|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001931132
007 cr mnu|||uuuuu
008 080321s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001728
Nikjeh, Dee Adams.
Vocal and instrumental musicians :
b electrophysiologic and psychoacoustic analysis of pitch discrimination and production
h [electronic resource] /
by Dee Adams Nikjeh.
[Tampa, Fla] :
University of South Florida,
ABSTRACT: Neurological evidence indicates that instrumental musicians experience changes in the auditory system following skill acquisition and sensory training; yet, little is known about auditory neural plasticity in formally trained vocal musicians. Furthermore, auditory pitch discrimination and laryngeal control are recognized as essential skills for vocal musicians; however, the relationship between physiological variables, perceptual abilities, and vocal production is unclear. Electrophysiologic and psychoacoustic measures were used to examine pitch production accuracy as well as pre-attentive and active pitch discrimination between nonmusicians and two classes of musicians. Participants included 40 formally trained musicians (19 vocalists/21 instrumentalists) and 21 nonmusician controls. All were right-handed young adult females with normal hearing.^ stimuli were harmonic tone complexes approximating the physical characteristics of piano tones and represented the mid-frequency range of the untrained female vocal register extending from C4 to G4 (F0 = 261.63-392 Hz). Vocal pitch recordings were spectrally analyzed to determine pitch production accuracy. Difference limens for frequency (DLFs) were obtained by an adaptive psychophysical paradigm. Pre-attentive auditory discrimination was assessed by auditory evoked potentials (AEPs), including the mismatch negativity (MMN). A standard tone (G4 = 392 Hz) and three deviants differing in frequency (1.5%, 3%, and 6% below) were presented in a multi-deviant paradigm. All musicians demonstrated superior pitch perception and vocal production compared to nonmusicians. Pitch perception and production accuracy did not significantly differ between vocalists and instrumentalists; however, pitch production accuracy was most consistent within the vocalist group.^ Music training appears to facilitate both auditory perception and vocal production regardless of music specialty. Pitch perception and production were correlated skills only for instrumental musicians. Vocalists demonstrated minimal variability for both skills so that perception and production were not correlated. These two skills may be independent abilities between which a relationship develops with training. AEP analysis revealed an influence of musical expertise on neural responses as early as 50 ms after onset of musically relevant stimuli. MMN responses indicate that vocal musicians as well as instrumental musicians have superior sensory memory representations for acoustic parameters of harmonic stimuli and imply that auditory neural sensitivity is developed by intense music training.
Dissertation (Ph.D.)--University of South Florida, 2006.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 212 pages.
Co-adviser: Stefan A. Frisch, Ph.D.
Co-adviser: Jennifer L. Lister, Ph.D.
Auditory evoked potential (AEP).
Event related potential (ERP).
Mismatch negativity (MMN).
Vocal pitch matching.
t USF Electronic Theses and Dissertations.
Vocal and Instrumental Musici ans: Electrophysiologic and Ps ychoacoustic Analysis of Pitch Discrimination and Production by Dee Adams Nikjeh A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Communicati on Sciences and Disorders College of Arts and Sciences University of South Florida Co-Major Professor: Stefan A. Frisch, Ph.D. Co-Major Professor: Jennifer J. Lister, Ph.D. Lynne Gackle, Ph.D. Arthur M. Guilford, Ph.D. Jane Scheuerle, Ed.D. Date of Approval: August 24, 2006 Keywords: auditory evoked potential (AEP), event related potential (ERP), frequency discrimination, mismatch negativity (MMN), vocal pitch matching Copyright 2006, Dee Adams Nikjeh
Dedication This great adventure is dedicated to my husband and best fr iend, Mr.-Dr. Nikjeh, who challenged and encouraged me to take my dream and make it an achievement. And In memory of my dad, Ronald H. Adam s, who taught me by word and example, To learn is to live, and to live is to learn.
Acknowledgments I have been considerably blessed to work with an outstanding and supportive group of researchers and educators. They have each made a personal and unique contribution to this research endeavor and all have my sincerest appreciation. I have greatly benefited from the wisdom and guidance of Arthur M. Guilford, Ph.D. His belief in my success remained steadfa st even when mine tended to waver. He has expertly guided me through an often tedious process so that I might keep the goal in focus. Dr. Guilford is, and always will be, a trusted advisor and a valued friend. I am indebted to my dissertation co-chair s, Stefan A. Frisch, Ph.D. and Jennifer J. Lister, Ph.D. for their unlimited assistance. Dr. Frischs keen insightfulness and provocative criticisms have enri ched the analyses of this research and challenged me to think beyond the obvious. Many thanks are dire cted to Dr. Lister for introducing me to an exciting world of psychoacoustics and elec trophysiology. I am truly grateful for her patient and skillful instruction. She has fore ver changed my perspective of audiology! The day I met Lynne Gackle, Ph.D., I met a kindred spirit. Her passion about music education and her enthusiasm for this research are responsible for the successful recruitment of musicians from the Univers ity of South Floridas School of Music. Appreciation is also given to Jane Scheue rle, Ed.D. It is an honor to have Dr. Scheuerle as a member of my dissertation co mmittee. I have tremendous respect for her
visionary insight and am sin cerely grateful for her editor ial contributions and astute suggestions. A special acknowledgment is ex tended to Joseph R. Stei niger, M.D. who has been most gracious with his time and assistance through every step of this project. He diligently read chapter after chapter and c ontributed a fresh insight from a medical perspective. Finally, the support of ones cohorts is always appreciated. Specifically, I thank Susan Fulton, who patiently gave of her time an d expertise to instruct and assist me with this research.
i Table of Contents List of Tables vi List of Figures vii Abstract ix Chapter One Introduction 1 Music and Neural Plasticity 2 Vocal Pitch Control 3 Auditory Pitch Perception 4 Relationship between Pitch Discrimination and Pitch Control 5 Statement of the Problem 6 Potential Application 8 Purpose of the Study 11 Research Questions 12 Chapter Two Review of the Literature 14 Music and Neural Plasticity 14 Knowledge of Music 14 Innate or learned 14 Musical meaning 15 Musical syntax 16 Music and Neural Correlates 18 Temporal processing 19 Pitch processing 20 Summary 24 Music Training and Neural Plasticity 25 Microstructural plasticity 25 Microstructural effects of aud itory training 26 Macrostructural plasticity 29 The Musicians Brain Anatomical Differences 30 Corpus callosum 30 Cerebral cortex 31 Gray matter volume 32
ii Cerebellum 33 Summary 33 Auditory Pitch Perception 34 Acoustical Perception of Music 35 Cortical Models of Pitch Perception 37 Modular theory 37 Parameter theory 39 Neural Specialization for Pitch Processing 40 High-level pitch processing 40 Low-level pitch processing 45 Vocal Pitch Control 51 Evolution of Singing 51 Laryngeal Anatomy for Pitch Control 52 Laryngeal cartilages 53 Laryngeal muscles 55 Physiology of Vocal Pitch Control 58 Vocal Pitch Control An Integrat ion of Systems 60 Phonatory monitoring systems 61 Auditory monitoring system 63 Kinesthetic feedback system 66 Summary 71 Development of Pitch Control 71 Sequence of development 71 Socio-cultural influence 73 Potential Factors Related to Vocal Pitch Control 73 Potential functional variables 74 Potential physical variables 75 Summary 77 Relationship between Pitch Discrimination and Pitch Control 77 Auditory Discrimination and Vocal Pitc h Control in Children 78 Summary 81 Auditory Discrimination and Vocal Pitc h Control in Adults 82 Summary 85 Chapter Three Methods and Procedures 86 Introduction 86 Purpose of the Study 88 Research Questions 89 Hypothesis 89 Questions 89 Null Hypotheses 90 Research Design 91 Variables 92
