is con-nected to different lexical choices, speech styles, andphrasings.

This notion by Campbell (2007) is consistent withtheoretical and empirical developments derived fromthe intersubjective approach, which support the linkbetween mutual regulation and the processes thatfoster change in psychotherapy. Each participant ofthe therapeutic dyad is believed to be affected bothby his/her own self-regulation behaviors, and bythose of his/her partner, in a process of mutualregulation (Tronick, 1989). This process is thoughtto happen moment by moment, mainly at a non-verbal level (Stern, 1998). Mutual regulation is

observed in the coordination sequences of (verbaland nonverbal) behavior between the dyad members,assuming that the regulatory behaviors of one parti-cipant can be predicted from those of the other, andvice-versa (Beebe, 2006; Jaffe, Beebe, Feldstein,Crown, & Jasnow, 2001). Regarding its connectionto change, Tronick (1998) has suggested that mutualregulation allows for the expansion of consciousnessof the dyads members. Each partner (mother andinfant, or therapist and patient) affects the othersstate of consciousness. As each affects the othersself regulation, each partners inner organizationis expanded into a more coherent as well as a morecomplex state. This happens because regulationwith another person in the interaction enables him/her to enrich and transform his/her own functioning(Beebe & Lachman, 1998).

Laura Rices Ideas Revisited

Rice andWagstaff (1967) focused on communicationand identified four CVQs in patients: (i) emotional,in which the speech pattern is disrupted or distortedto some extent by emotional overflow; (ii) focused;this category involves inward turning of attentionalenergy, deployed to tracking inner experience andfinding a way to symbolize it in words; (iii) external-izing; this category seems to involve a deploymentof attentional energy outward in an effort to producesome effect in the outside world; it has a premoni-tored quality, suggesting that the content beingexpressed is not being newly experienced and symbo-lized; and (iv) limited; this pattern seems to involve aholding back or withdrawal of energy. This patternsuggests a walking-on-eggs quality. The authorsestablished connections between CVQs presenceand the productivity of the sessions as assessed bythe therapists (Rice & Kerr, 1986; Rice & Wagstaff,1967). Subsequently, Wiseman and Rice (1989)described seven vocal quality patterns for therapists,which are different and non-comparable with theCVQs. They are: (i) softened; this pattern is char-acterized by a lax voice that creates an intimacy andinvolvement effect; (ii) irregular; this pattern ischaracterized by irregular intensity stresses withsome pitch variation; (iii) natural; this pattern ischaracterized by neither an overly tense nor relaxedvoice; the voice is unstrained and natural; the effect isone of interest; (iv) definite, a pattern characterizedby a voice full, measured, assured, generally onthe speakers pitch platform; this category can occa-sionally sound overbearing and confrontational;(v) restricted; this voice pattern is characterized asadequate to carry the content, but strained; the voicecan be slightly tremulous, whiny, droning, or sound-ing as though the air is escaping before the word is

The sound of change 3

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

formed; (vi) patterned; this category is patterned foremphasis, specially using pitch. The category as awhole sounds singsong; and (vii) limited; thiscategory is characterized by a low energy or flatstresses in the voice and a monotone pitch; there isjust not enough life in this voice (Rice & Kerr, 1986).Wiseman and Rice (1989) found that these vocalqualities of the therapist had significant effects onthose of patients. Specifically, they observed that aproductive vocal quality of the therapist (irregularTVQ) could predict a change in the patient towards afocused vocal pattern, one that revealed his/herinvolvement in the experience.

In a previous study Tomicic et al. (2011), follow-ing Rice and Wagstaff (1967), developed a VQPcoding system for patients and therapists whichcould be applied to psychotherapeutic dialog in thecultural context of Chile and other Spanish-speakingcountries. This work was inspired mainly by themethod Rice and colleagues designed to define theircategories. Nevertheless, it was not intended torepeat their categories in another language. Instead,a single VQP coding system for both participants wascreated, allowing a comparison of categories used byeach. While certain convergences can be found inboth categorical systems, for example between cat-egories CVQ emotional and VQP emotional-expressive, or TVQ definite and VQP affirmative,these similarities account for the capacity of bothsystems to capture expectable phenomena in thefield of sonorous properties of oral communication,such as speech modulation by emotion or certainty.Regarding the language and cultural context of VQPdevelopment, experimental studies support thehypothesis that the vocal expression of emotions ischaracterized by elements that are universally usedby speakers and can be decoded with accuracy levelsthat exceed those expected by chance, regardless ofthe language of the listener (Scherer, Banse, &Wallbott, 2001; Thompson & Balkwill, 2006). How-ever, evidence has also been reported that empha-sizes the comparative advantage of the recognition ofemotions expressed through speech prosody bylisteners who share the language of the speaker(Pell, Monetta, Paulmann, & Kotz, 2009). Alongthese lines, Pell et al. (2009) have argued that whilethe ability to recognize emotions through vocalexpression is partially independent of the listenerslanguage correspondence and implies universal prin-ciples, it is also true that this ability is modulated bylinguistic and cultural variables. From this perspect-ive the consideration of these variables when devel-oping this coding system helps to achieve VQPreliability and ecological validity on its application.

In the Tomicic et al. (2011) coding system,VQPs were defined as a combination of specific vocal

parameters in the utterances of speakers whosespeech gives a specific impression to a listener,regardless of the contents transmitted (Tomicic,Chacn, et al., 2009). Six VQPs were established andcharacterized: (i) Report, (ii) Connected, (iii) Affirm-ative, (iv) Reflection, (v) Emotional-Expressive, and(vi) Emotional-Restrained. In addition, in the case ofutterances in which VQP coding does not apply,the following categories were created: (vii) FullPause, (viii) Overlapping, and (ix) Non-Codable (seeTable IV).

As mentioned above, the system, which proved tobe highly reliable (Tomicic et al., 2011), has at leasttwo relevant differences with respect to those createdby Rice and collaborators: (i) it describes VQPsapplicable to the Spanish language and (ii) it proposesa single coding system for both patients and thera-pists, which makes it possible to conduct comparativeand sequential studies of the VQPs used by boththroughout the psychotherapeutic process.

The Study of Interaction in Psychotherapythrough Relevant Episodes

A common and useful way of studying the psycho-therapy process is to look at specific moments withinthe therapy. These moments are special episodes,chosen from a theoretical point of view, and include:Innovative Moments (Gonalves, Matos, & Santos,2009), Episodes of Rupture (Safran & Muran, 1996,2000, 2006), or in the case of this study, ChangeEpisodes (Krause, de la Parra, Arstegui, & Strasser,2006; Krause et al., 2007) and Stuck Episodes(Fernndez et al., 2012; Herrera et al., 2009). Theseepisodes are regarded as windows (Elliott, 1984;Timulak, 2007) that make it possible to understandthe relationship between the therapeutic exchanges(e.g., verbal and non-verbal behaviors of the partici-pants) and its outcome. The use of these relevantsegments allows one to solve the practical problem ofmanaging an excessive amount of information, and,at the same time, assigns a theoretical meaning tosuch sampling. In this study change and stuckepisodes were used as relevant episodes.

