00 izdebski vol2 i-xxiifm - plural publishing the estill voice model: physiology of emotion 83...

v

CONTENTS

FOREWORD by Hans von Leden viiPREFACE ixACKNOWLEDGMENTS xiiiCONTRIBUTORS xvINTRODUCTION xix

1 Autonomic Determinants of Vocal Expression of Emotions 1Branka Zei Pollermann

2 Brain Mechanisms for Social Perception 17Elizabeth J. Carter and Kevin A. Pelphrey

3 Child Speech and Emotions: A Cross-Linguistic Perspective 35Ioulia Grichkovtsova and Ineke Mennen

4 Conveyance of Emotion in Postlaryngectomy Communication 49Philip C. Doyle

5 Impact of Cleft Lip and Palate (CLP) and VelopharyngealInsufficiency (VPI) on Emotive Aspects of Voice and Speech 69Maria A. Hortis-Dzierzbicka and Krzysztof Izdebski

6 The Estill Voice Model: Physiology of Emotion 83Kimberly M. Steinhauer and Jo Estill

7 Neurology and Clinical Considerations of Affective Prosody 101Marilee Monnot, Diana Orbelo, Julie Testa, and Elliott Ross

8 Pathologic Phonation Connotes and Evokes Wrong Emotive Reactions in Listeners 119Krzysztof Izdebski

9 Perception of Emotion in Impaired Facial and Vocal Expression 129Ingrid Verduyckt and Dominique Morsomme

00_Izdebski_Vol2_i-xxiiFM 3/5/08 5:22 PM Page v

10 The Music of Language: The Importance of Prosody and Non-Verbal Communication in the Dyadic Interaction between Infant and Caregiver 145Victoria Stevens

11 Psychobiological Framework of Stress and Voice: A Psychobiological Framework for Studying Psychological Stress and Its Relation to Voice Disorders 159Maria Dietrich and Katherine Verdolini Abbott

12 The Role of Auditory Feedback in Nonverbal Vocal Behavior of Normally Hearing and Hearing Impaired Infants 179Elisabeth Scheiner and Kurt Hammerschmidt

13 The Vocal Dance of Seduction: From Being Anyone to Being Someone 199Luigi Anolli

14 Vocal Indicators of Coping Style in Patients with Breast Cancer: A Pilot Study 215Branka Zei Pollermann and Jürg Bernhard

15 Why Are We Attracted to Certain Voices? Voice as an Evolved Medium for the Transmission of Psychologicaland Biological Information 231Susan M. Hughes and Gordon G. Gallup, Jr.

16 From Perception to Communication: Sensolinguistic Therapy to Restore Voice and Vocal Emotions 239Lilla-Teresa Sadowski, Alois Mauerhofer, Sabine Hofmann, andKrzysztof Izdebski

17 Restoration of Vocal and Facial Emotive Production in Tardive Dyskinesia, Cerebral Palsy, and Closed Head InjuryUsing Selective Chemical Denervation 245Krzysztof Izdebski and Raul M. Cruz

18 A Comprehensive Model of How the Stress Chain Affects Voice 253Willy A. R. Wellens and Magda J. M. C. van Opstal

19 Vocal Psychodynamics in the Voice Clinic 273Mara Behlau and Gisele Gasparini

INDEX 283

vi VOICE AND EMOTION: VOLUME 2

00_Izdebski_Vol2_i-xxiiFM 3/5/08 5:22 PM Page vi

AUTONOMIC DETERMINANTS OF VOCAL EXPRESSION OF EMOTIONS 13

Conclusions

Our results support the main hypothesisthat diminished autonomic response isassociated with emotional blunting.Theyalso point to a major role played by theparasympathetic branch of the ANS. Morespecifically, patients with low parasym-pathetic tone feel lower intensity of angerand sadness, and display blunting ofvocally expressed emotional arousal.Thesympathetic branch of the ANS played arole only in the spectral differentiation ofemotional valence (anger vs. joy).

To the extent that an individual’s auto-nomic tone is inherent to his or her per-sonality, one could expect autonomiclesions to impact some of the prominentpersonality traits. In a recent study,Schweiger and colleges found that indi-viduals with high vagal tone are lessinhibited and do not try to deny feelings(Schweiger, Witling, Genzel, & Block,1998). It thus appears that our findingrelating a disease induced change in thevagal tone to the awareness of change in emotionality, may have psychologicalconsequences for the patients’ personal-ity. The assumption is based on the factthat emotion experiences provide conti-nuity in the functioning of personality.Emotional feeling states that remain con-stant over time link the stability of emo-tion experience to personality (Barrett & Campos, 1987; Izard, Libero, Putnam,& Haynes, 1993). Caroll Izard and col-leagues have convincingly demonstratedhow emotions become linked to partic-ular thoughts to form affective-cognitivestructures, which in turn combine intocognition-emotion-action patterns thatcharacterize the individual’s style ofadaptive behavior and contribute to thedevelopment of personality.

We can thus expect subjects withimpaired cardiovascular reactions to havealtered emotional experiences with sub-sequent consequences for the regulationof the organism’s interaction with physi-cal and social environment. Indeed, vocalsignaling of emotional states serves animportant function of regulating socialinteractions. It not only provides infor-mation about the speaker’s attitudes andbehavioral tendencies, but it is also inform-ative about the speaker’s acceptance ofsocial and cultural demands and expecta-tions (Gregory, Webster, & Huang, 1993;Rosenfeld, 1987). Altered vocal signalingof emotions may therefore impact theperson’s social functioning.

To conclude, we believe that damageto the nerves in the cardiovascular systemnot only affects the autonomic respon-sivity, its flexibility, and its balance, but italso induces a CNS-ANS dysregulation(Friedman & Thayer, 1998) resulting in adysregulation of the organism’s adaptivebiological and social response to environ-mental challenges.

Acknowledgments. I extend specialthanks to Professor Krzysztof Izdebskifor his constructive remarks in the elab-oration of the manuscript, to Dr. YvesMorel for his administration of the auto-nomic function tests, and to VéroniqueUllmann for her work on qualitative assess-ment of the patients’ verbal comments.

References

Banse, R., & Scherer, K. R. (1996). Acousticprofiles in vocal emotion expression. Jour-nal of Personality & Social Psychology,70(3), 614–636.

Barrett, K. C., & Campos, J. J. (1987). Per-spectives on emotional development. ii:

01_Izdebski_Vol2_1-16 3/5/08 5:22 PM Page 13

BRAIN MECHANISMS FOR SOCIAL PERCEPTION 27

Autism, Social Perception, andthe Superior Temporal Sulcus

Autism is a severe, pervasive develop-mental disorder characterized by deficitsin three domains: (a) social interaction,(b) communication skills, and (c) re-stricted interests and repetitive, stereo-typed behaviors [DSM-IV; American Psy-chiatric Association (APA), 1994]. Thereis a wide range of variation in abilitieswithin these domains; however, all threetypes of deficits are present to somedegree in individuals with autism. Addi-tionally, there is a large range of intellec-tual capacities: although many of theseindividuals exhibit mental retardation,others have average or even superiorabilities.

Social impairment is the primary,unique feature of autism (Kanner, 1943;Wing & Gould, 1979). Early problems insocial perception, such as joint attention,are indicators of autism (Baron-Cohen,1995; Dawson et al., 2002; Lord et al.,1997; Sigman et al., 1999).Also, these chil-dren fail to recognize biological motionin point-light displays, where lights areplaced on the joints of people in a darkroom and can be seen as they movearound (Blake, Turner, Smoski, Pozdol,& Stone, 2003; Klin, Jones, Schultz, &Volkmar, 2003). Additionally, individualswith autism have a hard time identifyingthe emotions portrayed by faces in photographs, particularly fear and anger(Adolphs, Sears, & Piven, 2001; Pelphreyet al., 2002).