iii Independent variable 92 Dependent variables 92 Procedures 93 Participant Selection 93 Exclusion criteria 93 Inclusion criteria 94 Formally trained vocal musicians 96 Formally trained instrumental musicians 96 Control subjects 96 Stimuli Considerations 97 Harmonic versus pure tone selection 97 Frequency selection 98 Equipment Common to All Tasks 99 Vocal Pitch Matching 99 Procedure for vocal pitch matching 99 Data analysis 100 Active Auditory Frequency Discrimination 101 Procedure for frequency difference limen (DLF) 101 Data analysis 103 Pre-Attentive Auditory Pitch Di scrimination 103 Mismatch negativity (MMN) 105 P100-N100-P200 108 Electroencephalographic recording 109 Procedure 110 Data analysis 113 Chapter Four Results 115 Participant Demographics 115 Psychoacoustic Measures 116 Vocal Pitch Matching 116 Statistical analysis 117 Active Auditory Frequency Discrimination 119 Statistical analysis 120 Comparison of DLF and PPA 121 Correlation Analysis of DLF and PPA 121 Correlations of DLF and PPA with Qu estionnaire Variables 123 Electrophysiological Measures 124 Pre-Attentive Auditory Pitch Di scrimination 124 Sensory perception P1-N1-P2 complex 124 Mismatch negativity-MMN 127 Statistical Analysis of the Mismatch Negativity 134 MMN Amplitude 134
iv MMN Latency 136 Correlation Analysis with Electrophys iological Variables 138 Psychoacoustic and Electrophysiolo gical Variables 138 Years of Music Training and Electro physiological Variables 139 Age Training Initiated and Electrop hysiological Variables 140 Summary of Findings 141 Chapter Five Discussion 144 Discussion of Findings in Relationship to the Research Questions 145 Vocal Pitch Production Accuracy 145 Active Auditory Pitch Discrimination 147 Electrophysiological Measures of Pre-Attentive Auditory Pitch Discrimination 152 Sensory perception P1-N1-P2 complex 152 Mismatch negativity (MMN) 154 MMN amplitude 154 MMN latency 156 Relationships between Auditory Perception and Pitch Production Across Groups 158 Correlations between DLF and PPA 158 Correlations between psychoacoustic and electrophysiol ogical variables 159 Summary of rela tionships among groups 160 Relationships between Auditory Perception and Pitch Production Within Groups 160 Comparison of DLF and PPA 161 Comparison of pre-attentive and active pitch discrimination 162 Summary of within group relationships 163 Limitations of the Study 164 Directions for Future Research 165 Conclusions 167 List of References 170 Appendices 187 Appendix A: Participant Screeni ng Questionnaire 188 Appendix B: Participant Information Questionnaire 189 Appendix C: Informed Consen t Form 190 Appendix D: Participant Profile Data 193 Appendix E: Individual Data for Pitch Production Accuracy 197 Appendix F: Individual Data for Diffe rence Limen for Frequency 206 Appendix G: P1 Individual Peak Amplitude at Fz 209
v Appendix H: P1 Individual Peak Latency at Fz 210 Appendix I: Mismatch Negativity Individual Peak Amplitude at Fz 211 Appendix J: Mismatch Negativity Indivi dual Peak Latency at Fz 212 About the Author End Page
vi List of Tables Table 1 Pitch Production Accuracy Group Data 116 Table 2 Difference Limen fo r Frequency Group Data 119 Table 3 Average Peak Latencies Deri ved from Grand Average Waveforms 134 Table 4 Rel DLF% and Rel PPA% Group Comparison 151
vii List of Figures Figure 1. Scatter Plot of DLF and PPA. 122 Figure 2-A. Comparison of Group Averag e Waveforms of P1-N1-P2 Complex for Deviant 1 Alone Condition ( 386.21 Hz, 1.5% deviance) at Fz between Controls (C-Dev1) and Musicians (M-Dev1) 125 Figure 2-B. Comparison of Group Averag e Waveforms of P1-N1-P2 Complex for Deviant 2 Alone Condition ( 380.58 Hz, 3% deviance) at Fz between Controls (C-Dev2) and Musicians (M-Dev2) 125 Figure 2-C. Comparison of Group Averag e Waveforms of P1-N1-P2 Complex for Deviant 3 Alone Condition ( 369.81Hz, 6% deviance) at Fz between Controls (C-Dev3) and Musicians (M-Dev3) 126 Figure 3. Grand Average Waveforms Examples of Fz, Cz, and M2 128 Figure 4-A. Control Groups Mismatch Response (MMN) 129 Figure 4-B. Instrumentalists Mismatch Response (MMN) 130 Figure 4-C. Vocal Musicians Mismatch Response (MMN) 131 Figure 5-A Grand Average Difference Waveforms for Deviant 1 (386.21 Hz, 1.5%) comparing control group (C), instrumentalists (IN), and vocalists (V) 132 Figure F-B Grand Average Difference Waveforms for Deviant 2 (380.58 Hz, 3%) comparing control group (C), instrumentalists (IN), and vocalists (V) 132 Figure 5-C. Grand Average Difference Waveforms for Deviant 3 (369.81Hz, 6%) comparing control group (C), instrumentalists (IN), and vocalists (V) 133 Figure 6-A. MMN Response Amplitudes by Group and Deviant Condition 135
viii Figure 6-B. MMN Response Amplitude by Musician Genre and Deviant Condition 136 Figure 7. MMN Response Latency by Group and Deviant Condition 137
ix Vocal and Instrumental Musici ans: Electrophysiologic and Psychoacoustic Analysis of Pitch Discrimination and Production Dee Adams Nikjeh ABSTRACT Neurological evidence indicates that in strumental musicians experience changes in the auditory system following skill acquisiti on and sensory training; yet, little is known about auditory neural plasticity in forma lly trained vocal musicians. Furthermore, auditory pitch discrimination and laryngeal cont rol are recognized as essential skills for vocal musicians; however, the relationship be tween physiological variables, perceptual abilities, and vocal production is unclear. Electrophysiologic and psychoacoustic m easures were used to examine pitch production accuracy as well as pre-attentive a nd active pitch discrimination between nonmusicians and two classes of musicians. Participants included 40 formally trained musicians (19 vocalists/21 instrumentalists) a nd 21 nonmusician controls. All were righthanded young adult females with normal hearing. Stimuli were harmonic tone complexes approximating the physical character istics of piano tones and represented the mid-frequency range of the untrained female vocal register extending from C4 to G4 (F0 = 261.63-392 Hz). Vocal pitch recordings were spectrally analyzed to determine pitch production accuracy. Difference limens for frequency (DLFs) were obtained by an adaptive psychophysical paradigm. Pre-attentiv e auditory discrimination was assessed by
x auditory evoked potentials (AEPs), incl uding the mismatch negativity (MMN). A standard tone (G4 = 392 Hz) and three devian ts differing in frequency (1.5%, 3%, and 6% below) were presented in a multi-deviant paradigm. All musicians demonstrated superior pitch perception and vocal production compared to nonmusicians. Pitch pe rception and production accuracy did not significantly differ between vo calists and instrumentalists ; however, pitch production accuracy was most consistent within the vocalist group. Music training appears to facilitate both auditory per ception and vocal production re gardless of music specialty. Pitch perception and production were correlated skills only for instrumental musicians. Vocalists demonstrated minimal variability for both skills so that perception and production were not correlated. These two skills may be i ndependent abi lities between which a relationship develops with training. AEP analysis revealed an influence of musical expertise on neural responses as early as 50 ms after onset of musically relevant stimuli. MMN responses indicate that vocal musicians as well as instrumental musicians have superior sensory memory representa tions for acoustic parameters of harmonic stimuli and imply that auditory neural sens itivity is developed by intense music training.