A change episode is an interaction segment in apsychotherapeutic session in which a representa-tional-level change is observed in the patient. Themethod for determining change episodes, derivedfrom this definition, is based on the subjectivetheories notion of change and operationalized viaGeneric Change Indicators (see method section)(Krause et al., 2006).

We developed the following questions with respectto these episodes: Are certain specific VQPs morelikely to occur in change episodes? Are these VQPsthe same for the patient and for the therapist? And, is

4 A. Tomicic et al.

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

it possible to track regulatory exchanges expressed byVQP sequences within patient-therapist interactions?

Meanwhile, stuck episodes represent a temporaryhalting of the patients process of change due to thereedition of the problem during the therapeuticsession. They characterize this reedition as thepersistence of forms of understanding, behavior andemotion that contribute to sustaining the problemand stopping the progression of change as describedby change indicators (Fernndez et al, 2012; Herreraet al., 2009; Krause et al., 2006).

We developed the following questions with respectto these episodes: Are certain specific VQPs morelikely to occur in stuck episodes? Are these VQPs thesame for the patient and for the therapist? And, is itpossible to track regulatory exchanges expressed byVQP sequences within patient-therapist interactionsin these episodes?

Considering both episodes (change and stuck), thehypotheses were: (i) There are differences betweenchange and stuck episodes with respect to theprobability of some specific VQPs; (ii) There aredifferences in the VQPs used by patients andtherapists; and (iii) There are differences withrespect to regulatory VQP sequences in patient-therapist interactions in change and stuck episodes.

Method

The study design was transversal in order to deter-mine and compare the characteristics of vocal qualityin change and stuck episodes gathered throughindividual therapies. Furthermore a mixed designwas employed, combining qualitative and quantitat-ive methodologies. The former was oriented towardsdescribing and categorizing the patterns of vocalcharacteristics observed in patient-therapist interac-tions. The latter was aimed at establishing differ-ences between change and stuck episodes withrespect to patient-therapist interaction sequences ofvocal properties.

Participants. The change and stuck episodeswere taken from six successful short -term psycho-analytic therapies. The ages of the six patientsranged from 21 to 42 (average 34.7). Four of themwere female and two male. Patients received between15 and 23 sessions of therapy with a psychodynamicfocus in the context of outpatient treatment. Thetherapists, five males and one female, had between3 and 15 years of practical professional experience.All the treatments were evaluated by means of anoutcome measurement using the Outcome Ques-tionnaire 45.2 (OQ-45.2; Lambert & Burlingame,1996; von Bergen & de la Parra, 2002) (see Table I).

Procedures. Each of the 50-minute sessions ofeach therapy took place in a one-way-mirror room.All sessions were video and audio recorded for lateranalyses. Patients and therapists were extensivelyinformed before commencing therapy and consentedto video and audio recordings and data collection atall times. All participants provided their writteninformed consent concerning the use of their datafor research purposes.

The unit of analysis was the relevant episode,specifically change and stuck episodes. The methodfor determining change episodes is based on thesubjective theories notion of generic change (Krause,2005; Krause et al., 2007). Subjective change isoperationalized by means of Generic Change Indi-cators (Krause et al., 2006), which make it possibleto identify a change moment based on its content (seeTable II). In turn, a change episode is an interactionsegment where a change moment takes place. In therating procedure, this moment marks the end of theepisode. At this point, a rater establishes the begin-ning of the episode by tracking back to when theparticipants start conversing about the content of thechange (Krause et al., 2006).

Considering the conceptual meaning of each GCI,the aforementioned hierarchy was organized by Alti-mir into three levels (see Table II) that reflect moreclearly the different phases of the psychotherapeutic

Table I. Characterization of the participants.

TherapyPatients

sexPatients

age Presenting problemTherapists

sexTherapists years of

practiceOQ at

beginningOQ

at endOutcome(RCI)

1 Female 37 Adaptative disorder Male 10 115 71 442 Female 36 Depression by

mourningMale 10 68 48 20

3 Female 40 Depression Male 15 111 91 204 Female 42 Anxiety disorder Male 15 52 30 225 Male 32 Adaptative disorder Female 10 128 46 826 Male 21 Adaptative disorder Male 3 47 14 33

Note. The diagnostic was reported by each psychotherapist. Outcome: Successful therapies were defined as those that met the criterion tosubmit a Reliable Change Index (RCI = 15 or more) (von Bergen & de la Parra, 2002).

The sound of change 5

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

change process (Altimir et al., 2010). The first level(Level I) is named Initial consolidation of thestructure of the therapeutic relationship. becausethe change indicators that belong to this categoryreflect the relevance of the establishment of thepsychotherapeutic contract and relationship beforethe start of the process of constructing new explana-tions about the patients problems. The second level(Level II) is considered a medial one, and was namedIncrease in permeability. The change indicatorsof this cluster refer to several kinds of observed

transformations with respect to patients emotions,cognitions, and behaviors. Finally, the change indi-cators corresponding to the third level (Level III),named Construction and consolidation of a newunderstanding, reflect the patients elaboration of anew subjective theory and his or her autonomous useof the psychological context of interpretation.

Stuck episodes coincide with periods of timeduring the session, in which there is a temporarydetention of the patient change process characterizedby the lack of new understandings. This involves anargumentative persistence which opposes the con-struction of new understandings, and does notcontribute to the focus of change (Herrera et al.,2009). For its selection, the stuck episode must havea minimum duration of 3 minutes, and a distance of10 minutes from any episode of change. To delimitthe start and the end of the episode, the fragment isselected that better represents the stuck occurrence,therefore those segments in doubt or for which noconsensus among coders is achieved are left out(Fernandez et al., 2012).

For the selection and temporal delimitation of thechange and stuck episodes six pairs of coders trainedby the Chilean Research Program on Psychotherapyand Change analyzed videotapes and transcriptionsof the therapeutic sessions and carried out anintersubjective validation procedure.1

The sample for this study comprised 31 episodestaken from a universe of 152 episodes from sixtherapies. Table III presents the total of change andstuck episodes for each therapy together withsampled episodes. As can be seen, from a total of93 change episodes, 16 were selected; from a total of59 stuck episodes, 15 were selected. The selection ofthese episodes was deliberate, to choose changeepisodes belonging to each of the phases of thepsychotherapy process: Initial, medial, and final.Also, the selection was deliberate to match thechoice of change episodes from the initial phasewith those with a Level I type of change, changeepisodes from the medial phase with those with a

Table II. Generic Change Indicators (GCI).

Level Indicators

I. Initial consolidation ofthe structure of thetherapeutic relationship.

1. Acceptance of the existence of aproblem

2. Acceptance of his/her limits andof the need for help.

3. Acceptance of the therapist as acompetent professional.

4. Expression of hope5. Questioning of habitual

understanding, behavior andemotions.