In order to explore emotional recogni-tion deficits in individuals with autism,we quantified visual attention to faces(Pelphrey et al., 2002). In an eye-trackingstudy of adults looking at photographs of

faces, we found that the adults withautism spent more time looking at thenonfeature portions of the face (e.g.,forehead, cheeks) than did typical adults.Additionally, there was a correspondingdecrease in the visual examination of corefeatures (i.e., the eyes, nose, and mouth)by the individuals with autism relative tothe individuals without autism. This sug-gested disorganized face processing andcould be related to their overarchingsocial perception problems. Similar resultswere reported in another study of scan-paths over faces by individuals withautism by Klin and colleagues (2002).

An impaired understanding of theimplications of gaze direction is also char-acteristic of autism. Baron-Cohen (1995)showed pictures of a cartoon face namedCharlie looking at one of four types ofcandy to children with typical develop-ment, mental retardation, or autism.When asked, “What is Charlie’s favoritecandy?” the typically developing childrenand those with mental retardation cor-rectly responded with the name of thecandy at which Charlie was looking.However, the children with autism failedto do this. However, when shown a different set of pictures with faces look-ing towards or away from them, all of the children could correctly answer thequestion, “Which one is looking at you?”Together, these results suggest that chil-dren with autism can perceive gazedirection, but they do not link it with itssocial and mental significance.

These deficits, combined with our priorfindings on gaze perception in the supe-rior temporal sulcus region, suggestedthat dysfunction in this area could relateto autism (Pelphrey, Morris, & McCarthy,2005). In order to test this hypothesis,we used fMRI and our congruent versus


Introduction

The area of affective speech research hasreceived much attention over the lastyears. Nevertheless, the effects of indi-vidual and cultural differences on affec-tive speech production are still largelyneglected. Most cross-linguistic studiesconcern the perception of vocal emo-tions (Scherer, Banse, & Wallbott, 2001).They show that humans can recognizevocal emotions with an accuracy percent-age well above chance, even in unknownlanguages, though the level of accuracyis higher in native and close to native lan-guages. The research suggests that vocalrealization of emotions is based both onuniversal mechanisms and cultural con-ventions. A large number of acousticparameters were identified as possiblemeans to express emotions in speech:pitch range, intonational patterns, voicequality, rhythm, and so on. At the sametime, the usage of these means, theirlevel of importance, and their meaningvary across languages (Abelin & All-wood, 2000; Chen, 2005).

Studies of cross-linguistic productionare still quite rare (Boula de Mareuil,Célérier, & Toen, 2002). They usuallyinvolve a very small number of speakersper language and disregard a widely ob-served phenomenon in affective speech—a high interspeaker and intraspeakervariability. Individual speakers have alarge number of possible acoustic meansto express emotions and a high flexibil-ity of choice and manner of usage thatcan be considered appropriate in theirlanguage and speech community. Theexisting research is primarily concernedwith adults, and studies of child affectivespeech are very rare and mostly based onperception only (Clément, 1999; Van derMeulen, Janssen, & Den Os, 1997).

36 VOICE AND EMOTION: VOLUME 2

The main objective of this research isto investigate the production of affectivespeech by bilingual (raised in two lan-guages from birth) and monolingual chil-dren cross-linguistically in Scottish Englishand French. Socio-emotional competenceis essential for effective communication,and it is one of the priorities in childdevelopment. Bilinguals communicatetheir emotions in two different languages,and they have to use acoustic meansrelated to affective speech appropriately.

Being raised in two different languagesand cultures, bilinguals may be compe-tent to express affective states in theirtwo languages. So, our first hypothesis isthat bilingual affective productions mayfall in the acceptable range in each oftheir languages. On the other hand, bilin-guals produce two languages that differin some acoustic correlates for affectivestates. Bilinguals may have a wider rangeof means and ways to express emotions,and their affective realization may notstrictly correspond to those of monolin-guals in each of their two languages.Thisphenomenon has been attested to in bi-lingual studies of other phonetic aspects(Aoyama, Flege, Guion,Akahane-Yamada,& Yamada, 2004; Mennen, 2004). Thus,our second hypothesis is that bilingualsmay represent a particular group of speak-ers who express vocal emotions in a dif-ferent manner than monolinguals.

Methodology

The starting point for this study was todevelop an appropriate methodology to record child affective speech. Thoughspontaneous speech is claimed to containthe most authentic emotions, its collec-tion gives rise to numerous difficulties:speakers should be recorded without


CONVEYANCE OF EMOTION IN POSTLARYNGECTOMY COMMUNICATION 61

how, or perhaps better stated where,speech occurs in the respiratory cycledoes arise. In fact, clinical experiencehas shown that users of an EL, regardlessof whether it be a neck-type or intraoraldevice, often suspend respiration alto-gether at points of speech output. Thisstrategy is at times used to discouragethe speaker from having respiratorynoise compete with the EL signal. Thissame problem of “stoma noise” has beenreported to be a real and significantimpediment to successful acquisition ofesophageal speech in that stoma noise orstoma blast secondary to excessive expi-ratory force may mask the less intenseesophageal speech signal (Doyle, 1994).

In summary, three methods of post-laryngectomy speech rehabilitation exist,each with inherent advantages and disad-vantages. Because of the disconnectionbetween the primary airway and thestructures comprising the vocal tract,secondary limitations specific to the res-piratory system do emerge. The conse-quences of this disruption clearly havean effect on respiratory patterns, whichmay in turn create limitations in the con-veyance of emotion following total laryn-gectomy. The essence of this disruptionis seen primarily in alterations in the synchrony with which respiration andsound production may be achieved.However, the larger issue relative to theexpression of emotional states in thepresent context is that the respiratorysystem is no longer linked to sound gen-erating structures. When coupled withalterations in respiratory phase, the com-bined influence of these factors on emo-tional expression are substantial. Further

information concerning associated prob-lems that arise is presented in the subse-quent section.

Problems Conveying Emotion:Postlaryngectomy

Limitations and Restrictions

Evolving from our knowledge of the nor-mal and postlaryngectomy speech produc-tion system, and regardless of alaryngealmode, let’s now consider two commonacts of emotional expression as theymight exist on a continuum, laughingand crying. Although both can be donevolitionally and sometimes without true,internalized, and spontaneous emotionalinvolvement (e.g., consider the polishedactor), they more regularly exist as invol-untary acts of emotional expression inresponse to a particular event or stimulus.

Laughing in Those Who AreLaryngectomized

The more pleasant of these two emotions,laughing, requires considerable integra-tion of several physical structures.8 Therespiratory system in coordination withthe abdomen and thoracic wall con-tribute greatly to the act of laughing.Thisinvolves active participation of a numberof muscles in order to control move-ments associated with laughing. In thenormal system, respiratory airflow ismodulated through the upper airwayand larynx, prior to moving into thevocal tract. In contrast, however, in those

8From a physiologic perspective, laughing and crying may share many common characteristics relative to airway and respiratory control mechanisms.

04_Izdebski_Vol2_49-68 3/5/08 5:23 PM 1

who are laryngectomized, while muscu-lar movements associated with the act oflaughing do occur, the product of thismovement terminates through an openand unrestricted tracheostoma. For thisreason, one might view the act of laugh-ing in the individual who is laryngec-tomized as a paroxysmal respiratoryevent without vocalization of any form.More simply stated, laughing for thosewho are laryngectomized is tantamountto going through the respiratory motionswithout the benefit of acoustic speechoutput observed in the normal system.A common clinical complaint regardingthis deficit is reported to be the lack ofsimple satisfaction of completing the actof laughing as it would have occurrednormally.

Clearly, laughing may be characterizedby variance relative to the intensity ofthe response and we have all observedthe continuum from a subdued, lightchuckle to the almost infectious belly-laugh. The duration of the response maybe brief or extended, dependent uponthe circumstances and how receptiveand/or appropriate the behavior is for agiven environment.