1 Chapter One Introduction Music is a universal occurrence in all human cultures. Throughout history, on every part of the earth, in ev ery past and present culture, i ndividuals have enjoyed music. Examination of the neural basis of music pr oduction and investigations of neural changes related to music training provide opportuni ties for researchers to study the interaction between brain development and environmental in fluences. It is argued that because of the intense training and skill acquisition that a musician receives from an early age, the musicians brain serves as an excellent m odel for the study of neuroplasticity (Gaser & Schlaug, 2003; Mnte, Nager, Beiss, Schr oeder, & Altenmller, 2003; Pascual-Leone, 2001; Schlaug, 2001; Zatorre, 2003). Current res earch indicates that trained instrumental musicians have superior auditory pitch disc rimination ability relative to nonmusicians (Kishon-Rabin, Amir, Vexler, & Zaltz, 2001; Speigel & Watson, 1984). Pitch is the auditory perception of a tones frequency. Electrophysiological data have shown that long-term music training modifi es neural processing of acoustic input. In addition, it has been found that instrumental musicians have faster neural responses for pitch changes than nonmusicians (Koelsch, Schmidt, & Kansok, 2002; Shahin, Bosnyak, Trainor, & Roberts, 2003).
2 It is questioned whether the neurologi cal and anatomical differences between musicians and nonmusicians are inherent neurop hysiological distincti ons or secondary to training-induced neurological changes. While the literature cont ains comprehensive evidence for neural plasticity in trained in strumental musicians, little is known about formally trained vocal musicians. Vocal musicians receive intense music training comparable to instrumental musicians. The vocalists musical instrument is the larynx, the biological organ within the human body re sponsible for voice production. Auditory pitch discrimination (perception) and vocal pitch control (production) have been identified as related abilitie s and essential skills for succes sful vocal musicians. Vocal pitch control requires the integration of the bodys motor and sensory systems. The interactions and relationships between the auditory system and the laryngeal system necessary for instant and exact vocal pitc h productions are areas of interests for researchers and educators (Amir, Amir, & Kishon-Rabin, 2003; Mrbe, Pabst, Hofman, & Sundberg, 2004; Wyke, 1974). Music and Neural Plasticity The presence of neurologi cal and anatomical differences between musicians and nonmusicians supports the prem ise of functional and structural experience-dependent plasticity in the auditory system (Pantev, Engelien, Candia, & Elbert, 2001; PascualLeone, 2001; Schlaug, 2001; Schn, Magne, & Be sson, 2004; Trainor, Shahin & Roberts, 2003). Electrophysiological studi es of auditory responses reveal differences between musicians and nonmusicians and parallel anatom ic studies show cortical enlargement of
3 auditory areas important for music perception (Elbert et al ., 1998; Pantev et al., 2003; Trainor et al., 2003). Taken together, these studies provide evidence of enhanced preattentive auditory processing in musicians compared to non musicians, suggesting that fundamental auditory abilities to process pitch and temporal features can be facilitated by music training and supporting the theory of training-induced co rtical plasticity. Specifically, electroencephalography (EEG ) and magnetoencephalography (MEG) data suggest that musical expertise influences pitch processing by refining the neural frequency-processing network (Koelsch, Schr ger, & Tervaniemi, 1999; Pantev et al., 2001; Schn et al., 2004; Shahin et al., 2003; Te rvaniemi, 1993; Traino r et al., 2003). Vocal Pitch Control The vocal production of music requires the integration of multiple brain systems including the sensorimotor, auditory, limbic, and executive systems (Mnte et al., 2003). There are those within our population who express an ex ceptional ability to produce musical modulations of the voice for singing. The ability to sing with accurate pitch control is considered the most basic featur e that distinguishes singers from nonsingers (Murry, 1990; Titze, 1994; Watts, Barnes -Burroughs, Adrianopoul os, & Carr, 2003). Physiologically, the act of singing involves control and coordination of several neuromuscular systems. In addition to re spiration, resonance, and articulation, vocal pitch precision relies on pre-phonatory tuning of the laryngeal musculature, laryngeal reflex modulation, and an auditory gove rnance system (Elman, 1981; Jrgens, 2002; Kirchner & Wyke, 1965; Sundberg, 1987; Wyke, 1967; Wyke, 1974).
4 Auditory input, neuromuscular pitch me mory, and kinesthetic feedback of the laryngeal system contribute to pitch cont rol (Amir et al., 2003; DiCarlo, 1994; Jones & Munhall, 2000; Mrbe et al., 2004; Titze, 1994; Ward & Burns, 1978). It has been suggested that the developmen t of kinesthetic feedback or internal models of pitch control assists trained sing ers in controlling fundamental frequency and maintaining targeted pitches more accurately than non-tr ained singers (Murry, 1990; Sapir, McClean, & Larson, 1983; Ward & Burns, 1978; Watts, Murphy, & Barnes-Burroughs, 2003). DiCarlo (1994) writes that vocal instruction and reflex condi tioning train the professional singer to associate an auditory image with an internal sensation. Similarly, an internal model for the control of pitch has been proposed by Jones and Munhall (2000). This model corresponds to a neural representa tion of the spatial, dynamic, and/or proprioceptive characteristics that provide an internal pitch reference to the nervous system to predict and plan for vocal frequency control. In other words, singers match the perceived pitch to a reference pitch in the brain. Longitudinal studies of vocalists in training indicate that accuracy for the absolu te neuromuscular memory of pitch increases with music education (Mrbe, Pa bst, Hofman, & Sundberg, 2003, 2004). Auditory Pitch Perception Speigel and Watson (1984) describe a re lative acuteness of the ears and a mystique associated with the listening abilities of performers, conductors, and composers of classical music (p. 1690). A physical characteristic important for the perception of speech and music is a change in fundamental frequency; that is, a change in
5 pitch (Novitski, Tervaniemi, Huotilainen, & N tnen, 2004). Pitch extraction is basic to the perception of speech intonation, but prec ise pitch perception is crucial to the processing of music. Musical melodies us e much smaller pitch intervals than speech intonation contours (Ayotte, Peretz, & Hyde, 2002). The ability to perceive and discriminate pitch differences is regarded by music educators as a fundamental capacity for musical talent and an implicit skill of a successful performer (Bentley, 1966; Geringer, 1983; Seashore, 1919). Psycho acoustic studies comparing frequency discrimination thresholds, also known as di fference limens for frequency (DLFs), for musicians and nonmusicians report significantly smaller discrimination thresholds for musicians (Kishon-Rabin et al., 2001). Researchers have also investigated pr ocesses of auditory pitch perception for music through neuropsychological studies, neural imaging and, more recently, electrophysiology. EEG studies of pre-atten tive pitch discrimination indicate superior pre-attentive discrimination by musicians as compared to nonmusicians suggesting training induced modification of pre-attentive auditory neural proces sing (Koelsch et al., 1999; Shahin et al., 2003). Relationship between Pitch Di scrimination and Pitch Control Auditory pitch discrimination and vocal pitch control reflect abilities necessary for accurate integration of sensory percep tion, motor planning, and execution of vocal production. Intuitively, it seems these two abilities are directly related. A positive relationship between auditory pitch discri mination and vocal pitch matching skills is
6 reported for instrumentalists (Amir et al., 2003). Amir and colleagues (2003) postulate that intense music training fine-tunes the c oordination between audi tory perception and motor-production skills. The authors stated that musici ans are more perceptive to acoustic parameters in vocal productions comp ared to nonmusicians despite the fact that the musicians in the study were instrume ntal musicians and had no previous vocal training. Although many researchers acknowledge a relationship between auditory pitch discrimination and vocal pitch control, research is sparse and the nature and development of this relationship is uncertain (Amir et al., 2003; Geringer, 1983; Goetze, Cooper, & Brown, 1990; Watts et al., 2003; Yarbrough, Green, Benson, & Bowers, 1991). Amir and colleagues (2003) and Goetze and colleague s (1990) suggest a plausible relationship between auditory pitch discrimination and vocal pitch matching ab ilities; however, the relationship does not appear to be reciprocal. Physiologi cal, perceptual, and production variables may be independent abilities between which relationships form as a result of training-induced neural changes in the auditory system. Statement of the Problem Neuroplasticity of the human brain reflects dynamic neural changes and reorganization as an effect of training and experience (Merzenich et al., 1996; Pantev et al., 2003; Teter & Ashford, 2002). Investigations comparing the brains of musicians and nonmusicians have identified anatomical and physiological differences in the cortex and cerebellum (Schlaug, 2001). Data from esta blished neurophysiol ogical techniques,