6. Expression of the need forchange.

7. Recognition of his/her ownparticipation in the problems.

II. Increase in permeabilitytowards newunderstandings.

8. Discovery of new aspects of self.9. Manifestations of new behaviors

and emotions.10. Appearance of feeling of

competence.11. Establishment of new

connections.12. Reconceptualization of

problems and/or symptoms.13. Transformation of valorizations

and emotions in relation toself or others.

III. Construction andconsolidation of a newunderstanding.

14. Creation of subjective constructof self through theinterconnection of personalaspects and aspects of thesurroundings, includingproblems and symptoms.

15. Founding of the subjectiveconstructs in own biography.

16. Autonomous comprehensionand use of the context ofpsychological meaning.

17. Acknowledgment of helpreceived.

18. Decreased asymmetry betweenpatient and therapist.

19. Constructions of abiographically groundedsubjective theory of self andothers and of the relationshipwith surroundings.

Note. Taken from Altimir et al. (2010).

Table III. Characterization of study sample.

TherapyChange episodes Stuck episodes Total

Total Sample Total Sample Total Sample

1 10 3 15 3 25 62 14 3 7 2 21 53 24 3 12 3 36 64 20 3 12 2 32 55 11 1 4 2 15 36 14 3 9 3 23 6

Total 93 16 59 15 152 31

6 A. Tomicic et al.

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

Level II type of change, and change episodes fromfinal phase with those with a Level III type of change.The rationale for this sampling strategy was torepresent to some extent the ideal evolution ofthe process of change in psychotherapy. The selec-tion criterion for stuck episodes was that they shouldoccur within the nearest session of the changeepisode previously selected.

Data analysis. Each episode was analyzed by tworaters trained in the VQP coding system. They wereblind to type and phase of the episode. With the VQPcoding system the raters categorized patients andtherapists speech in terms of its aural quality. Thissystem identifies six VQPs: (i) Report, (ii) Connected,(iii) Affirmative, (iv) Reflection, (v) Emotional-

Expressive, and (vi) Emotional-Restrained. Also, forthe utterances in which the VQP coding does notapply, the following categories were created: (vii) FullPause, (viii) Overlapping, and (ix) Non-Codable (seeTable IV).

The VQP coding procedure was carried out infour analytic steps for each episode:

1. Listening to the full episode analyzed, so as tobecome familiar with the timbre of theparticipants voices.

2. Listening from the start of the episode ana-lyzed, reading the text speaking turn byspeaking turn, and performing a preliminarysegmentation, considering changes or break-downs in vocal quality as revealed by changes

Table IV. Characterization of Vocal Quality Patterns.

VQP Characterization

Report It adds to the speech the quality of something already known, of a disconnected speech of what is being said and/or acertain emotional distance. It sounds as if the speaker was reporting, narrating, or exploring content without anyemotional involvement. In this pattern, the central element is the listeners impression of a disconnected speech.Main Vocal Parameters: INTENSITY: Increased volume and high variations; DURATION: Speed augmented.

Connected It conveys the quality of being oriented toward the other (the partner in the dialogue) and of being developed while it isuttered. In this pattern, the central element is the listeners impression of an elaborative speech geared towards thepartner in the dialogue.Main Vocal Parameters: TONE: Dynamic-agogic accent, end of phrase half-suspended anti-cadence; INTENSITY:Volume increased sustained-crescendo dynamics and low variations.

Affirmative It conveys the quality of certainty and conviction. It sounds as if the speaker were teaching or instructing the listener, oras if he/she was very sure of what he/she is saying. In this pattern, the central element is the listeners impression of asecure and instructive speech.Main Vocal Parameters: TONE: Dynamic-tonic accent and end of phrase suspended; INTENSITY: Dynamicssustained-crescendo; DURATION: Hard vocal attack.

Reflection It conveys the quality of being directed toward oneself (the speaker). It sounds as if the speaker was connected with her/his internal world or in a dialog with her/himself. In this pattern, the central element is the listeners impression of anintroverted speech.Main Vocal Parameters: TONE: Dynamic-agogic accent and end of phrase half-suspended cadence; INTENSITY:Volume decreased and low variations; DURATION: Speed reduced.

Emotional-Expressive

It conveys affection and/or that the speech has a heavy emotional load. It sounds like the speakers emotion (joy, anger,sadness, fear, etc.). In this pattern, the central element is the listeners impression of an emotionally charged speech,regardless of the type of emotion.Main Vocal Parameters: TIMBRE: Clear / Bright; Clear / Opaque; Dark / Bright; and Dark / Opaque.

EmotionalRestrained

It conveys affection and/or that the speech has a heavy emotional load. However, even though in this case the speakersemotion is not audible, what does impress the listener is an effort to contain her/his emotion. In this pattern, the centralelement is the listeners impression of suffocation and control to avoid being overwhelmed by emotion.Main Vocal Parameters: DURATION: Speed decreased, not fluid pace, and long pauses.

Exclusion categories for VQPs

Overlapping It is an instance of simultaneous speech, which, in VQP coding, makes it impossible to distinguish the vocalcharacteristics of the participants in a full segment or speaking turn. When coding this conversation phenomenon, theoverlapping of the actors is noted.

Full Pause Short utterances with para-verbal content (hmm, aha, okay). They are usually ways of agreeing, showing attention,disagreeing, or displaying the wish to end a conversation. Their meaning depends mainly on the context and on certainvocal characteristics of the utterance; however, due to their brevity, they are hard to analyze in terms of the vocalparameters that define the VQPs described.

Non-Codable These are units of analysis which do not meet the phenomenological characteristics and the parameters of the VQPs. Itcan also apply to the cases in which the recording is not completely audible, due to ambient noises, mispronunciations,or other errors by the speakers. They are neither full pauses nor instances of overlapping.

The sound of change 7

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

in a vocal parameter (for example, speed,volume, or rhythm variations).

3. Listening from the start of the episodeanalyzed, speaking turn by speaking turnand segment by segment, and performing apreliminary coding, considering the pheno-menological description of VQPs.

4. Listening from the start of the episode ana-lyzed, speaking turn by speaking turn, andconfirming or discarding the presence of theVQP coded in step 3, considering the audit-ory perception of the vocal quality parametersinvolved, and attempting to reach a con-sensus when considering more than oneVQP per segment.

The VQP coding systems reliability in this sam-ple was assessed using Cohens kappa (Cohen,1968) to measure independent raters agreement.The analysis was performed with SPSS 14.0 andComKappa (where raters categories were notexactly the same). We used the 15% of the totalsample (n = 31). Five episodes (three changeepisodes and two stuck episodes) from four therapies(II, V, X, and XVI) were randomly chosen. VQPcoding in the chosen episodes resulted in k = .80,p < .05.

To compare the frequencies and the probability ofoccurrence of the different VQPs in change andstuck episodes, and in patients or therapists, weemployed the non-parametric chi-square test to-gether with logistic regressions.