For the laryngectomized person, laugh-ing is in essence manifest as solely aresponsive change in breathing dynamics—short, quick inspiratory and expiratorymovements secondary to adjustments inthe abdominal and thoracic walls. Simi-larly, the disconnection from the abilityto generate sound in response to respira-tory actions is what creates the greatestlevel of loss for these individuals basedon direct report from those who are lar-yngectomized. Consider laughing without


the ability to make any type of vocalizednoise and one can see that the con-veyance of what might be viewed as asimple but primary human emotion isdisrupted substantially. What this sug-gests to us in the case of laughing is thatthe ultimate satisfaction from this behav-ior is found in the direct linkage to andultimate representation in some type ofvocalization.9 Shanks (1994) has recom-mended that attempts at compensatingfor the loss of normal laughter via use ofa “body shake and thigh slapping,” inaddition to potential facilitation of thisloss of emotional expression though“stoma exhalation concomitant withesophageal phonation” (p. 215) are ofvalue. However, my own clinical inter-actions with those who are laryngec-tomized suggest that such levels of compensation are not satisfactory. Thisobservation lends further credence tothe suggestion that the loss of this abilityto convey a common behavioral activitythat is directly related to an affectivestate is reduced considerably if notentirely eliminated in the clinical popu-lation. Those who are laryngectomizedfrequently report that the loss of “nor-mal” laughter is not insignificant relativeto the nature and effectiveness of socialinteractions.

Crying in Those Who AreLaryngectomized

Despite sharing a number of featuresfrom a physiologic perspective, crying isindeed different from laughing. Similar tolaughing, crying necessitates a substan-

9One need only consider his or her group of friends and acquaintances to accept the notion that the mannerof how one laughs is indeed as unique as the human voice itself and carries considerable satisfaction to theindividual.


The Value of Phonation in CLP Population

The value of phonation in current soci-ety cannot be underestimated, and voicevalue is no longer restricted to singers,actors, preachers, or other professionalvoice users, but is of indisputable valueto many people. It is now clear that voiceand vocal emotions play an enormousrole in social relations and may often bea factor in determining the extent of suc-cess, job placement, and societal statusand even in the choice of a sexual part-ner or a mate. (For more on this subjectsee Chapters 13 and 15 in this volume.)

It is now evident that CLP voice qual-ity can frequently be a cause of socialexclusion (at all age stages) and can be adecisive factor in professional achieve-ment or employment. When present,even discreetly, voice/speech problemscan prevent the CLP patient from access-ing the job market. Some of our patientsreport that even minor or discreet hyper-nasality and vocal stridency perceived by the listener on the phone can causeinquiry to the management by clientsabout the apparent deviancy of thespeaker, or can simply result in a hang-upor aborted call or a request to speak tosomebody else. (For more on the role ofvoice quality while on the phone seeChapter 8 of this volume.)

The voice and speech distortions pres-ent in CLP patients are often causative ofreading difficulties, and are of impor-tance in school settings for all CLPpatients because of how teachers andpeers react to their oral presentations(Richman, Eliason, & Lindgren, 1988).When a child with CLP is forced to talkor to read aloud in front of classmates,which is a particularly cruel task but


which unfortunately happens rather fre-quently, this poses an enormous emo-tional burden on the child. In such a situation, a child with CLP may react by either increasing vocal efforts in anattempt to increase the chance of beingunderstood, or by not being able to uttera single word.These efforts may manifestthemselves in various types of vocal mis-use, including excessive vocal hard attack,wrong voice pitch, excessive loudness,or combinations of all of these factors.If this defense behavior is frequentlyrepeated it can create phonotrauma andform vocal fold edema, chronic laryngi-tis, vocal fatigue, muscular tension dys-phonia, or vocal cord nodules. Thosechanges in turn will affect the voice qual-ity, and the vicious circle continues.

Conclusions

The knowledge of how the voice in the CLP population hinders expressionof emotions is not clear, judging fromcurrently available literature. The datareviewed here show that vocalization in children with CLP may be tainted byintrinsic restrictions, hence providingfalse information about the emotivestates of the speaker. We attempted hereto show how important the knowledgeof phonation is in consideration of howto improve treatment to provide the bestanatomical condition for execution ofnearly normal voice and speech. Basedon our experience, we believe that earlyand properly executed palatal surgery maynot only enable speech and languagedevelopment under desirable anatomicaland functional conditions, but can alsoprevent grave psychosocial maladjust-ment with all its consequences on the


NEUROLOGY AND CLINICAL CONSIDERATIONS OF AFFECTIVE PROSODY 109

cally to the control group on vocabulary,abstract reasoning, and basic cognitivefunctions, as measured by the ShipleyInstitute of Living Scales. This informa-tion, along with publications showingthat emotional facial expression recogni-tion is also deficient in detoxified alco-holics, indicates that emotion cognitionis damaged by early alcohol abuse (Korn-reich et al., 1998; Kornreich et al., 2002;Kornreich, Blairy, Philippot, Dan, et al.,2001; Kornreich, Blairy, Philippot, Hess,et al., 2001).

In addition, alcoholics also appear tobe deficient in vocal production of affect(Monnot, Orbelo, et al., 2001). Four ex-perienced raters listened to the vocalproduction of the 82 subjects in the alco-hol abuse study (described above) andidentified what emotion was projectedusing a forced choice of the six emotionsrepresented in the Aprosodia Battery.Their inter-rater reliability correlationwas 0.82. Results showed detoxifiedalcoholics were significantly impaired intheir ability to produce understandableand accurate emotion in their speech [F (2,54) = 7.0, p = .0036]. In addition,these pragmatic ratings of subjects’ abil-ity to inject emotion into the voice showa positive, significant relationship toacoustic analyses assessing fundamentalfrequency variability [F (2,54) = 24.58,R2Adjusted = 0.52, p < .0001]. A regres-sion analysis showed a significant associ-ation between the Comprehension andProduction components of the Aproso-dia Battery for all 82 subjects [F (1,54) =9.27, p < .003]. Subjects’ sex, education,ethnic origin, and other drugs of abusewere not predictive of results. However,age was negatively associated with accu-racy of affective prosodic repetition [F (1,54) = 4.96, p < .03] ( Monnot, 2002;Monnot, Orbelo, et al., 2001). These

results indicate that detoxified alcoholics(and those with putative prenatal expo-sure) are impaired in their ability to pro-duce understandable affective prosodyin speech and to comprehend emotionin the voice of a communication partnerand, thus, this may explain partially thepoor social skills and altered personaland work relationships so often noted in alcoholics.

Alzheimer’s Disease and Affective Prosody

Alzheimer’s disease (AD) is a progressiveneurodegenerative disease characterizedby declines in memory and cognitiveabilities, such as language and problemsolving (McKhann, Folstein, Katzman,Price, & Stadlan, 1984). In addition tothese cognitive deficits, some patients withAD experience disturbances in behavior,mood, and emotional regulation (Cum-mings, 1996; Piccininni, Baldereschi,Zaccara, & Inzitari, 2005; Shimabukuro &Matsuoka, 2005; Zubenko & McPherson,2003). Patients with AD can have diffi-culty expressing and comprehendingemotional information through facial ex-pressions and affective prosody (Albert,1991; Allender, 1989; Bucks, 2004;Cadieux, 1997; Cohen & Brosgole, 1988;Hargrave & Stone, 2002; Lavenu, 2005;Roberts, 1996), which worsen as demen-tia severity increases, and these deficitshave been associated with mood andbehavior disturbances (Allender, 1989;Roberts, 1996).

From a pathological standpoint, affec-tive prosody deficits should reflect theunderlying distribution and severity ofneurofibrillary tangles. It is well knownthat in late-onset AD, pathological changesfirst appear in the medial and orbitofrontal


Introduction

It is commonly agreed that certain voicequalities, prosodic features, and temporalvocal characteristics are associated on analmost universal basis with perception ofspecific emotions, like anger, joy, fear,happiness, and sadness (Shearer, Banse,& Wallbott, 2001). When vocalization isaffected by pathological processes, theresultant dysphonic voice assumes deviantqualities, some of which may approxi-mate the voice qualities associated withvarious emotive states. Hence, even dur-ing an emotionally neutral speaking con-dition a dysphonic patient may sound asif he or she is expressing a specific emo-tion, when indeed there is no such intenton the part of the afflicted speaker. Thesituation may be even more convolutedwhen the vocal condition does not matchthe gender of the speaker. Moreover,with certain pathological conditions,this incongruence is more pronouncedwhen speaking on the phone, and maybe amplified further in face-to-face con-versation, the latter occurring when theaccompanying facial expressions bringthe total percept (voice-face-posture)closer to the assumed emotional quality,or when they are in conflict with theacoustic signal produced.