7 including EEG and MEG, suggest that expe rience-dependent functi onal and structural plasticity occurs in the aud itory system of musicians. These differences support the premise that neural changes occur in th e human brain following skill acquisition and sensory stimulation. Although musician includ es instrumentalists and vocalists, there is a paucity of comparative research including fo rmally trained vocal musicians. Previous neurophysiological research has focused pre dominantly on instrumental musicians (e.g., violinists, keyboard players) ra ther than vocal musicians (M nte et al., 2003; Schlaug, 2001; Zatorre, 2003). Vocal musicians adhere to the same rigorous training as other musicians; however, the auditory system of vocal musicians has been studied to a much lesser extent. The overall objective of this study is to take an initial step to contribute to the body of basic research regarding the per ception and production abilities of formally trained vocal musicians. A review of the available literature indi cates an inadequate understanding of the relationship between physiological variables, perceptual abil ities, and pitch production of formally trained vocal musicians. Investig ators and educators have identified auditory pitch discrimination (perception) and vocal pitc h control (production) as related abilities and essential skills for vocal musicians. Auditory pitch discrimination and vocal pitch control contribute to the professional singers laryng eal neuromotor performance; however, the strength and nature of this relationship is unclear. Evidence suggests that long-term vocal training influences aud itory abilities for pitch perception and discrimination. Previous studies have examined these skills separate ly using a variety of tasks in the musically trained population; however, no previous investigation has
8 attempted to relate pre-attentive auditory neural responses to active auditory pitch perception and vocal pitch matching abilities in the population of formally trained vocal musicians. What is the relationship among th ese abilities in this population as compared to formally trained instrumental musicians and musically untrained subjects? Do the vocal musicians have an identifiable pattern of abilities? Does skill in one area precede ability in another? Specifically, objective information regard ing the relationship between auditory pitch discrimination and vocal pitch control, as well as the effects of long-term vocal training on the neurophysiology of the auditory sy stem is needed. This study is designed to contribute to a growing body of research identifying relationships that may have implications for vocal perfor mers and music educators. Potential Application A consideration of human differences in a ny domain invariably leads to the issue of nature versus nurture. Reliable neuroimaging techniques reveal anatomical and neurological differences between musicians and nonmusicians. Elect rophysiological data provide evidence of enhanced pr e-attentive auditory processing in musicians compared to nonmusicians. Existing evidence indicates that formally trained professional singers control fundamental frequency and maintain accurate pitch better than untrained singers (Dejonckere, 1995; Jones & Mundall, 2000; Leydon, Bauer, & Larson, 2003; Wyke, 1974). Whether these differences and superior abilities are inherent and/or dependent on training and neural plasticity is a controversial issue.
9 Are these differences determined by a ge netic code whose expressions guide the decision to seek musical training and become a professional musician ; or alternatively, do these attributes arise from modifications of synaptic connections or neural growth influenced by sensory input from musi c training at an early age (Buonomano & Merzenich, 1998; Monaghan, Metcalfe, & R uxton, 1998; Shahin, et al., 2003)? Investigating neural changes associated w ith the acquisition and mastery of new skills represents one experimental model used to determine whether or not functional and anatomical markers of exceptional skills exist or develop (Gaser & Schlaug, 2003). Similar to an athletic scout who searches fo r certain indicative qualities such as height, speed, and balance, identification of predictive variables of musical skill may assist music educators in the early identification of children with potential musical ability. Investigating the relationship betwee n the components of our physiological mechanisms which discriminate and control fundamental frequency may expand existing data on the function of audito ry feedback and vocal product ion in those individuals who lack these abilities. For those who can sing eas ily and accurately, the failure of others to do the same is baffling. Much of the existi ng research in music education has focused on children who cannot match pitch (Apfelstadt, 1984; Geringer, 1983; Goetze, Cooper, & Brown, 1990; Green, 1990; Howle, 1992; Joyner 1969; Moore, 1994; Pedersen & Pedersen, 1970; Porter, 1977; Yarbrough et al., 1991). The possibility that some children may not learn to sing accurately is considered a major problem for music educators (Yarbrough et al., 1991).
10 There is a growing body of evidence sugges ting that music training of children promotes cognitive development including reading and math achievement, as well as critical thinking abilit ies, motor skills, and social ab ilities (Weinberger, 1994). Music lessons require focused atte ntion and daily practice. Music training involves a multiplicity of experiences including comprehe nsion of musical notation and structures (e.g., musical symbols for notes and timing, chords, scales, clefs), memorization of musical passages, progressive mastery of fi ne-motor skills and emotional expression during performance (Schellenberg, 2004). Comp ared with groups of children engaged in nonmusical activities, those children who r eceived music training demonstrated greater increases in full-scale intelligen ce quotients (Sch ellenberg, 2004). Music and language are intimately related and share similar neural substrates (Friederici, Pfeifer, & Hahne 1993; Koelsch et al ., 2000; Tervaniemi & Brattico, 2004). Acoustically, signals of musi c and language consist of va riations of intensity and frequency as a function of time which are pe rceived by the brain as sound. Cognitively, both have rules of syntax and are dependent on memory (Tervaniemi & Brattico, 2004). Neural imaging data from EEG and functi onal magnetic resonance imaging (fMRI) imply there is considerable overlap of neural structures and pr ocesses underlying the auditory perception of music and language (Friederic i et al., 1993; Koelsch et al., 2000). EEG data comparing formally trained musicians to nonmusicians indicate that extensive music training facilitates pitch anal ysis by refining the auditory frequency-processing network not only for music, but also for language. If music training stimulates neuroanatomical changes in the cerebral cortex, then further id entification and differen tiation of the neural
11 substrates for music processing and vocal pitch control may have implications for treatment strategies for individuals w ho have neurological, language, and vocal impairments, such as dyslexia, aphasia, Parkinsons disease, and hearing impairment (Ayotte et al., 2003; Overy, 2003; Rami g, Yoshiyuki, & Bonitati, 1991). While no single investigative technique is sufficient to provide more than a small piece of the puzzle, converging evidence from a variety of methods is needed to provide a comprehensive and robust understanding of th e relationship between music training and neurological structures. The global picture that emerges fr om studies of music and its neural substrate is far from complete; however each piece of information contributes to our overall comprehension of the complex st ructure and function of the human brain. Purpose of the Study The purpose of this study was to assess, compare and correlate three identified variables of perception and production that contribute to the performance of the singing voice. This study proposed a causal-compara tive (ex post facto) design to assess and compare the assigned variables of active au ditory pitch discrimi nation, pre-attentive auditory pitch discrimination, and vocal pitc h matching accuracy w ithin and between the following groups: formally trained vocal mu sicians, formally tr ained instrumental musicians, and a matched unt rained control group. This investigation sought to determin e whether a significant difference exists between formally trained vocal musicians, fo rmally trained instrumental musicians, and nonmusicians for the abilities of vocal pitch control, active auditory pitch discrimination,
12 and pre-attentive auditory pitch discrimi nation. Furthermore, by examining these abilities across populations, it is possible to assess what, if any, relationships exist among these abilities. It was also questioned whethe r formally trained vocal musicians, similar to instrumental musicians, experience traini ng-induced neural plas ticity in the auditory system. This study was a beginning step of i nquiry into the effects of intensive music training on the auditory neural f unction of vocal musicians. There was one independent variable, subject group which was sub-divided into formally trained vocal musicians, formally trained instrumental musicians, and a matched control group of musically unt rained subjects. Three depend ent variables were measured and reported as: (1) relative accuracy for vocal pitch produc tion accuracy in percentage ( rel PPA%), (2) relative difference limen for frequency in percentage ( rel DLF%), and (3) latency and amplitude of the mismatch nega tivity (MMN); that is, an auditory evoked potential (AEP) associated with pre-a ttentive auditory ne ural responses. Research Questions The relationship between physiological, pe rceptual, and production abilities for musical stimuli between formally trained musi cians and musically untrained subjects was examined. Specifically, relationships be tween vocal pitch matching accuracy, active auditory pitch discrimination, and pre-atte ntive auditory pitc h discrimination among formally trained vocal musicians, formally trained instrumental musicians, and a matched
13 control group of musically unt rained participants were investigated. This study was designed to answer th e following questions: 1. Is there a difference in vocal pitch ma tching accuracy between musicians and the control subjects and furthermore, is ther e a difference between the instrumental and vocal musician groups? 2. Is there a difference in active auditory frequency discrimination ability between musicians and the control subjects? Mo reover, is there a difference between the instrumental and vocal musician groups? 3. Is there a difference in pre-attentive aud itory neural responses to pitch change (i.e., pre-attentive auditory pitch disc rimination for musical stimuli) between musicians and the control subjects and pa rticularly between the instrumental and vocal musician groups? 4. Is there an overall correlation betwee n perception and production variables across the groups? 5. Is there a correlation between percepti on and production variables within each subject group (i.e., controls, instrument al musicians, and vocal musicians)?