To determine regulatory VQP sequences inpatient-therapist interactions in both types of epi-sodes, Lag Sequential Analysis (LSA) was applied,using the Generalized Sequential Querier program(GSEQ 5.0; Bakeman & Quera, 1995). LSA is astatistical methodology which makes it possible todescribe behavior sequences in a given interaction,determining the probability of occurrence of a givenbehavior (Lag 0) before the occurrence of a targetbehavior (Lag 1) (Bakeman, Deckner, & Quera,2005). This does not involve causality; instead, itcan be regarded as a prediction relationship.

The analysis produces two types of statistical data.First, descriptive information: Yules Q (valuesbetween 1 and 1, which are interpreted as a correla-tion) and Odds Ratio. Second, inferential informa-tion: Pearsons chi-square and the Adjusted Residualstest (Z > 1.96), recommended for the analysis ofcategory data and the exploration of sequentialrelationships, respectively (Paul & Liker, 1982;Bakeman, Adamson, & Strisik, 1995; Bakeman,Quera, McArthur, & Robinson, 1997).

Results

Regardless of the type of episode, patients andtherapists VQP repertoire displayed significant dif-ferences in distribution, 2 (8, N = 1353) = 165, 81,p = .000). In comparison with therapists, patients aremore likely to use report, reflection, and emotional-expressive VQPs. In contrast, therapists are morelikely to use full pauses. Patients and therapists useaffirmative and connected VQPs in similar proportions(see Table V).

Regarding the use of VQPs by patients in com-parison with therapists, and irrespective of the typeof episode, patients are more likely to use report (OR.320), reflection (OR .276), and emotional-expressive(OR .300) VQPs. Also, It is more probable thatpatients verbalizations correspond to non-codablecompared to therapists vocal quality (OR .397). Incontrast, as Table VI shows, regardless of the type ofepisode, therapists are more likely to display affirm-ative (OR 1.409) VQP and full pauses (OR 3.080).

Considering the episode type variable only, VQPdistribution in change and stuck episodes also dis-played significant differences, 2 (8, N = 1353) =50.352, p = .000). In change episodes, in comparisonwith stuck episodes, connected VQP is more prevalent.In contrast, in stuck episodes, affirmative VQPand full pauses are more frequent. In both episodes,report, reflection, emotional-expressive, and emotional-restrained VQPs, as well as the overlapping andnon-codable categories, appear in similar proportions(see Table VII).

With respect to the probability of appearance ofVQPs in change episodes in comparison with stuckepisodes, regardless of the participant using them, itis more likely for a connected VQP (OR .410) to beobserved in change episodes, whereas an affirmativeVQP (OR 1.474) is more likely to appear in stuckepisodes (see Table VI).

In order to analyze the effect of the type of episodeon the probability of using each VQP, the actorsand type of episode variables were regressed on

Table V. VQP repertoire in patients and therapists.

Patients Therapists

Report 13.7% 5%Connected 20% 31.1%Assertive 26.9% 32.7%Reflection 7.7% 1.7%Em-Expressive 8.4% 2.8%Em-Restricted 0.7% 0.7%Full Pause 4.8% 16.6%Overlapping 3.9% 4.7%Not Coded 13.9% 4.7%

100% 100%

8 A. Tomicic et al.

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

the probability of using different kinds of VQP.As Table VI shows, no interaction effects wereobserved between these variables in relation to theoccurrence of VQPs.

In order to explore the sequential associationbetween therapists and patients VQP categories, atable with all vocal behavior categories of eachmember of the therapeutic dyad was produced. Init, all VQP categories of therapists and patients areregarded as given behaviors (Lag 0) and targetbehaviors (Lag 1). The chi-square statistic for thislag table indicated that the degree of association

between the VQP categories was not determined bychance, 2 (289, N = 1353) =1640.58, p .01).Having established the global sequential associationbetween the VQP categories, a more specific exam-ination into pairs of associated categories was per-formed. As shown in Table VIII, 12 temporalsequences between VQPs were statistically signific-ant. These temporal associations between VQPs arereferred to as regulatory microsequences. In turn, weclassify them as self-regulation sequences if the tem-poral association between the VQPs occurs in thesame utterance by the patient or the therapist, or asmutual regulation sequences if the temporal associationbetween the VQPs matches the interaction betweenthe members of the therapeutic dyad.

Regarding self-regulation microsequences, in the case ofpatients only the temporal association between con-nected and reportVQPs was statistically significant, witha moderate degree of association (Yules Q = .35).Regarding therapists, two statistically significant self-regulation microsequences were observed. The firstwas the temporal association between report andaffirmative VQPs, and the second sequence was anoverlapping followed by affirmative VQP, both with ahigh degree of association (Yules Q = .67 and .59,respectively) (see Table VIII).

The mutual regulation microsequences were subdi-vided depending on whether a given behavior corre-sponded to a patient or a therapist VQP. When agiven behavior corresponded to a patient VQP, fivetemporal mutual regulation sequences wereobserved: Connected VQP for the patient and con-nected VQP for the therapist, connected VQP for thepatient and full pause for the therapist, affirmativeVQP for the patient and connected VQP for thetherapist, affirmative VQP for the patient and fullpause for the therapist, and finally emotional-expressiveVQP for the patient and connected VQP for thetherapist. As Table VIII shows, the strength of the

Table VI. Probability of occurrence of VQPs by actor and episodetype.

B Wald Sig Exp(B)

VQP AccountActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

1.141.214.073

15.376.347.030

.000**.556.863

.3201.2391.076

VQP ConnectedActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

.517.892.150

6.14524.996.320

.013*

.000**.571

1.677.4101.162

VQP AffirmativeActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

.343.892.150

4.2065.237.294

.040*.022*.587

1.4091.474.878

VQP ReflectionActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

1.286.265.532

7.631.188.623

.006**.664.430

.2761.304.588

VQP Emotional-ExpressiveActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

1.204.149.095

8.665.096.030

.003**.757.863

.300.8621.100

VQP Emotional-RestrictedActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

.1821.184.296

.0561.046.034

.813.306.854

.8343.2661.344

Full PauseActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

1.125.412.745

19.8153.7002.733

.000**.054.98

3.0801.5102.106

OverlappingActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

.114.222.165

.098.352.093

.754.553.761

1.1201.2491.179

Not CodedActor: Therapist (1) Patient(0) Episode: Stuck (1)Change (0)Actor Episode

.924.364.525

9.658.9291.452

.002**.335.228

.3971.438.592

**p

temporal association of these sequences ranged fromlow to high levels.

When a given behavior corresponded to a therapistVQP, four temporal mutual regulation sequenceswere observed: Affirmative VQP for the therapist andaffirmative VQP for the patient, affirmative VQP forthe therapist and full pause for the patient, full pausefor the therapist and connected VQP for the patient,and full pause for the therapist and affirmative VQPfor the patient. As Table VIII shows, the strength ofthe temporal association of these sequences rangedfrom moderate to high levels.

In order to compare the proportion of micro-regulatory VQP sequences in change and stuckepisodes, the aforementioned sequences were iden-tified in each type of episode and the chi-square

statistic was used to check for an association betweeneach sequence and the type of episode.