The scenario of voice-gender mismatchis typical when a female voice drops to alow F0, whether due to polypoidal vocalcords, or a virilized voice, or a personwho looks female but has a male voicequality (in a transgender patient; Flower& Izdebski, 1979; Izdebski, 2007).

The emotional mismatch scenario canbe present in a variety of dysphonias,and can be amplified further in neurolog-ical cases such as in the various dystoniasand dyskinesias, in various craniofacial


anomalies, or in Parkinsonian syndromepatients, where facial expressions do notmatch the intended vocal signals. Thismismatch, when present in nonpathologicpopulations, is known as the McGurkeffect (McGurk & MacDonald, 1976).

In this chapter I will discuss how thepathologic voice may be misinterpretedby the naïve listener as being indicativeof a “wrong” emotional or physical statethat is not representative of the dyspho-nic patient. I will demonstrate that thisscenario often brings serious embarrass-ment to the afflicted patient (the pro-ducer) and the listener alike, and in manycases carries profound psychosocial consequences for the afflicted patient(Krischke et al., 2005).

Vocal Movement Dysphonias

Typical vocal movement dysphoniasinclude focal dystonia of the larynx represented by adductor (ADDSD) andabductor (ABDSD) spasmodic dysphonia,vocal tremor, or mixed ADDSD with VoT(Izdebski, 1981; 1992; 1997; Izdebski &Ward, 1997). In all these conditions, symp-toms are organized phonetotopically,meaning that the voice can be normal or abnormal depending on predictableF0/dB levels or as a function of phonemiccontext (Izdebski, 1996; 2007).

ADDSD

The distinguishing quality of ADDSD isvocal overpressure (Izdebski, 1992). Over-pressure results from increased glotticcompression and increased subglotticpressure (Alku, Svec, Vilkman, & Sram,2000; Shipp, Izdebski, & Schutte, 1988).


PERCEPTION OF EMOTION IN IMPAIRED FACIAL AND VOCAL EXPRESSION 139

conversational appropriateness to thebrain dysfunction associated with PD.

The Parkinsonians in this study werealso asked to rate their pragmatic skillson an auto evaluation scale. The spousesalso rated the pragmatic skills of theiraffected partners with the same scale.The comparison of the results indicateda lack of awareness of the pragmatic dif-ficulty in the patients. However, thesame analysis was not conducted in thecontrol group, and the results mightneed to be interpreted with care.

Steinert’s Disease

Steinert’s disease is a hereditary myo-tonic atrophy. It is characterized by aslowness and a difficulty in relaxing themuscles during voluntary movements. Itis not accompanied by pain but is accen-tuated by cold and fatigue. The severityof the disease is variable from one indi-vidual to the other and ranges from aminimal functional impairment to aninability to walk. Observed symptomsare repeated infections of the respiratorysystem, a perturbation of the cardiovas-cular apparatus (particularly the heart),and ocular symptoms such as cataracts.The central nervous system and the endo-crine system are also often impaired.Thedisease touches various muscles and particularly the ones of the hand, themouth, and the tongue.This yields chew-ing and swallowing difficulties as well asvocal difficulties. The voice is monotoneand nasal. The mouth is often slightlyopened and the cheeks are hollow.

In a study by Morsomme et al. (2007),the vocal and facial emotion expressionof a patient with Steinert’s disease wascompared to an actor’s. The patient’scomplaint, besides an action myotony

increasing with cold, swallowing prob-lems, and a general fatigue, was his diffi-culty in adopting a facial mimicry reflect-ing his oral message.This is impairing hisprofessional occupation, where he is oftenleading meetings. The facial mimicry ofthe patient is affected by the muscularrigidity caused by his myotonic dystro-phy. On the vocal level, articulation diffi-culties and a characteristic hyponasalityare observed. In order to establish theorigin of the deficit of emotion expres-sion in the patient, it was important toisolate the different channels of emotionexpression according to the three modes:audio, visual, and audiovisual.

The tasks of the two subjects includedreading a neutral content text with fourdifferent emotions: joy, anger, sadness, andfear, and with a neutral expression. Thetwo subjects were recorded during thetask and the recordings were then judgedby adult and adolescent raters regardingthe category of the expressed emotionand the intensity of the expression.

The results showed that the actorobtained higher ratings than the patientin the gradation of the intensity. As forthe categorical choice, there was no sig-nificant difference between the subjects,except for the visual mode. The judgeshad difficulties identifying the expressedemotions on the sole basis of the patient’sfacial mimicry. That corroborates theauthors’ hypothesis that the patient’sdeficit would impair the judges’ ability to correctly identify the emotion. It alsoobjectivated the patient’s self-reporteddifficulty in adopting a facial mimicrythat reflects his intended message. Thelow intensity scores yielded by the patientas compared to the actor also suggest acertain difficulty for him to clearlyexpress his emotions, on both the facialand vocal level.


PROSODY AND DYADIC INTERACTION BETWEEN INFANT AND CAREGIVER 147

the mother’s voice (Piontelli, 1992). Allof this begins to occur around the age of4 months and is experienced while float-ing in the amniotic fluid, which also am-plifies the internal and external sounds.

Research has shown that shortly afterbirth, infants turn their heads in the direc-tion of the mother or the father whosevoice they heard while in the womb(DeCasper & Fifer, 1980). Therefore, animprint of the timbre, pitch, and melodyof the familiar voice appears to be en-coded in the infant, although it is difficultto specify how that early experienceaffects the developing brain-mind of thefetus. Once a child is born, these intonedwords are experienced within a rich mixof nonlinguistic sounds, facial expres-sions, gazes, gestures and physical hold-ing, caressing and touching—all of whichmake up the sensory-motor-affectivematrix of our earliest experiences ofother human beings, the world, and ournascent self.

Prosody includes vocal rhythm, pitch,melody, tempo, and volume as part of thecommunication of linguistic and para-linguistic (emotional and attitudinal)information. This prosodic “speech” ischaracterized by elevated pitch, simplifiedpitch contours, expanded pitch range,decreased tempo, and repetitiveness thatconvey meaning by varying stress andpitch irrespective of the words and gram-matical construction (Mitchell, Elliott,Barry, Crittenden, & Woodruff, 2003).

Prosody is thought to be processedmainly in the right brain along withother crucial socio-emotional informa-tion and body states that are central tothe regulation of emotions and the originof a sense of self (Schore, 2005a). Thisright brain regulatory capacity is depend-ent on experience—particularly the expe-

rience of attunement and synchrony witha caregiver— and develops primarilynonverbally through playful interactionsinvolving gesture, rhythm, melody, tempo,facial expressions, and movements thatform patterns of interactions between thedeveloping child and others. A mother’srhythm, melody, pitch, volume, and tempoare ideally finely tuned to the infant’sneeds and response; even deaf mothersvocalize to their young deaf infants,although neither can hear the sounds(Sieratzki & Wolf, 1996).

Researchers in the field of infantdevelopment and linguistics state that themost essential aspects of prosody occurduring the preverbal period of infancy andthat music is instrumental to the devel-opment of language in that the orderingof pitched sounds made by the humanvoice is the first thing we learn whenacquiring a language (Papousek, 1996).It has also been claimed that it is difficultto discern which sounds made by infantsare premusical and which are prelinguis-tic and that this lack of clear distinctioncontinues throughout the language andmusic acquisition processes. Starting ataround 21⁄2 years, they begin to separateout into language as the means of com-municating semantic content and mean-ing, with music and prosody continuing torepresent and express emotional contentand meaning throughout the lifespan(Steinke, Cuddy, & Holden, 1997).

The Intersubjective Musical Matrix

Bargiel (2005) states that most develop-mental researchers believe that we areborn with a biological predisposition for


investigate physiological stress reactivitypatterns in extroverted and introvertedindividuals and the specific role of neu-roticism in that context.