14 Chapter Two Review of the Literature Music and Neural Plasticity Knowledge of Music Innate or learned. What is the interaction betwee n genetics and the environment that produces distinct musical abilities? Music is recognized as a universal characteristic occurring in all human societie s, both past and present. Cross-cultural ev idence supports the innateness of music and indi cates that certain features of music, such as interval scales, are universal regardless of the mu sical genre or style (Hauser & McDermott, 2003; Tillman, Bharucha, & Bigand, 2000). An inte rval, as it relates to music, refers to the distance between sounds played simultane ously or successively and is crucial for scales and harmony (Pantev et al., 2003; Tervaniemi & Br attico, 2004). Certain acoustic stimuli are recognized as music by most member s of a given culture, even if these sounds have never been heard before; and convers ely, there are acoustic stimuli that humans recognize as nonmusical or diss onant (Hauser & McDermott, 2003). Therefore, even if a particular melody has never been heard, a di ssonant tone may be detected based on an internal musical representation (Tervaniemi & Brattico, 2004). This representation may correspond to a neural template hardwired in the brain or may become automatic
15 secondary to implicit neuronal models that develop from exposure to music in the environment (Tervaniemi & Brattico, 2004). Studies of infant auditory perception de monstrate seemingly innate traits. Young infants prefer consonant musical intervals rather than dissona nt intervals (Schellenberg & Trehub, 1996; Trainor & Heinmiller, 1998; Trehub, 2001) and they are capable of detecting the smallest differen ces that are musically meani ngful in any culture (Trehub, Schneider, & Henderson, 1995). However, the fetu s can hear a filtered version of sounds in the external environment by the third tr imester of pregnancy (DeCaspar & Fifer, 1980). Learning occurs during the fetal peri od and the nature of this learning with respect to music depends on the musical s ound environment before birth (Tervaniemi & Huotilainen, 2003). Thus, it is possible that seemingly innate traits are actually the result of early exposure to music. Musical meaning. Musical meaning is understood within the context of an arrangement of acoustic events, such as a scale or melody. The melody, referred to as the musical structure, has two components, rhyt hm and pitch (Pantev et al., 2003). Rhythm refers to timing and/or beat. Pitch is perceived as a tones highness or lowness. It is the perceptual correlate of frequency which pertai ns to the sounds physical structure (i.e., the number of cycles per second) (Patel & Balaban, 2001). The pitch produced by a persons voice is measured as the fundamental frequency (F0). Pitch structure has contour and an inte rval code. Contour re fers to the up and down pattern of pitch changes common to speech prosody and music. Interval code is the distance between two sounds on a musical scal e. The perception of pitch along musical
16 scales is central to pitch orga nization. A musical scale refers to the use of a small subset of pitches in a given musical piece. Scale tones are not equivale nt and are organized around a central tone, called the tonic. Th is tonic hierarchy of pitch facilitates perception, memory and performance of mu sic by creating expectancies (Peretz & Coltheart, 2003). Although the commonly used scales differ fr om culture to cult ure, most musical scales use pitches of unequal intervals or ganized around five to seven focal pitches (Tillman et al., 2000). The seven tones above or below a given tone in a scale form an octave. In Western culture, sp eech intonation contours use va riations in pitch that are larger than an octave to c onvey relevant information. In contrast, musical melodies of Western culture use smaller pitch intervals approximately 1/6th to 1/12th of an octave (Ayotte, Peretz, & Hyde, 2002). In other culture s, such as Arabic, Indian, and Chinese, the musical pitch intervals are even smalle r (Tervaniemi & Brattico, 2004). Thus, the auditory processing of pitch for music is necessarily more sensitive than for speech. Musical syntax. Like language, music is rule-gov erned. Each musical style has a relatively small set of rules to generate an infinite variety of musical compositions (Trehub, 2003). The rules that govern musical st ructure are referred to as musical syntax. Musical syntax does not imply that it is a li nguistic syntax in musi cal terms; rather, it reflects that music is structured according to complex regularities similar to language (Koelsch & Friederici, 2003). The ability of listeners to expect sp ecific musical events according to complex musical regularities and to detect violations of harmonic
17 expectancies within a musical sequence is an example of musical syntax (Bharucha & Krumhansl, 1983; Koelsch, Schmidt, & Kansok, 2002; Tillman et al., 2000). When a person sings, plays an instrument or speaks a sentence, a succession of acoustic events constitutes a context whic h is understood by others. To understand a musical context, listeners extract a tonal center by perceiving the musical relations between notes; that is, the interval (K rumhansl & Kessler, 1982). The mental representation of tonality and mu sical context is quickly esta blished by the listener; thus, there is an expectancy of what tone come s next. The dominant t onic progression at the end of a harmonic sequence is considered a basic syntactic structure for major-minor tonal music. For listeners, the sound of a c hord that violates musi cal regularities of major-minor tonal music is perceived as unexpected (Bharucha & Krumhansl, 1983; Koelsch, Schmidt, & Kansok, 2002; Tillman et al., 2000). Similar to language, culturally specif ic aspects of music are dependent on knowledge acquired through prior experience. Music percepti on is molded by implicit and/or explicit experience and is founded on early automatic functions of the auditory system that dynamically organize and stor e separated sounds (Tervaniemi & Brattico, 2004). Thus, in theory, knowledge of musi c, including musical meaning and syntax, may be acquired through normal exposure to mu sic within a culture without training just as linguistic knowledge is acquired through exposure independent of education.
18 Music and Neural Correlates The structural and functional organizati on of the human auditory system for the processing of music has been an issue of res earch dating back to the beginning of the 19th century. Franz Joseph Gall (1758-1828), a physic ian from the University of Vienna, established the idea that the brain was th e organ of the mind and as such, it was comprised of multiple organs so that the different functions of the brain were situated in specific sites (Bentivoglio, 2003). In Galls classificati on, the organ of music was responsible for the relationships between sounds, musical memory, and emotions of melody and harmony. This organ was located la terally in the suprao rbitary region, at the border between the inferior frontal and supe rior temporal regions. Gall identified this organ by palpation of the head of several musically talented i ndividuals and first recognized it in Mozarts head (Bentivogli o, 2003). Galls concep t of localization of mental function has had a lasting impact. Even in the 21st century, the issue of localizing musical structures and f unctions in the human brain is still debated. During the 1960s, brain lesion research by experimental psychologists supported the distinction between music and language by locating each of these functions in a different hemisphere (Platel, 2002). A widely held view attributes linguistic function to the left cerebral cortex and other non-verbal auditory functions, such as those involving music and environmental sounds, to the right cerebral cortex (Ligois-Chauvel, Peretz, Baba, Laguitton, & Chauvel, 1998). However, subsequent research suggests music processing is complex, bi-hemispheric, and interactive.