As shown in Table IX, only the self-regulationmicrosequence of connected and report VQPsobserved in patients showed an association withepisode type, with a higher proportion in changeepisodes. Four of the five mutual regulation micro-sequences in which the given behavior was a patientVQP displayed an association with episode type. Thesequences of connected VQP for the patient andconnected VQP for the therapist, connected VQP forthe patient and full pause for the therapist, andaffirmative VQP for the patient and connected VQPfor the therapist, were more often found in changeepisodes. In contrast, the sequence of affirmativeVQP for the patient and full pause for the therapist

Table VIII. VQP Lag 0 VQP Lag 1 temporal sequences.

Lag 0 Lag 1 N Adjusted residual p Value Yules Q Odds Ratio

Patient VQP PatientVQPConnected Report 17 2.63 .01 .35 2.09Therapist VQP Therapist VQPReport Affirmative 15 4.98

was more frequently found in stuck episodes. Themutual regulation microsequences in which thegiven behavior was a therapist VQP were mostlypresent in stuck episodes.

Discussion

The question addressed in this study was whether itwas possible to observe differences between changeand stuck episodes with respect to the probability ofspecific VQPs, the VQP used by patients andtherapists, and with respect to regulatory VQPsequences in patient-therapist interactions.

VQP Differences Between Patients andTherapists

The results did not reveal an interaction effect in theprobability of finding a specific VQP considering thetype of episode and actor variables. This means thatthe difference in the probability of encountering aspecific VQP in patients or therapists is the same inchange and stuck episodes. Thus, patients tendencyto use report, reflection, and emotional-expressive VQPsmay be connected to the fact that, in their role, theyare expected to present variations in their speech,with a detached vocal quality as in report VQP, withmoments of more introverted vocal quality as inreflection VQP, and others full of feeling, as inemotional-expressive VQP. Also, therapists more fre-quent use of affirmative VQP and full pause may beassociated with the role characteristics imposed bypsychotherapeutic tasks, which are expressedthrough a convincing and committed speech, as inaffirmative VQP. Full pause, which is usuallyexpressed with para-verbal vocalizations (e.g.,mmh, aha) may be linked to the therapists task ofdirecting the rhythm of conversational exchange inthe session.

VQP Differences Between Stuck and ChangeEpisodes

In addition, the differences observed in specificVQPs probability during change and stuck episodescan be understood given the operational character-istics of each VQP. Therefore, the higher probabilityof connected VQP in change episodes is consistentwith the idea that they are interaction segmentsduring the session which display changes in thepatients subjective theory about him/herself, his/herrelationship with others, and his/her problems(Krause et al., 2007). This consistency becomesevident when we consider that connected VQP sug-gests to the listeners that speech is prepared as it isuttered, being therefore perceived as capable of

transmitting its potential for innovation and trans-formation due to its aural quality. The higherprobability of affirmative VQP in stuck episodes iscoherent with the idea that they are interactionsegments during the session in which the patientpersists in his/her usual forms of understanding,behaving, or feeling, which solidifies his/her problemand halts progress towards therapeutic change(Fernndez et al., 2012; Herrera et al., 2009). In thiscase, consistency becomes apparent when we con-sider that affirmative VQP adds certainty andconviction to speech, and, therefore, is capable oftransmitting a lack of openness towardsreelaboration.

Regulatory VQP Microsequences

Regarding VQP regulation sequences, two typeswere identified: Self-regulation sequences (if the tem-poral association between the VQPs takes place inthe same patient or therapist utterance) and mutualregulation sequences (if the temporal associationbetween VQPs corresponds to the interactionsbetween the members of the therapeutic dyad).However, this distinction is more complicated toobserve when the sequences include full pauses andoverlapping, since these conversational phenomenacan be interpreted as regulation strategies withoneself and with somebody else, without necessarilybeing linked to utterances spoke by one or bothmembers of the dyad. Thus, the VQP sequencecomposed of the overlapping category temporallyassociated with affirmative VQP, both in the therap-ist, is classified as an instance of self-regulation, butthe presence of overlapping as given behavior mayalso suggest its identification as regulation with thepatient, for example, in the case of turn-taking.

Another way of distinguishing the VQP sequencesobserved, regardless of the regulation process ascrib-able to them, is to focus on the similarities ordifferences between the specific VQPs that theyinvolve. Hence, it is possible to distinguish symmet-rical sequences, in which similar VQPs are temporallyassociated, and asymmetrical sequences, in whichdifferent VQPs are temporally associated. Given thecoding rules in which a different VQP is coded in thesame speaking turn only when vocal quality changes,we only encountered symmetrical sequences inpatient-therapist mutual regulation VQP sequences.In contrast, it is possible to identify asymmetricalVQP sequences both in self-regulation as well as inmutual regulation sequences.

Bearing these distinctions in mind, some hypo-theses can be advanced about the associationbetween the different VQP sequences observed inthe data and the type of episode. Thus, that only

The sound of change 11

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

patients self-regulation VQP sequences and notthose of therapists are associated with the type ofepisode may be due to the fact that the lattercorrespond to a speech that characterizes a generalform of therapeutic intervention, whereas the formerare linked to a self-regulatory process in a patientopen to change.

Similarly, considering the two symmetrical mutualregulation VQP sequences, it is interesting to note thatthe sequence connected VQP in the case of the patientand connected VQP for the therapist are associated withchange episodes, and that the opposite occurs with thesequence affirmative VQP for the therapist and affirm-ative VQP for the patient. This situation could beexplained through the interaction effect of two vari-ables present in the makeup of these sequences. A firstvariable might correspond to the specific VQPs thatconstitute these symmetrical sequences. Descriptiveresults have shown that connected VQP is associatedwith change episodes and that affirmative VQP islinked to stuck episodes. We observed that in four ofthe five mutual regulation VQP sequences that includethe affirmative VQP (as given or target behavior) theyappear more often during stuck episodes, and that thesequence that breaks this pattern presents connectedVQP for the therapist.

A second variable to consider is the actor whoinitiates the sequence. In the results, it is noteworthythat all the mutual regulation VQP sequences inwhich the given behavior is a therapist VQP appearmore frequently in stuck episodes.

Therefore, both variables (specific VQP and actorwho initiates the sequence), through their interac-tion, may help explain the fact that two symmetricalVQP sequences are associated in opposite ways withrespect to the type of episode. That is, in addition tothe regulatory nature that the VQPs may have bythemselves, these results show the importance ofmutual regulation if it is the patient or the therapistwho is using a particular VQP and at what point ofthe therapeutic dialogue.