In addition, results from research onthe relation between type A personality,as predominantly seen in males, andspeaking style and facial behavior can beinformative. Type A personalities arecharacterized by an aggressive and hostileattitude, including nonverbal facial ten-sion such as teeth-clenching (Chesney,Ekman, Friesen, Black, & Hecker, 1997).Research on speaking style has shownthe largest differences between type Aand control personalities for syllabicemphasis, voice loudness, voice harsh-ness, hostility, and speaking rate, indicat-ing an increase in all of these factorswith type A personality (Chesney et al.,1997). Interestingly, investigations onpsychological characteristics of patientswith vocal fold granulomas seem to par-allel these results (Kiese-Himmel, Pralle,& Kruse, 1998). Thus, future researchcould incorporate the type A constructas a potential predictor of some voicedisorders.

Psychopathology in Voice Disorders

Although stress and anxiety are closelyrelated, stress can be present withoutconcurrent psychological distress suchas anxiety and depression (Cohen &Williamson, 1988). However, allostaticload or chronic stress can be conceptu-alized as a precursor predicting a risk for future psychiatric disease (Cohen,Kamarck, & Mermelstein, 1983). Morespecifically, research has investigated therelation between stress and psychiatricdiseases and results have pointed to a


stress-induced, biological basis of anxietyand depression, by way of HPA dysregu-lation (Cohen et al., 1995).

A recent study obliquely related tothis notion was conducted on the fre-quency of perceived stress as comparedto anxiety and depression in patientswith voice disorders (Dietrich,Verdolini,Gartner-Schmidt, & Rosen, in press). Thestudy aimed to explore (a) the relation-ship between stress and laryngeal diag-nosis, and (b) the relation between stressand anxiety and depression in relation tolaryngeal diagnosis.Two hundred eighty-one new patients (65% female) presentingto a voice center completed self-reportquestionnaires that assess perceived globalstress within the preceding 1 month(Perceived Stress Scale-10, PSS-10; Cohen& Williamson, 1988) and anxiety anddepression based on symptoms from thepreceding week (Hospital Anxiety andDepression Scale, HADS; Zigmond &Snaith, 1983). Findings indicated the fol-lowing results. About 25% of all patientswith MTD (primary and secondary) andvocal fold lesions, about 30% of patientswith laryngopharyngeal reflux (LPR),and about 40% of patients with paradox-ical vocal fold motion disorder (PVFMD)produced stress scores equal to or greaterthan one standard deviation above themean for normative data. Further, fully50% of all patients with vocal fold lesions,PVFMD, LPR, and primary MTD (as com-pared to about 30% of patients with sec-ondary MTD), had at least mild anxiety.Only a small number of patients haddepression scores indicating at least milddepression: about 15% of patients withprimary MTD and PVFMD and 10% of allother patients. In comparison, patientswith glottal insufficiency due to paralysisand other causes had significantly lowerstress and anxiety scores compared to


PSYCHOBIOLOGICAL FRAMEWORK OF STRESS AND VOICE 171

subjects in other diagnostic categories,and also had the lowest depressionscores. In conclusion, primarily elevatedanxiety and to a lesser degree stressappeared to co-occur with voice disor-ders in roughly the following descendingorder of magnitude of relation: PVFMD,LPR, vocal fold lesions, and MTD (Diet-rich et al., in press).

Only one other study of which we areaware investigated the prevalence ofadjustment-related stress, anxiety, voiceuse, and somatic complaints in a popula-tion of individuals with voice disorders(Schedule of Recent Experiences, SRE;Davis, Eschelman, & McKay, 1988; Gold-man et al., 1996). In that study, patientswith vocal fold nodules showed no higherstress scores compared to controls,whereas a comparison group composedof individuals with hyperfunctionalvoice disorders (MTD, vocal fold lesionsother than nodules) did show higherstress scores. In addition, both groupsshowed higher state and trait anxietyscores as compared to healthy controls(Goldman et al., 1996).

Comparing the two studies, the com-monalities are elevated anxiety scores forindividuals with hyperfunctional voicedisorders with and without lesions.Results surrounding stress are weakerand less easily comparable, but researchshould pursue the relationship betweenanxiety and stress in voice disorders.The pervasiveness of anxiety and stressin individuals with voice disorders isstriking, and possibly may influence theoccurrence and maintenance of a widerange of voice disorders relevant in clin-ical practice. However, no causality canbe implied at this point, based on thenoted cross-sectional studies, and no pre-dictions can be made about risk factorsfor future voice disorders due to anxiety

or stress. Further, the likely bidirectional-ity between stress and voice disorders onone hand, and voice disorders and stresson the other hand, makes it difficult todisentangle the pathways. For instance,it has been established that voice disor-ders decrease an individual’s quality oflife (Smith et al., 1996), and thus mayincrease stress.

Physiological Pathways in Voice and Stress

Past research around psychological fac-tors involved in certain individuals withvoice disorders has focused primarily onpersonality issues and psychopathologyusing self-report measures as previouslydiscussed, and on the general impor-tance of stress in professional voice users(Rosen & Sataloff, 1997). New frontiersfor investigation include studies around(a) the specific role of stress in voice dis-orders, and (b) the potential pathwaysby which stress may affect voice. Stateddifferently, no physiological or biologi-cal research exists to date of which weare aware that specifically studies thelink between stress and voice disordersusing a psychobiological frameworkincorporating complex stress-emotion-voice interactions. However, recent workhas made some forays in that direction.For example, one study looked at theeffects of “vocal constitution” (i.e.,dynamic range), autonomic reactivity,and state and trait anxiety on vocalendurance in female student teacherswithout vocal complaints (Schneider et al.,in press). Interestingly, autonomic stressreactivity did not predict fundamentalfrequency changes during actual teach-ing in the classroom—although it shouldbe noted that other vocal parameters


A UNIFIED MODEL OF COGNITION, EMOTION, AND ACTION 187

Vocal Expression of Emotionsin Normally Hearing Infants

Up to now, studies that focus on thequestion of to what extent infant vocal-izations contain information about differ-ent emotional states were done mainlyon crying behavior. The results of thesestudies are controversial. According toWasz-Höckert et al. (1968), four types ofcries can be discriminated: birth, hunger,pain, and pleasure cries. However, Mulleret al. (1974) and Murry et al. (1975)reported that mothers were not able toidentify the stimulus that elicited cryingin their infant. Other authors suggestthat infant crying is a graded signal, mir-roring a continuum of emotional statesreaching from arousal to urgency (Bren-nan & Kirkland, 1982; Porter, Miller, &Marshall, 1986; Protopapas & Eimas,1997; Zeskind, Sale, Maio, Huntington, &Weiseman, 1985). Porter et al. (1986), forinstance, investigated infant cries duringpainful circumcision procedures. Theyfound correlations between the acousticstructure of the elicited cries and thedegree of intrusiveness of the particularprocedure. In the perceptual test, thecries elicited by the most intrusive pro-cedures were judged by adult listenersalso to be the most urgent ones.

According to Keller and Schölmerich(1987), parents interpret the vocalizationsof even 2-week-old infants as expressionsof emotional states and respond in a dif-ferentiated way to different vocaliza-tions. There is no general agreement,however, whether infants during theirfirst months of life are able to expressspecific emotions in their behavior at all(for an overview, see Strongman, 1996).Some authors suggest that, in the first

months, the expressive behavior is to alarge extent random. This assumption isbased on the observation that specificvocal or facial patterns often occur with-out a specific stimulus preceding them(for an overview, see Camras, 1994). Onereason for this low predictability of spe-cific behavior patterns could be a lowdegree of emotional differentiation inearly infanthood. Most authors agree thatfrom the very beginning, there is a differ-entiation into at least two emotionalstates: aversive and nonaversive (Giblin,1981; Izard & Malatesta, 1987; Lewis,1993; Malatesta-Magai, Izard, & Camras,1991; Sroufe, 1979). Some authors, how-ever, argue that more than two emo-tional states can be distinguished (Giblin,1981; Izard & Malatesta, 1987; Malatesta-Magai et al., 1991). This disagreementabout the expression of infant emotionsmay be partly due to the difficulties inestimating the emotional state of aninfant, because we rely on assumptionsabout the infants’ internal state. Theseassumptions can be strengthened by con-textual information (e.g., certain contextsin which specific emotions are proba-ble), or by comparing the facial expres-sions of infants with the facial expressionsof emotions of adults (see Ekman & Oster,1979; Izard, Huebner, Risser, McGinnes,& Dougherty, 1980). A further possibilityto judge the emotional state of infants isto use their parent’s ratings (Wasz-Höckert,Lind, Vurenkoski, Patanen, & Valannen,1968), since in a communication processthe receivers of a signal should be able todecode the transmitted information.