19 The primary auditory cortical areas responsible for auditory processing are located in the left and right temporal lobe s in the middle and superior temporal gyri, including the associative areas which expand to the posterior sites of the temporal lobes (Tervaniemi & Hugdahl, 2003). The primary aud itory cortex is mainly engaged in the early stages of processing for pitch, durat ion, intensity, and spatial location; whereas, more complex features involvi ng temporal patterns are proc essed via neurons within the associative areas. In a truly linear system, resolution of tim e and pitch has an inverse relationship so that as one is enhanced, the other is imped ed (Zatorre, 2001). In theory, the auditory nervous system is a highly nonlinear and di stributed system. Music is acoustically complex and requires neurophysiological proce ssing of multiple components including fundamental frequency, pitch contour, inte nsity, timbre and rhythm. Evidence supports functional asymmetry of the auditory cortices suggesting that temporal resolution occurs more rapidly in the left auditory cortical areas and spectral resolu tion is stronger in the right auditory areas (Dalla Bella & Peretz, 1999; Tervan iemi & Hugdahl, 2003; Zatorre, Belin, & Penhune, 2002). Zatorre and Belin (2001) speculate that neurons in the right auditory cortex compared with those in the le ft have increased synaptic densities, more closely spaced cortical columns, and compar atively less myelination which may reflect a specialization of these neurons for processing spectral information. Temporal processing. Ligois-Chauvel and colleagues examined the human auditory cortex by means of intracerebrally recorded auditory evoke d potentials in both hemispheres (Ligois-Chauvel, Giraud, Badi er, Marquis, & Chauve l, 2001). Findings
20 indicated that neurons in the ri ght auditory cortex were more sharply tuned to pitch than neurons in the homologous regions of the left hemisphere re vealing a functional asymmetry of the auditory co rtex and suggesting a preference for frequency (spectral) processing in the right Heschls gyrus. Anatomically, magnetic resonance imaging (MRI) shows the volume of white matter underl ying Heschls gyrus to be significantly greater on the left than on the right in two independent samples of right-handed subjects (Penhune, Zatorre, MacDonald, & Evans, 1996). If white matter volum e is indicative of myelination and thus greater speed of pro cessing, then these findings suggest faster transmission of acoustically relevant inform ation occurs on the left which supports the theory of rapid temporal processi ng in the left auditory cortex. Pitch processing. Whereas the analysis of speech requires good temporal resolution to process ra pidly changing formants, it can be argued that music processing requires good pitch resolution (Zatorre et al ., 2002). Pitch varia tion is an essential element of all music compositions, and as a re sult of this variation, structures such as melodies are created. Pitch processing is a central feature of music and is amenable to study because the physical paramete rs are easily manipulated. Pitch can be neurologically disassociated from the other perceptual functions and broken into a hierarchy of le vels (Foxton, Dean, Gee, Pere tz, & Griffiths, 2004; Peretz, 1990; Zatorre, 2001). Low-level pitch proces sing includes basic tasks such as the discrimination between two sounds or detection of pattern change. The discrimination of pitch sequence patterns and the organizat ion of sounds into melodies and harmony
21 require higher levels of pro cessing and may include interacti on with other cortical areas (Foxton et al., 2004; Zatorre, 2001). The nature of the neural processes un derlying basic pitch processing and the manner in which pitch is perceived is deba table (Gelfand, 1998; Zatorre et al., 2002). The precise interaction of frequency and tem poral coding is speculative. Traditionally, pitch processing has been dominated by two ma in theories: (a) the place theory, and (b) the temporal or rate coding theory (Lig ois-Chauvel, Giraud, Badier, Marquis, & Chauvel, 2001). The place theory postulates exclusively tonotopic coding throughout the pathways of the auditory system. A complex sound is broken down into its frequency components. These frequencies excite different places along the basilar membrane of the cochlea which resonate in response to a particular frequency (Gel fand, 1998). The place theory assumes that the pitch of a sound is directly related to this exc itation pattern (Moore, 1997). Thus, an incoming stimulus results in the vibration of those parts of the basilar membrane whose natural frequencies correspon d to the components of the stimulus. The temporal coding theory assumes that the pitch of a sound is related to the time pattern of the neural impulses evoke d by that sound (Moore, 1997, p. 143). The temporal theory proposes that the hair cells of the cochlea transmit all parameters of the signal to the central auditory nervous system for processing. This is accomplished by the volley principle which states th at groups of neurons work to gether so that the single response of the group is a spike corresponding to each cycle of the stimulus. It is believed that the rate at which a neuron fires correlates to the frequency of the stimuli; that is, the
22 impulses are phase-locked to the frequency. Neurons can only respond in an all-or-none manner. The absolute refractory period of the neuron corresponds to a maximum firing rate of 1000 times per second. Physiologic data from auditory nerve fibers indicate that the maximum pure tone frequency for which th e nerve fibers can pr eserve the period via phase locking is approximately 5000 Hz. Thus, temporal coding alone cannot account for the perception of pitch for pure tones having frequencies greater than 5000 Hz (Gelfand, 1998; Moore, 1997). Moore (1997) speculates that both codi ng processes may occur depending on the task and proposes a combination theory calle d the spectral-temporal theory that accounts for most of the existing data on the pitch perception of complex tones. This combination theory assumes that information from both low and high harmonics contributes to the determination of pitch. The place theory has b een shown to work best for the processing of higher frequencies and the temporal codi ng mechanism is best for coding frequencies in the lower frequencies; however, there is no agreement to the exact borderline between these registers (Gelfand, 1998; Novitski, Te rvaniemi, Huotilainen, & Ntnen, 2004; Moore, 1997; Zeng, 2002). Novitski and colleagues (2004) examined auditory frequency discrimination as indexed by electrophysiological measures. Data indicated that pre-a ttentive, auditory change-related responses; that is, the mismat ch negativity responses (MMN), recorded at lower frequencies (250 and 500 Hz) differed sign ificantly from those responses recorded at higher frequencies (2000 Hz and 4000 Hz). They found a glaring discrepancy in the MMN amplitudes and latencies as a functi on of frequency. From 250 to 1000 Hz, the
23 amplitude of the MMN was higher and the latency decreased i ndicating stronger and faster neural responses; however, as fre quency increased beyond 1000 Hz, the amplitude of response continued to grow while the latency increased, implying a stronger but slower neural response. Novitski et al. (2004) speculated that this changing point at 1000 Hz may indicate a transition between the pl ace and temporal mechanisms of pitch discrimination supporting a theory that the border between these two mechanisms is approximately 500 to 1000 Hz. Beyond basic pitch perception, neural co rrelates for low level pitch tasks (e.g., discrimination of two sounds or detection of pattern change) have also been examined. Data from electrophysiology and neural imag ing techniques support ea rlier brain lesion studies. Thus, although music processing e ngages components lateralized in both hemispheres, it is the posterior portion of the superior tempor al gyrus (STG) in the right hemisphere that is especially important fo r low-level pitch processing tasks (Johnsrude, Penhune, & Zatorre, 2000; Lig ois-Chauvel et al., 1998; Peretz, 1990; Zatorre, Evans, & Meyer, 1994). Perry et al. (1999) were th e first to use positron emission tomography (PET) scans to examine regional cerebral blood flow (rCBF) duri ng rudimentary singing of a single pitch and vowel. In contrast to a pitch perceptio n baseline, singing resulted in greater activation of th e right primary auditory cortical regions (Heschls gyrus). For higher level pitch processing tasks i nvolving melodies an d pitch sequencing, neural imaging reveals multiple areas of activ ation including bilateral activation of the superior parietal areas near the angular gyrus and activation of the occipital lobe. This implies integration with visual associativ e functions in the brain and frontal lobe
24 activation supporting the inter active roles of memory and attention (Gaab & Schlaug, 2003; Schmithorst & Holland, 2003; Zatorre et al., 1994). Summary. Based on reviewed evidence, the foll owing is a summary of generally accepted neural correlates for speech and mu sic (Alho et al., 1998; Anderson, Brown, & Tallal, 1993; Dalla Bella & Pe retz, 1999; Gelfand, 1998; Li gois-Chauvel et al., 2001; Steinmetz et al., 1989; Tervaniemi & Brattic o, 2004; Zatorre et al ., 2002; Zatorre et al., 1994): 1. The exact neural processes underlyi ng basic pitch perception and the precise interaction of frequency and temporal coding are uncertain. 2. The left auditory hemisphere is im plicated for speech processing, while the right auditory hemisphere is implicated for music. 3. The left auditory hemisphere is resp onsible for processing fast temporal information intrinsic to speech, while the right auditory hemisphere is responsible for processing minute cha nges in pitch (spectral information) intrinsic to music. 4. The right auditory hemisphere is domi nant for directing spatial attention. 5. The superior temporal and inferior front al cortices in the right hemisphere interact for the active retention of pitch. 6. This pattern of hemispheric functiona l asymmetry is consistent for both attentive and pre-attentive levels of musical cognition.