In sum, as was our purpose, we were able toobserve the process of change embodied in theexpressive vocal styles of the participants, and touncover regulatory sequences between them. Theresults consistently show that it is possible to detectthe emergence of regulatory patterns in therapeuticinteractionin this case, in the form of the vocalexpression of the participantsand that these pat-terns are involved in the process of change inpsychotherapy. Nevertheless, this study has somelimitations, mainly therapy type homogeneity (all thepsychotherapies studied were psychodynamicallyoriented) and sample size (six therapies). Forexample, one could ask whether therapists fromother than psychodynamic approaches would show

different VQPs. Another limitation has to do withthe code procedure of the VQPs, which did notcompletely ensure the independence of the observa-tions, because the same raters coded the vocalquality of both therapists and patients.

For these reasons, the findings presented must becarefully interpreted when weighing their externalvalidity. On the other hand, there might be alegitimate question regarding independence betweenthe coding of sound of the voice and its verbalcontent. We agree that its difficult for humans toseparate the two dimensions of communication,both for the one who speaks, and for the one wholistens. Moreover, it is unquestionable that in theproduction of speech sounds there is a modulationgiven by the content. Usually in every conversation acorrelation between both dimensions is expected, somuch so that when they appear to be in dissonancethe existence and the degree of independencebecome evident. This distinction between the soundof speech and its content could, for example, allowthe therapist to mark significant moments in thedialogue with the patient. In this study, coders weretrained to focus on the sound dimension of speech,making a conscious effort to leave the contentbetween parentheses.

With this study, the groundbreaking ideas of Riceand her collaborators are confirmed as a current anduseful way to explore not only therapeutic interac-tion but also the manifestations of the self in thevocal expression of the participants. In 1967, LauraRice and Alice Wagstaff stated that their system wasfocused on the attempts of patients to expressthemselves, and that such attempts exerted aninfluence on the therapist beyond the contents ofcommunication. The results presented not onlyrevisited these notions to understand how vocalexpression participates in the patient-therapist rela-tionship and in the regulation between its membersduring the psychotherapeutic process, but also con-firmed and expanded them.

Funding

This manuscript has been developed with support ofthe National Council for Science and Technology ofChilean Government, Project Fondecyt N1110361,and the Millenium Scientific Initiative of the Minis-try of Economy, Development and Tourism [grantnumber NS100018].

Note1 Intersubjective validation is a process in which the observationsby a researcher or rater are compared with the independentobservations of other researchers or raters. The validation of

12 A. Tomicic et al.

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

observations is attained through consensus or agreementbetween these different perspectives (see Flick, 2009).

References

Altimir, C., Krause, M., de la Parra, G., Dagnino, P., Tomicic, A.,Valds, N., et al. (2010). Clients, therapists and observersperspectives on moments and contents of therapeutic change.Psychotherapy Research, 20, 472487. http://dx.doi.org/10.1080/10503301003705871.

Andersen, P. (1999). Nonverbal communication: Forms and fuctions.Mountain View, CA: Mayfield.

Bachorowski, J., & Owren, M. J. (1995). Vocal expression ofemotion: Acoustic properties of speech are associated withemotional intensity and context. Psychological Science, 6, 4,219224. http://dx.doi.org/10.1111/j.1467-9280.1995.tb00596.x.

Bachorowski, J., & Owren, M. J. (2008). Vocal expression ofemotion. In M. Lewis, J. M. Haviland-Jones, & L. Feldman(Eds.), The handbook of emotions (pp. 196210). New York:Guilford Press.

Bachorowski, J., & Owren, M. J. (2009). Emotion in speech.In R. Larry (Editor-in-Chief), Encyclopedia of Neuroscience(pp. 897901). Oxford: Academic Press.

Bady, S. (1985). The voice as a curative factor in psychotherapy.Psychoanalytic Review, 72, 479490.

Bakeman, R., & Quera, V. (1995). Analyzing interaction. Sequentialanalysis with SDIS and GSEQ. New York: Cambridge Univer-sity Press.

Bakeman, R., Adamson, L. B., & Strisik, P. (1995). Lags andlogs: Statistical approaches to interaction (SPSS version). InJ. M. Gottman (Ed.), The analysis of change (pp. 279308).Mahwah, NJ: Erlbaum.

Bakeman, R., Deckner, D. F., & Quera, V. (2005). Analysis ofbehavioral streams. In D. M. Teti (Ed.), Handbook of researchmethods in developmental psychology (pp. 394420). Oxford:Blackwell.

Bakeman, R., Quera, V., McArthur, D., & Robinson, B. F.(1997). Detecting sequential patterns and determiningtheir reliability with fallible observers. Psychological Methods, 2,357370. http://dx.doi.org/10.1037/1082-989X.2.4.357.

Bnzinger, T., & Scherer, K. (2005). The role of intonation inemotional expressions. Speech Communication, 46, 252267.doi:10.1016/j.specom.2005.02.016.

Bauer, S., Tomicic, A., Martnez, C., Reinoso, A., Schfer, H.,Peukert, A., & Frangkouli, A. (2010). Die Bedeutung vonStimme und StimmklangimpsychotherapeutischenProzessausder Sicht der Patienten und Patientinnen [The meaning ofthe voice and the tone of voice for the psychotherapeuticprocess from the patients point of view]. Psychoanalyse &Krper, 17, 2750.

Beebe, B. (2006). Co-constructing mother-infant distress inface-to-face interactions: Contributions of microanalysis. Infant Observa-tions, 9, 151164. http://dx.doi.org/10.1080/13698030600810409.

Beebe, B., & Lachmann, F. (1998). Co-constructing inner andrelational processes: Self and mutual regulation in infant researchand adult treatment. Psychoanalytic Psychology, 15, 480516.http://dx.doi.org/10.1037//0736-9735.15.4.480.

Campanelli, L., Iberni, E., Sarracino, D., Degni, S., & Mariani,R. (2007). Semiotic of the non verbal vocal expression ofemotions and research into the psychotherapy process: A pilotstudy. Revista di Psicologia Clinica, 1, 102115.

Campbell, N. (2004a). Listening between the lines: A study ofparalinguistic information carried by tone-of-voice. InternationalSymposium on Tonal Aspects of Languages, 2831. Available athttp://www.isca-speech.org/archive/tal2004/tal4_013.pdf.

Campbell, N. (2004b). Accounting for voice-quality variation.Speech Prosody, 2326.

Campbell, N. (2007). On the use of non verbal speech sounds inhuman communication. Verbal and Nonverbal CommunicationBehaviours, LNAI4775, 117128.

Cohen, J. (1968). Weighed kappa: Nominal scale agreementwith provision for scaled disagreement or partial credit.Psychological Bulletin, 70, 213220. http://dx.doi.org/10.1037/h0026256.

Elliot, R. (1984). A discovery-oriented approach to significantchange events in psychotherapy: Interpersonal process recall andcomprehensive process analysis. In L. Rice & L. S. Greenberg(Eds.), Patterns of change: Intensive analysis of psychotherapy process(pp. 249286). New York: Guilford.

Fernndez, O., Herrera, P., Krause, M., Prez, J. C., Valds, N.,Vilches, O., & Tomicic, A. (2012). Episodios de cambio yestancamiento en psicoterapia: Caractersticas de la comunica-cin verbal entre pacientes y terapeutas [Change and stuckepisodes in psychotherapy: Characterization of verbal commun-ication between patients and therapists]. Terapia Psicolgica, 30,522. http://dx.doi.org/10.4067/S0718-48082012000200001.