In our study (Scheiner et al., 2002;2004; 2006), we used a combination ofparent ratings and contextual informa-tion to judge the emotional state of theinfants.We tested the four most common


= 3.66; for Unit B, M = 26.57 s, SD = 4.76;and for Unit C, M = 31.83, SD = 5.4).

Each vocal data set was analyzed withCSL (Computerized Speech Lab, model4300 B, software version 5.X, Kay Ele-metrics Corp.) for 11 acoustic parame-ters regarding time (rate of articulation),pitch (fundamental frequency, F0), andenergy (intensity in dB). For both pitchand energy, the mean, range (gamma +,gamma −), and standard deviation werecalculated.

Furthermore, the vocal data were sub-jected to a participant-by-participantanalysis, to determine the vocal profilesof the male seducers (Ekman, O’Sullivan,Friesen, & Scherer, 1991). This analysisonly took into consideration variationsbetween each individual participant’sown speech samples. In this way, eachparticipant effectively acted as statisticalcontrol for himself. Thus, different cate-gories of participants were identifiedfrom the analysis of the variations fromthe respective baseline values.The analy-sis consisted of (a) quantification of thedifferential values of each participant(i.e., the differences between the rawvalues of the seductive speech units andthe raw values of the baseline speech);(b) calculation of the differential thresh-old for the definition of the vocal profilecategories (the differential thresholdvalue was equal to or greater than twicethe standard error); and (c) the definitionof the vocal profile categories as “partic-ipants +” (those who had positive differ-ences equal to or greater than twice thethreshold value), “participants −” (thosewho had negative differences equal to orgreater than twice the threshold value),or “participants =” (those who showedpositive or negative differences lower thantwice the standard error). Based on thisanalysis, several further categories were


identified for the vocal profiles (high-pitched, low-pitched, variable, mono-tone, and intermediate voice for pitch;loud, soft, variable, stable, and intermedi-ate voice for energy; fast, slow, and inter-mediate rate of articulation).

Main Vocal Features of Seduction

As general result, the voice of seductionin the standard utterance was character-ized by a high modulate pitch (F0 =156.32), loud speech (dB = 58.23), and afast rate of articulation (M = 8.97 sylla-bles/s). Generally, when inviting the part-ner to meet him again, the potentialseducer spoke at a higher pitch, higherintensity, and accelerated rate of articula-tion, compared to his standard speech.

More interestingly, the temporal dis-position of the vocal profiles in threeperiods (beginning, middle, and end) ofthe experimental situation revealed therichness of vocal variations in the seduc-tive behavior of the male participants. Atthe very beginning, most participants(57.89%) spoke in a higher voice thanthey did in baseline speech, whereasonly a minority (15.79%) lowered thepitch. In any case, all participants tendedto modulate their pitch in a variable way.Conversely, in the middle period, therewas a drop in the percentage of partici-pants who spoke at a high pitch (26.32%)compared to those who spoke at a low or intermediate pitch. During thisperiod, the voice became more monoto-nous. In the final period, an increase inthe trend shown in the middle periodwas observed: participants spoke to thepartner with a low and monotone (whis-pered) pitch.

As regards voice intensity, initiallymost participants (52.63%) spoke at a


THE VOCAL DANCE OF SEDUCTION: FROM BEING ANYONE TO BEING SOMEONE 209

high volume intensity with a loud andvariable voice, whereas only a minority(21.05%) spoke in a low-intensity voice.In the middle period, the vocal conditionwas reversed and most participants(57.89%) spoke at a low volume, using avariable soft voice. In the last period,most of the participants spoke consis-tently at a lower volume.

With regard to speed (rate of articula-tion), in the first period only a very smallminority (5.26%) spoke slowly, whilemost participants (94.74%) spoke at amedium or fast rate. In the middle period,there was an increase in the number ofparticipants who spoke slowly (21.05%).Conversely, in the last period, the num-ber of participants who spoke quicklyincreased (52.63%),

In sum, during the seductive interac-tion, the male partners produced a highlevel of vocal variations, as revealed bythe participant-by-participant analysis.At the very beginning of the interaction,most of participants spoke louder and ata higher pitch than their standard. In themiddle and at the end of the interaction,their voices became progressively lowerand softer. Conversely, the rate of articula-tion increased constantly during the seduc-tive process, even more so at the end.

The Voice of Successful andUnsuccessful Seducers

The follow-up data enabled the “effi-cacy” of the seductive strategies used bymale participants during the experimen-tal situation to be tested. On the basis of this efficacy, it was possible to distin-guish between successful and unsuccess-ful seducers, as outlined above.As shownin Figure 13–2, the differences betweenthe vocal cues employed by successful

and unsuccessful seducers during theseductive interaction were general andmarked.

As regards pitch, at the very begin-ning, most of the successful seducersspoke at a high pitch and none used a low pitch, whereas more than half of those unsuccessful spoke at a low ormedium pitch. In the middle of the inter-action, the percentage of successfulseducers who spoke at a high pitchdropped compared with the previousperiod, while only 14.29% of them spokeat a low pitch. Conversely, more than90% of the unsuccessful seducers spokeat a low or medium pitch in the middleof the interaction. This trend continuedinto the last period for the unsuccessfulseducers, while the successful seducerschanged their voice, lowering furthertheir pitch.

As far as intensity is concerned, at thestart of the interaction, successful seduc-ers spoke in a loud voice, whereas only41.67% of unsuccessful ones spokeloudly, and 25% spoke at a low volume.In the middle of the interaction, the suc-cessful seducers reduced the intensity oftheir voice (only 14.29% went on speak-ing in a loud voice, while 42.86% spokein a soft voice). Among the unsuccessfulseducers, a greater percentage loweredthe volume of their voice in the middleof the interaction. In the last period, thistrend was consolidated by the unsuc-cessful seducers (75% of them used alow-intensity voice), while the successfulones spoke at a medium intensity.

With regard to the rate of articulation,most of the successful seducers spokequickly and none spoke slowly at thebeginning of the interaction. Conversely,only one third of unsuccessful seducersspoke quickly. In the middle of the inter-action, successful seducers markedly


Theoretical Framework forMeasuring Vocal Indicators of

Coping Styles

The patient’s way of coping with illnessis of major importance for both psycho-social (e.g., well-being) and biomedical(e.g., compliance with treatment) aspectsof the disease. Obrist (1981) describedpassive coping as a state of enduring astressor without expending effort to actupon it, while active coping was definedas tonic readiness to act upon an event.This corresponds to the action readinessdimension of Cannon’s emergency re-sponse (Cannon, 1929). Generally speak-ing, active coping styles are thought tobe more appropriate for dealing withstressful events, while avoidant copingappears to be a psychological risk factorfrom a long-term perspective (Holahan &Moos, 1987).

An important determinant of copingreactions is the feeling of potency or con-trol over the illness.The feeling of control,or lack of control, can induce cognitive-affective states of hopefulness or hope-lessness (Ortony, Clore, & Collins, 1988,p. 132) and thus significantly influencethe patients’ coping reactions, whichinclude both somatic and autonomicreactions. The latter can be induced byanticipated activity (Obrist, Webb, Sut-terer, & Howard, 1970). ß-adrenergicsympathetic effects on the heart have


been found to predominate during activerather than passive coping, particularlywhen environmental uncertainty and be-havioral uncertainty are high (Bongard,1995; Brener, 1987; Kelsey, 1991; Kelseyet al., 1999; Light & Obrist, 1980; Sher-wood, Allen, Obrist, & Langer, 1986).Obrist (1981) also provided evidenceregarding the role of parasympatheticphysiological reactions in subjects withpassive coping style.