25 Music Training and Neural Plasticity The ability of a sensory or motor system to adjust or adapt to environmental stimuli, or a compensation of a cerebral struct ure for another impaired area due to injury is referred to as neuroplasticity (Schlaug, 2001). Synaptic changes at a cellular level refer to microstructural plasticity. Neuroplasticit y at the cellular level is described as a continuous process in reaction to neurona l activity (Teter & Ashford, 2002, p. 405). Long-term cortical modification is k nown as macrostructural plasticity. Continuous neuroplasticity at the cellular level may underlie functional and structural cortical re-organization. Ne ural representations are dynamic and continuously modified by experiences including intens e auditory and peripheral se nsory stimulation generated by music training and performance (Merzenich et al., 1996; Pantev et al., 2003; Pantev, Engelien, Candia, & Elbert, 2001). Thus, neurophysiological re search comparing musicians and nonmusicians serves as an exce llent tool for the st udy of neuroplasticity (Mnte et al., 2003; Schlaug, 2001; Zatorre, 2003). Microstructural plasticity. Microstructural plasticity is observed when there is a change in the efficiency of transmission at a cellular level; that is, changes in the firing probabilities, changes in activa tion strength between synapses or structural adjustments in the connections between groups of neurons (Calford, 2002; Robertson & Murre, 1999). Functional systems in the brain retain flex ibility at the cellula r level throughout life (Edelman, 1987; Plante, 2000). New variations of synapses continue to occur between interacting neural networks and w ithin hierarchies of networks.
26 Neurophysiological studies of musi cians who play instruments support microstructural changes in the brain as a result of music training (Schlaug, 2001). Pascual-Leone et al. (1995) showed that as subjects learned a five -finger exercise on the piano over the course of five days, the cort ical representation on the sensorimotor area targeting the long finger flexor and extensor muscles enla rged. Thus, training-induced microstructural plasticity can occur with in a short time period (Pantev et al., 2001; Pascual-Leone et al., 1995; Tr ainor, Shahin, & Roberts, 2003 ). Rauschecker (200l) views training-induced neural change as a common occurrence. He states, Of course, even the ability to learn and memorize a simple tune is an expression of the brains ability to change with musical experience (p.330). Microstructural effects of auditory training. Whether musical abilities of and neural differences in musicians are due exclusively to learning, or whether these differences reflect innate capacities enhanced by early music training is unknown (Gaser & Schlaug, 2003; Pantev et al., 1998; Schlaug, 2001; Trainor et al., 2003; Zatorre, 2003). To examine the neuroplastic effect of audito ry experience independent of innate abilities that may be present in musicians, resear chers have studied auditory training in nonmusicians (Brattico, Tervaniemi, & Picton, 2003; Menning, Robert s, & Pantev, 2000; Trainor et al., 2003). Electroencephalography (EEG) has been used to compare the effects of musical context and musical syntax on neural respons es of pitch perception in musicians and nonmusicians (Brattico, Ntnen, & Terv aniemi, 2001; Koelsch, Schmidt, & Kansok, 2002; Lopez et al., 2003). EEG is a non-invasive method of recording electrical activity
27 and changes to real-time cognitive processi ng. Event-related potentials (ERPs) reflect this electrical activity in waveforms with positive and negative peaks. An ERP is a sequence of voltage changes that are time -locked to a stimulus event (Koelsch & Friederici, 2003; Mody, 2004). Trainor, Shahin and Robert s (2003) compared seven 4year old children taking Suzuki music lessons with 6 age-matched c ontrol children who were not studying music. ERPs were recorded when subjects just bega n Suzuki lessons and one year later. The children listened to three di fferent tones matched in loudness: violin, piano, and pure tones. The P1-N1-P2 complex was examined. The P1 occurs approximately 50 100 ms af ter the onset of an auditory stimulus and is interpreted as an indi cator of preferential attenti on (Key, Dove, & Maguire, 2005). It is frequently associated with audito ry inhibition and s uppression of unattended information (Key et al., 2005). The N1 is a negative wave peak typically recorded at about 100 ms after the stimulus onset. It reflects activation of the large neuronal population in regions of the auditory cortex on the superior surface of the temporal lobe. The N1 is sensitive to attention and may be augmented by plasticity occurring either cortically or at subcor tical sites projecting to the auditory cortex (Menning et al., 2000). The P2 occurs between 150-275 ms after stim ulus onset and is sensitive to physical parameters of the stimulus, such as pitch and loudness (Key et al ., 2005). The P2 has been found to differ between musicians and nonmusicians and has been enhanced in nonmusician adults with aud itory training (Bosnyak, Eaton, & Roberts, 2002 as cited in Trainor et al., 2003).
28 In the Trainor et al. study, there were no significant differenc es between the two groups of children prior to music training. For all children, ERP responses were most robust to piano tones with clear P1, N1, and P2 components suggesting an increased cortical response to sounds rich in harmonics. Responses to pure tones were least robust and only clear P1 components were present. ERP responses were measured again when the students were 5 years of ag e (i.e., after one year of traini ng for the Suzuki students). ERP responses differed between the groups onl y for the piano tones. The P2 response was stronger and the N1 component emerged ear lier in the Suzuki-tra ined children. Thus, auditory cortical responses can be differen tiated between groups of children as young as 4 or 5 years old. ERP differences between the children lend strong support for cortical changes as a result of music training; however, the authors ca ution that the influence of genetic factors cannot be dismi ssed (Trainor et al., 2003). Menning, Roberts, and Pantev (2000) investig ated plasticity of th e auditory cortex in nonmusicians through intensive frequency discrimination traini ng. Ten right-handed volunteers were trained for 15 days to detect progressively smaller deviant stimuli. Frequency discrimination diminished to about 30% of its initial value and thresholds stabilized in about 10 sessions near 2 Hz. Data from ma gnetic electroencephalography (MEG) indicate increased streng th of pre-attentive audito ry neural responses during training and three weeks after training compar ed to pre-training data. Thus, results support training-induced neuroplas ticity of the auditory sy stem and suggest that the neural processes responsible for detection of pitch irregu larities may be enhanced by auditory discrimination training.