Flick, U. (2009). An introduction to qualitative research. Lon-don: Sage.

Gobl, C., & Chasaide, A. (2003). The role of voice quality incommunicating emotion, mood and attitude. Speech Commun-ication, 40, 189212. http://dx.doi.org/10.1016/S0167-6393(02)00082-1.

Gonalves, M., Matos, M., & Santos, A. (2009). Narrativetherapy and the nature of innovative moments in theconstruction of change. Journal of Constructivist Psychology, 22,123. http://dx.doi.org/10.1080/10720530802500748.

Herrera, P., Fernndez, O., Krause, M., Vilches, O., Valds, N.,& Dagnino, P. (2009). Revisin terica y metodolgica de lasdificultades en psicoterapia. Propuesta de un modelo ordena-dor [Theoretical and methodological review of difficulties inpsychotherapy. Proposing an organizer model]. Terapia Psico-lgica, 27, 169179.

Jaffe, J., Beebe, B., Feldstein, S., Crown, C. L., & Jasnow, M. D.(2001). Rhythms of dialogue in infancy. Monographs of theSociety for Research in Child Development, Serial No. 265, 66, 2.

Knapp, M. (1988). La comunicacin no verbal [Non-verbal com-munication]. Barcelona: Editorial Paids.

Knoblauch, S. (2005). Body rhythms and the unconscious: Towardan expanding of clinical attention. Psychoanalytic Dialogues, 15,807827. http://dx.doi.org/10.2513/s10481885pd1506_2.

Knoblauch, S. H. (2000). The musical edge of therapeutic dialogue.Hillsdale, NJ: Analytic Press.

Krause, M. (2005). Psicoterapia y cambio. Una mirada desde laSubjetividad [Psychotherapy and change. A view from the subject-ivity]. Santiago: Ediciones Universidad Catlica de Chile.

Krause, M., de la Parra, G., Arstegui, R., & Strasser, K. (2006).Hacia una prctica efectiva en psicoterapia: estudio de episodios decambio relevantes en diferentes tipos de psicoterapia y sus efectos enlos resultados teraputicos [Towards an effective practice in psycho-therapy: Study of relevant change episodes in different types ofpsychotherapies and its effects on treatment outcomes] (ProyectoFONDECYT N1030482). Santiago, Chile: Pontificia Uni-versidad Catlica de Chile, Escuela de Psicologa.

Krause,M., de la Parra,G., Arstegui, R.,Dagnino, P., Tomicic, A.,Valds, N., et al. (2007). The evolution of therapeutic changestudied through generic indicators. Psychotherapy Research, 17, 6,673689. http://dx.doi.org/10.1080/10503300601158814.

Kucharska-Pietura, K., Nikolaou, V., Masiak, M., & Treasure, J.(2004). The recongnition of emotion in the faces and voice ofAnorexia Nervosa. Eating Disorders, 35, 4247. http://dx.doi.org/10.1002/eat.10219.

Lambert, M., & Burlingame, G. L. (1996). Outcome Questionnaire(OQ-45.2). Washington, DC: American Professional Creden-tialing Services.

The sound of change 13

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

http://dx.doi.org/10.1080/10503301003705871http://dx.doi.org/10.1080/10503301003705871http://dx.doi.org/10.1111/j.1467-9280.1995.tb00596.xhttp://dx.doi.org/10.1037/1082-989X.2.4.357http://dx.doi.org/10.1016/j.specom.2005.02.016http://dx.doi.org/10.1080/13698030600810409http://dx.doi.org/10.1037//0736-9735.15.4.480http://www.isca-speech.org/archive/tal2004/tal4_013.pdfhttp://dx.doi.org/10.1037/h0026256http://dx.doi.org/10.1037/h0026256http://dx.doi.org/10.4067/S0718-48082012000200001http://dx.doi.org/10.1016/S0167-6393(02)00082-1http://dx.doi.org/10.1016/S0167-6393(02)00082-1http://dx.doi.org/10.1080/10720530802500748http://dx.doi.org/10.2513/s10481885pd1506_2http://dx.doi.org/10.1080/10503300601158814http://dx.doi.org/10.1002/eat.10219http://dx.doi.org/10.1002/eat.10219

Malloch, S. (1999). Mother and infants and communicativemusicality. In I. Delige (Ed.), Rhythms, musical narrative, andthe origins of human communication. Musicae scientiae, specialissue, 19992000 (pp. 2957). Lige: European Society for theCognitive Sciences of Music.

Martinez, E. (2003). El sonido en la comunicacin humana:Introduccin a la fontica [The sound in human communication:Introduction to phonetics]. Barcelona: Editorial Octaedro.

Moneta, M. E., Penna, M., Loyola, H., Buchheim, A., & Kchele,H. (2008). Measuring emotion in the voice during psychotherapyinterventions: A pilot study. Biological Research, 41, 389395.http://dx.doi.org/10.4067/S0716-97602008000400004.

Nespor, M., & Vogel, I. (1994). La prosodia [The prosody].Madrid: Visor Distribuidores.

Paul, A., & Liker, J. (1982). Analyzing sequential categorical dataon dyadic interaction. Psychological Bulletin, 91, 393403.doi:10.1037/0033-2909.91.2.393.

Papouek, M. (2007). Communication in early infancy: An arenaof intersubjective learning. Infant behavior & development, 30,258266. http://dx.doi.org/10.1016/j.infbeh.2007.02.003.

Pell, M. D., Monetta, L., Paulmann, S., & Kotz, S. A. (2009).Recognizing emotions in a foreign language. Journal of Non-verbal Behavior, 33, 107120. http://dx.doi.org/10.1007/s10919-008-0065-7.

Rice, L., & Wagstaff, A. (1967). Client voice quality andexpressive style as indexes of productive psychotherapy. Journalof Consulting Psychology, 31, 557563. http://dx.doi.org/10.1037/h0025164. doi:10.1037/h0025164.

Rice, L. N., & Kerr, G. P. (1986). Measures of client andtherapist vocal quality. In L. S. Greenberg & W. M. Pinsof(Eds.), The psychotherapeutic process: A research handbook (pp.73105). New York: Guilford.

Russel, J. A., Bachorowski, J., & Fernandez-Dol, J. (2003).Facial and vocal expressions of emotion. Annual Review ofPsychology, 54, 329349. http://dx.doi.org/10.1146/annurev.psych.54.101601.145102.

Safran, J. D., & Muran, J. C. (1996). The resolution of ruptures inthe therapeutic alliance. Journal of Consulting and Clinical Psycho-logy, 64, 447458. http://dx.doi.org/10.1037/0022-006X.64.3.447.

Safran, J. D., & Muran, J. C. (2000). Negotiating the therapeuticalliance: A relational treatment guide. New York: Guilford Press.