As voice and speech production mech-anisms (respiration, phonation, and artic-ulation) are strongly influenced by thesubjects’ autonomic reactions, studyingthe patients’ vocal expression appearedto be a more direct method of assessingthe emotional factors associated withcoping styles.

What Is Measured: The Voice ofFull-Blown Emotions or the Voiceof Affective Dimensions?

Research on vocal indicators of affectivestates has been dominated by designsbased on the relations between acousticfeatures and semantic tasks of emotionlabeling. Psychologists and speech scien-tists rely on the subjects’ intuitive labelingof the states called “emotions,” often inforced choice recognition tasks (Cowie& Douglas-Cowie, 1996; Kienast & Sendl-meier, 2000; Mozziconacci & Hermes,1999; Pereira, 2000; Schröder, 2001;Yuan,

higher VDI scores than those with passive/maladaptive coping(median VDI = 21.3). In conclusion, the results show a strongtendency for the vocal arousal to be associated with thepatients’ coping style.


Introduction

Quite independent of the content ofspeech, the tonal features of the humanvoice convey a rich array of nonverbalinformation about the speaker. Consider,for instance, answering the telephone asa case in point. Even though you may notknow or have ever met the person whois calling, as soon as he or she starts totalk, you can immediately tell if you arespeaking to a male or a female, or to anadult or a child.

Voices contain much more embeddedinformation than just gender and devel-opmental status. Studies have shown thatlisteners who hear voice samples canaccurately infer the speaker’s race (Lass,Tecca, Mancuso, & Black, 1979), gender(Lass, Hughes, Bowyer,Waters, & Bourne,1976), socioeconomic status (Ellis, 1967;Harms, 1963), personality traits (Adding-ton, 1968; Allport & Cantril, 1934; Zuck-erman & Driver, 1989), and emotionaland mental state attributes related todeception (Ekman, Friesen, & Scherer,1976; Streeter, Krauss, Geller, Olson, &Apple, 1977). Listeners are also capableof estimating the age, height, and weightof speakers with the same degree ofaccuracy achieved by examining photo-graphs (Krauss, Freyberg, & Morsella,2002; Lass & Colt, 1980; Lass & Davis,1976). Independent raters are capable ofmatching a speaker’s voice with the per-son’s photograph over 75% of the time(Krauss et al., 2002). Furthermore, thereis evidence that the sound of an individ-ual’s voice is an external attribute that isused as a cue for assessing potentialmates and plays an important role ininterpersonal attraction (Miyake & Zuck-erman, 1993; Oguchi & Kikuchi, 1997).


Voice Attractiveness

What makes some voices sound moreattractive than others? There have beena variety of investigations on this matterusing acoustic and sound spectrographicanalyses, but, as yet, no one has been ableto identify or agree on all of the physicalproperties that distinguish attractivefrom unattractive voices. For instance,with regards to pitch, Daniel and McCabe(1992) found that both sexes rate mid-pitch voices as being the “most sexy,”whereas Oguchi and Kikuchi (1997)found that both sexes evaluated a low-toned voice with a small pitch range asattractive, but the exact opposite wasshown by Oksenberg, Coleman, and Can-nell (1986), where a higher pitch withgreater variation was associated withvoice attractiveness.

Whatever the underlying tonal/acousticfeatures may be that constitute attractivevoices, there is high inter-rater reliability(i.e., consensus) with regards to whichvoices sound attractive (Hughes,Dispenza,& Gallup, 2004; Hughes, Harrison, &Gallup, 2002; Zuckerman & Driver, 1989;Zuckerman, Hodgins, & Miyake, 1990).Not only do people tend to agree aboutwhich voices are attractive, they are usu-ally capable of making this decision assoon as they hear a person begin to talk.In addition, individuals rate the attractive-ness of same-sex and opposite-sex voiceswith the same degree of consensus(Addington, 1968; Hughes et al., 2002).

Vocal attractiveness appears to exertan influence on mate selection, and indi-viduals with voices that sound attractiveare perceived as having more desirablepersonality characteristics (Zuckerman& Driver, 1989). Individuals with attrac-


WHY ARE WE ATTRACTED TO CERTAIN VOICES? 233

tive voices are thought to be warmer,more likable, honest, dominant, and morelikely to achieve (Berry, 1990; Zuckerman& Driver, 1989). Thus, a “vocal attractive-ness stereotype” appears to exist (Zuck-erman & Driver, 1989; Zuckerman et al.,1990). Furthermore, the higher the rat-ing of vocal attractiveness, the more sim-ilar the speaker is judged to be to like thelistener and the more the listener wouldlike to affiliate with the speaker (Miyake& Zuckerman, 1993). Favorable impres-sions of speakers with attractive voiceshave been obtained under conditionswhere observers only heard the speaker’svoice, as well as when they both saw andheard the speaker (Miyake & Zucker-man, 1993).

There is evidence suggesting that reli-able impressions of other personalityattributes can also be inferred on the basisof the sound of a person’s voice (Allport& Cantril, 1934; Berry, 1990; Zuckerman& Driver, 1989). Addington (1968) foundthat when making inferences about per-sonality based upon vocal samples, theaverage inter-rater reliability coefficientwas a .81 across 40 measured personalitytraits.When listeners form an impressionof the speaker’s personality traits, theyare influenced more by the speaker’stone of voice than by the content of thespeech (Yogo, Ando, Hashi, Tsutsui, &Yamada, 2000). Furthermore, the effectsof manipulating a single vocal sample bychanging the rate, mean, and variance offundamental frequency all showed a cor-responding change in personality ratings(Brown, Strong, & Rencher, 1974).

Most studies that investigate vocalattractiveness use voice samples that areneutral (such as a standardized sentence,word count, or vowel recitation) andattempt to control for such things as

content, speech patterns, dialect, accents,visual information, intonations, and in-flections (Collins & Missing, 2003; Daniel& McCabe, 1992; Hughes et al., 2002;Hughes et al., 2004; Zuckerman, et al.,1990; Zuckerman & Miyake, 1993). Butin certain contexts, it can be argued thatit is possible that individuals modify theirvocal patterns in order to persuade,seduce, or influence potential mates. Forexample, there is evidence showing thatmales’ vocal style in seductive speechproduces a profile that differs significantlyfrom the one used in normal speech(Anolli & Ciceri, 2002). On the otherhand, the sound of a person’s voice canonly be manipulated within limits, andmuch of the variance in vocal quality isundoubtedly influenced by one’s biologyand genetic makeup. (To learn moreabout vocal attractiveness, see Chapter 2in Volume 1: Editor.)

Voice and Physical Features

A person’s voice not only conveys psy-chological information about that per-son, but it can also reveal informationabout one’s basic biological features. Forinstance, both height and weight can beinferred from a person’s voice with thesame degree of accuracy as viewing aphotograph (Krauss et al., 2002; Lass &Colt, 1980; Lass & Davis, 1976).

There are also several studies showingthat ratings of voice attractiveness pre-dict certain physical features. For instance,we have shown that voice attractivenessis related to bilateral symmetry. Individu-als who possess symmetrical bilateraltraits (such features as length of the fin-gers, hand widths, etc.) tend to have


thyroarytenoid, bilateral depressor angulioris, procerus/corrugators, and frontalismuscles were injected. This has resultedin moderate improvement in her dys-phonia and reduction in unwanted facialcontortions. She is pleased with theresults and continues to return for peri-odic injections.

Summary and Conclusions

Several clinical conditions can cause sig-nificant disruptions of voice, speech, andassociated facial expression that are dif-ficult to correct purely through behavioralor speech therapy. These conditions canalso affect emotive expression in thesepatients either by introducing unwantedrepresentation of emotion, especiallyduring declarative speech, or by magni-fying intended emotions. The idea forusing botulinum toxin type A injectionswas based on our long history of treatingADDSD using these techniques. The useof these injections in MD patients hasbeen described before (Arikan et al., 2006;Rapaport et al., 2000; Schneider et al.,2006; Suskind & Tilton, 2002;Tarsy et al.,1997). Botulinum treatment of droolingin CP patients has been reported (Arikanet al., 2006), but not for vocal or facialproblems. Finally, we were not able tolocate any reports of the use of this ther-apy for these indications in CHI patients(http://www.ncbi.nlm.nih.gov/PubMed),May, 2007. Our results have been encour-aging to date. In select patients manifest-ing these problems, botulinum toxininjection therapy may prove useful. Allthree of our patients feel they have ben-efited and have continued periodic treat-ment for over a year. Specifically, patientswith spasmodic-like dysphonias and


facial grimacing and in particular whoexperience difficulty with emotional ex-pression may benefit from this therapy.

References

Aminoff, M., Izdebski K., & Dedo, H. H. (1978).Clinical aspects of spasmodic dysphonia.Journal of Neurology, Neurosurgery, andPsychiatry, 41(4), 361–365.

Arikan, O. K., Tan, F. U., Kendi, T., & Koc, C.(2006). Use of botulinum toxin type A forthe treatment of masseteric muscle hyper-trophy. Journal of Otolaryngology, 35(1),40–43.

Dedo, H. H. (1976) Recurrent laryngeal nervesection for spastic dysphonia. Annals ofOtology, Rhinology, and Laryngology,85, 451–459.

Dedo, H. H., & Izdebski, K. (1981). Surgicaltreatment of spastic dysphonia. Contem-porary Surgery, 19, 75–90.

Dedo, H. H., & Izdebski, K. (1983). Interme-diate results of 306 recurrent laryngealnerve sections for spastic dysphonia.Laryngoscope, 93(1), 9–16.

Dedo, H. H., & Izdebski, K. (1983). Problemswith surgical (RLN section) treatment ofspastic dysphonia. Laryngoscope, 93(3),268–271.

Dedo, H. H., & Izdebski, K. (1984). Evaluationand treatment of recurrent spasticity afterrecurrent laryngeal nerve section.A prelim-inary report. Annals of Otology, Rhinology,and Laryngology, 93(4, Pt. 1), 343–345.

Gharabawi, G. M., Bossie, C. A., Lasser, R. A.,Turkoz, I., Rodriguez, S., & Chouinard,G. (2005). Abnormal Involuntary Move-ment Scale (AIMS) and ExtrapyramidalSymptom Rating Scale (ESRS): Cross-scalecomparison in assessing tardive dyskine-sia. Schizophrenia Research, 77(2–3),119–128.

Izdebski, K. (1992). Symptomatology ofadductor spasmodic dysphonia: A physio-logic model. Journal of Voice, 6, 306–319.


A COMPREHENSIVE MODEL OF HOW THE STRESS CHAIN AFFECTS VOICE 261

ence; in the long term being estimatedas an excellent singer, actor, lawyer,teacher; earning a living as a profes-sional voice user;

■ On a creative level: the intrinsic pleas-ure of mastering a technique, puttingon a vocal performance that transcendsthe ordinary; in the long term spiritualwell-being or fulfillment.

The negative effects of maladaptivebehavior are:

■ On a physical level: vocal fatigue, vocaldiscomfort, and edema of the vocalfold can be defined as short-term, andvocal nodules as a long-term effect ofvocal misuse and abuse;

■ On a personal level: doubts about one’ssocial position, occupational incapac-ity; in the long term work incapacity;

■ On a creative level: suffering fromburnout.

To summarize: the dynamic stresschain consists of: stressors; neuroen-docrinocrine reactions; physical signsand organic changes; cognitive aspects,feelings, and emotions; instrumentalbehaviors; short- and long-term effects,which meet to some extent the needs ofsafety, security, and creativity.

Figure 18–1 and Figure 18–2 summa-rize the biological and mental stresschain.

Interaction of Biological andMental Activation

The interaction of primarily biological andmental excitement and the most commontypes of secondary kinesthetic emotionalexcitement are listed in Table 18–1.

Interference of Stress withVocal Performance

The obvious relationship between vocalcommunication and stress is commonlystated (Brockway, 1979). Even in the fieldof aviation and space medicine there existdata about the correlation of voice pitchwith emotional excitation (Johannes,1994; 1995; Ruiz, 1990).

Stress affects the physiological andmotor mechanism of vocal performance(Sundberg, 1987).This is reflected in sev-eral acoustic parameters such as pitchand intensity variations, vocal range (Sei-dner, 1981), timbre and prosodic changes,jitter, shimmer, (Fuller, 1992; Mendoza,1998), duration of the syllables, and otherspeech aspects (Engstrand, 1988; Fuller,1992). Moreover, voice and speech mod-ulations are the vehicles to communi-cate, for adequately expressing the broadspectrum of human feelings and emo-tions. The listener can easily detect theemotional state of a speaker and singerjust by listening and watching the mimic(Sundberg, 1987). Voice, posture, andmimic are very precise emotion indica-tors.The emotional state of a speaker andsinger has a considerable effect on theway the vocal tract is used (Cosmides,1983; Sundberg, 1987).

Changes of fundamental frequency (F0),in particular, have significant relationshipwith emotional load. The maximum F0

constitutes the primary indicator of theperception of emotional stress (Cosmides,1983; Protopapas, 1997; Wittels, 2002).

The influence of life events has beenshown to have an important effect onthe etiopathogenesis of functional dys-phonias (Butcher, 1993; Freidl, 1990;1993; House, 1988). Voice pitch mode


is to analyze the information containedin characters’ voices in foreign films orsoap operas and even news announcerstransmitting happy and sad events. Moreimportant than identifying the vocal psy-chodynamics characteristics is to use thisinformation positively to improve com-munication competence.

Vocal PsychodynamicsApproach

The goal of the Psychodynamic Approach(PA) is to offer feedback to let patientsrecognize the specific impact of theirvoice on the listener. The patient isguided in identifying the revealed impact,and the clinician makes it clear that vocalparameters are conditioned throughoutthe life or may be the result of a pathol-ogy and do not always represent thepatient’s emotion or true nature. For thisprocess, examples can provide a greathelp: a loud voice is common in childrenthat grow up in noisy families; fast speechrate may be due to stressful situations;uncoordinated breathing can be due to a


vocal fold paresis and not to anxiety;rough voice can be the result of a dis-torted mucosal wave from vocal nodulesand not due to an abrasive manner. So,it’s important to make the patient com-prehend that the psychological impactdoes not always represent the real self;for example, a closed jaw articulationdelivers aggression even if the cause istemporomandibular joint dysfunction.

Voice is a primary and automatic instru-ment for emotional expression, becauseit has a physiological basis and a commu-nicative function (Scherer, 1995). Thevoice clinician must be aware that anybehavioral change is submitted to thepsychological acceptance of the patient.Voice is a strong mirror of one’s self.Everybody influences and is influencedby others’ vocal patterns.To omit consid-eration of the vocal psychodynamic of a dysphonic patient can restrict or evenimpair the vocal rehabilitation outcome.

Table 19–1 summarizes the mainnaïve interpretations of different types ofvoice, and Table 19–2 matches frequentvocal problems and common features to their main vocal psychodynamicsinterpretation.

Table 19–1. Type of voice and main vocal psychodynamic interpretation

Type of Voice Main Interpretation

Roughness Aggression, anger, and/or fatigue

Breathiness Fatigue, sensuality (discrete deviation), lack of power, and/or shyness

Strain Effort, difficult life, relationship problems, and/or distress

Nasality Sensuality (discrete deviation), lack of social skills, and/or limitedintelligence

Tremulous Emotional liability, lack of confidence, conflict, and/or fear

Asteny Lack of energy, shyness, and/or lack of social skills or confidence

Creaky/Basal Aggression, attempt to control or to express power, villainy


00 izdebski vol2 i-xxiifm - plural publishing the estill voice model: physiology of emotion 83...

Documents