Safran, J. D., & Muran, J. C. (2006). Has the concept of thealliance outlived its usefulness? Psychotherapy, 43, 286291.http://dx.doi.org/10.1037/0033-3204.43.3.286.

Sapaly, J. (2005). Psicoacstica [Psychoacoustics]. In M. Jackson-Menaldi (Ed.), La voz normal [The normal voice] (pp. 47114).Buenos Aires: Editorial Mdica Panamericana.

Sarko, D., Roth, H., & Martin, E. (2010). Prosody-preservingvoice transformation to evaluate brain representations of speechsounds. IEEE Transactions on Audio, Speech and LanguageProcessing, 18, 10171029. doi:10.1109/TASL.2009.2035165.

Scherer, K. R. (1982). Methods of research on vocal communica-tion: Paradigms and parameters. In K. R. Scherer & P. Ekman(Eds.), Handbook of methods in nonverbal behavior research (pp.136198). Cambridge: Cambridge University Press.

Scherer, K. R., & Bergmann, G. (1990). Vocal communication.German Journal of Psychology, 8, 5790. http://dx.doi.org/10.1037/0033-2909.99.2.143.

Scherer, K. R., Banse, R., & Wallbott, H. G. (2001). Emotioninferences from vocal expression correlate across languages andcultures. Journal of Cross-Cultural Psychology, 32, 7692. http://dx.doi.org/10.1177/0022022101032001009.

Stern, D. N. (1998). The process of therapeutic change involvingimplicit knowledge: Some implications of developmental obser-vations for adult psychotherapy. Infant Mental Health Jour-nal, 19, 300308. http://dx.doi.org/10.1002/(SICI)1097-0355(199823)19:33.0.CO;2-P.

Thompson, W. F., & Balkwill, L. L. (2006). Decoding speechprosody in five languages. Semiotica, 158, 407424. http://dx.doi.org/10.1515/SEM.2006.017.

Timulak, L. (2007). Identifying core categories of clients-identifiedimpact of helpful events in psychotherapy: A qualitative meta-analysis. Psychotherapy Research, 17, 305314. http://dx.doi.org/10.1080/10503300600608116.

Tomicic, A., Bauer, S., Martnez, C., Reinoso, A., & Guzmn, M.(2009). La Voz como una herramienta psicoteraputica: Laperspectiva de los terapeutas [The voice as psychotherapeutictool: The psychotherapists view]. Revista Argentina de ClnicaPsicolgica, 18, 197208.

Tomicic, A., Chacn, D., Martnez, C., Guzmn, M., Bauer, S., &Reinoso, A. (2009). Sistema de codificacin de patrones de cualidadvocal de paciente y terapeuta (PCV-1.0): manual de entrenamiento yprocedimiento para la codificacin de patrones de cualidad vocal[Coding system of vocal quality pattern of patient and therapist(PCV-1.0): Training manual and coding process for vocal qualitypatterns]. Unpublished manuscript.

Tomicic, A., Martnez, C., Chacn, L., Guzmn, M., & Reinoso,A. (2011). Patrones de Cualidad Vocal en Psicoterapia: Desar-rollo y estudio de confiabilidad de un sistema de codificacin.Vocal Quality Patterns in Psychotherapy: Coding System Devel-opment and Reliability Analysis. Revista Psykh, 20, 1, 7793.http://dx.doi.org/10.4067/S0718-22282011000100006

Tronick, E. (1989). Emotions and emotional communications ininfants. American Psychologist, 44, 112119. http://dx.doi.org/10.1037/0003-066X.44.2.112.

Tronick, E. (1998). Dyadically expanded states of consciousnessand the process of therapeutic change. Infant Mental HealthJournal, 19, 290299. http://dx.doi.org/10.1002/(SICI)1097-0355(199823)19:33.0.CO;2-Q.

von Bergen, A., & de la Parra, G. (2002). OQ-45.2, cuestionariopara la evaluacin de resultados y evolucin en psicoterapia:Adaptacin, validacin e indicaciones para su aplicacin einterpretacin [OQ-45.2, questionnaire for the assessment ofoutcome and evolution in psychotherapy: Adaptation, valida-tion and guidelines for its application and interpretation].Terapia Psicolgica, 20, 161176.

Wiseman, H., & Rice, L. (1989). Sequential analyses of therapist-client interaction during change events: A task-focusedapproach. Journal of Consulting and Clinical Psychology, 2,281286. http://dx.doi.org/10.1037/0022-006X.57.2.281.

14 A. Tomicic et al.

Dow

nloa

ded

by [

Swin

burn

e U

nive

rsity

of

Tec

hnol

ogy]

at 1

7:51

24

Aug

ust 2

014

http://dx.doi.org/10.4067/S0716-97602008000400004http://dx.doi.org/10.1037/0033-2909.91.2.393http://dx.doi.org/10.1016/j.infbeh.2007.02.003http://dx.doi.org/10.1007/s10919-008-0065-7http://dx.doi.org/10.1007/s10919-008-0065-7http://dx.doi.org/10.1037/h0025164http://dx.doi.org/10.1037/h0025164http://dx.doi.org/10.1146/annurev.psych.54.101601.145102http://dx.doi.org/10.1146/annurev.psych.54.101601.145102http://dx.doi.org/10.1037/0022-006X.64.3.447http://dx.doi.org/10.1037/0033-3204.43.3.286http://dx.doi.org/10.1109/TASL.2009.2035165http://dx.doi.org/10.1037/0033-2909.99.2.143http://dx.doi.org/10.1037/0033-2909.99.2.143http://dx.doi.org/10.1177/0022022101032001009http://dx.doi.org/10.1177/0022022101032001009http://dx.doi.org/10.1002/(SICI)1097-0355(199823)19:33.0.CO;2-Phttp://dx.doi.org/10.1002/(SICI)1097-0355(199823)19:33.0.CO;2-Phttp://dx.doi.org/10.1515/SEM.2006.017http://dx.doi.org/10.1515/SEM.2006.017http://dx.doi.org/10.1080/10503300600608116http://dx.doi.org/10.1080/10503300600608116http://dx.doi.org/10.4067/S0718-22282011000100006http://dx.doi.org/10.1037/0003-066X.44.2.112http://dx.doi.org/10.1037/0003-066X.44.2.112http://dx.doi.org/10.1002/(SICI)1097-0355(199823)19:33.0.CO;2-Qhttp://dx.doi.org/10.1002/(SICI)1097-0355(199823)19:33.0.CO;2-Qhttp://dx.doi.org/10.1037/0022-006X.57.2.281

AbstractThe Relevance of the Voice in Human CommunicationThe Voice in the Psychotherapeutic DialogThe Study of the Voice in Psychotherapeutic InteractionLaura Rice's Ideas RevisitedThe Study of Interaction in Psychotherapy through Relevant EpisodesMethodOutline placeholderParticipantsData analysis

ResultsDiscussionVQP Differences Between Patients and TherapistsVQP Differences Between Stuck and Change EpisodesRegulatory VQP Microsequences

FundingNoteReferences