29 Brattico, Tervaniemi, and Picton (2003) examined whether the pre-attentive auditory neural response to different t one frequencies can be affected by brief discrimination training at one specific freque ncy. Eighteen volunteers of mixed gender and mixed handedness received a one-hour tr aining session. Immediate post-training EEG data indicated that the co rtical response to the learne d tone (1062 Hz) and repeated tone (1000 Hz) was as large in amplitude as be fore training; however, the auditory neural responses to the other test tones were diminished suggesting a counteraction of the sensitization effect by activation of neurons that previously did not respond. The authors conclude that these plastic changes may underl ie the long-term modification of cortical representation observed in mu sicians (Pantev et al., 2003). Macrostructural plasticity. A common finding across most skill acquisition studies is the functional enlargement of th e cortical representation area underlying a particular skill (Gaser & Schlaug, 2003). Also known as map extension, cortical representation demonstrates the flexibility of a functional brain re gion to enlarge on the basis of skilled practice or frequent exposure to a stimul us (Grafman, 2000). Elbert and colleagues (1995) found altere d representation for the fingers in the somatosensory cortex for professional musi cians who play stringed instruments. Specifically, neural imaging revealed increase d cortical representation of the fingers of the left hand in skilled violinists. These researchers also noted that functional enlargement of cortical representation was inversely correlated with the age at which musicians begin to practice suggesting microstructural adaptation evolving to macrostructural changes. Thus, consistent and intense practice of bimanual finger
30 sequences has been shown to alter the structur e of a musicians primary motor cortex and somatosensory cortex, especially when traini ng occurs during a cri tical period of brain development. Using magnetoencephalography (MEG), Pantev et al. (1998) compared the location and strength of the electrical source for neural representati on of piano tones and pure tones between musicians and nonmusicia ns. Supporting earlier research (Elbert, Pantev, Wienbruch, Rockstroh, & Taub, 1995), mu sicians demonstrated an increase in the size of cortical representation for the processing of piano tones. Once again, beginning age of music training was inversely correlated with neuronal representation. That is, the earlier the initiation of musical practice, the stronger the neuronal response was to the piano tones. The Musicians Brain Anatomical Differences The lifelong ability to adapt to environm ental demands and sensory stimulation is grounded in the dynamic capacity of the human brain to modify its structure and function. A musicians brain provides oppor tunities for researchers to study the interactions between inherent neurophysiologi cal distinctions and the impact of music training on structural adap tation and development. Corpus callosum. The corpus callosum is the main interhemispheric fiber tract responsible for interhemis pheric integration and comm unication. Structural and functional maturation of the corpus callo sum extends into late childhood and early
31 adolescence. Maturation coincides with the termination of the corpus callosum myelination cycle (Yakovl ev & Lecours, 1967). The anterior portion of the corpus callosum is the last subregion to mature. This portion contains fibers mainly from frontal motor-related regions and pre-frontal related regions. Functional magnetic resonance imaging (fMRI) reveals that th e anterior half of the corpus callosum is significantly larger in musicians compared to nonmusicians (Schlaug, 2001). This difference in callosal size may be due to (a) an increase in the number of fibers, (b) a larger proportion of thicker myelinated fibers with fast interhemispheric transfer, or (c) fibers with thicker axons or more axon collaterals (Schlaug, 2001). During music training for an in strument requiring the use of both hands, such as a violin or a piano, intense inte rhemispheric communication is necessary for management of complex bimanual motor sequenc es. Thus, music training is implicated in the determination of callo sal fiber size and composition. Furthermore, the anterior corpus callosum is significantly larger in musicians who begin training prior to age 7 compared to musicians who begin later or to nonmusician cont rols (Schlaug, 2001). Cerebral cortex. Neural imaging studies examin ing the primary motor cortex, somatosensory cortex and auditory cortex prov ide evidence of macrostructral differences between instrumental musicians (violin and keyboard players) and control subjects (Elbert et al., 1995; Pantev et al., 1998; Schlaug, 2001). All st udies found a strong negative relationship between the age that musi c training begins and the degree of neural alteration; that is, the earlier in life that pr actice begins, the greater the structural change.
32 These differences support the concept that intensive music training influences neural changes. Schlaug and colleagues (2001) examined th e intrasulcal length of the posterior bank of the precentral gyrus (ILPG) as a gro ss anatomical marker of the primary motor cortex. Data from functional magnetic resona nce images indicated a significantly greater intrasulcal length in the right hemisphere for musicians. There was no significant between-group difference in the left ILPG. Sc hlaug suggests that th e longer right ILPG in musicians is a training-induced adaptation of the motor area for greater control of the nondominant hand. Correlation analyses s upport a strong relationship between mean intrasulcal length and age of commencement of music training. Gray matter volume. Gaser and Schlaug (2003) compared the brain structures of professional musicians, amateur musici ans, and nonmusicians using voxel-based morphometry (VBM). VBM is a fully automatic technique for comput ational analysis of differences in local gray matter volume. Voxe l clusters are overlaid on a rendered cortex surface of a selected single subject. VBM pr ovides high resolution anatomical images of the whole brain using a magnetization prepar ed rapid acquisition gr adient echo sequence (Gaser & Schlaug, 2003). A significant positive correlati on between musician status and increase in gray matter volume were found in the peri-rolandic regions including the primary motor and somatosensory areas, pre-motor areas, anterior superior parietal area, and the inferior temporal gyrus bilaterally. A positive correlation indicates volume was highest in professional musicians; that is those with the most training, intermediate in the amateur
33 musicians, and lowest in those with no pr evious music training. Positive correlations between gray matter volume and musician stat us were found in the left cerebellum, left Heschls gyrus, and left inferior frontal gyrus. No significant correlation between white matter volume and musician stat us was indicated. It was su ggested that either the VBM is insensitive to white matter difference or that most of the presumed plastic changes occur in the cerebral gray ma tter (Gaser & Schlaug, 2003). Cerebellum. The cerebellum comprises only 1/10th of the brains total volume; however, the number of cells in the human cer ebellum exceeds the total number of cells in the cerebral cortex by four times (Ande rson, Korbo, & Pakkenberg, 1992). Because of its role in motor learning, movement coordi nation, and timing of sequential movements, it was questioned whether the cerebellum is st ructurally different between musicians and nonmusicians (Schlaug, 2001). F unctional magnetic resonance imaging data revealed a significantly higher mean relativ e cerebellar volume (5%) for male musicians (strings and keyboard players) compared to male nonmu sicians. A positive trend was noted between intensity of music training (practice time per day and across a lifetime) and relative cerebellar volume. Summary Music is a rule-governed, universal ch aracteristic of human culture acquired through implicit and explic it experience. The study of musi c and its neural correlates for perception and performance has yielded insights into the st ructural organization of the human brain. Neurological and anatomi cal differences between musicians and
34 nonmusicians have been reliably measur ed with established neurophysiological techniques (Elbert et al., 1998; Pantev et al., 2003; Schl aug, 2001). These distinctions serve as indicators of possibl e genetic factors or training -induced changes in neural structure indicative of neural plasticity (Zatorre, 2003). Because of the concentrated training and skill acquisition that a musician pursues from an early age, it is argued that neural development occurs differentially in response to performance demands; that is, macrostr uctural changes take place as a result of widespread microstructural adaptations (Gaser & Schla ug, 2003; Pantev et al., 1998; Pascual-Leone, 2001; Schlaug, 2001; Zatorre, 2003). The strong association between beginning age of music training and degree of cortical adaptation further supports the argument that these changes evolve over time as a consequence of training. It is yet unknown whether these musical abilities and neurological/anatomical differences of musicians are due exclusively to learning, or whether these distinctions reflect innate abilities and capacities that are advanced by early exposure to musi c (Gaser & Schlaug, 2003; Pantev et al., 1998; Schlaug, 2001; Zatorre, 2003). While a review of the literature reveal s comprehensive evidence supporting neural plasticity in trained instru mental musicians, little is known regarding trained vocal musicians. Trained vocal musicians present with exceptional abilities to perceive and perform music and adhere to the same rigor ous training as other musicians; however, there is a paucity of comp arative research and an in sufficient understanding of physiological variables in this population.
35 Auditory Pitch Perception Auditory perceptual abilities fo r music and speech emerge from neurophysiological structures and functions. Th e abilities to perceive and discriminate the components of music are regarded by ma ny music educators as fundamental abilities and implicit skills of a succe ssful performer (Bentley, 1966; Geringer, 1983). Carl Seashore (1919), psychologist and author, argued that th e capacity to hear pitch differences is a fundamental capacity in mu sical talent, and upon it rests most of the powers of appreciation and expression in musi c one must be guided by such hearing in playing and singing (p. 42, as cite d in Pedersen & Pedersen, 1970). Acoustical Perception of Music Acoustical perception of music includes th e basic perceptual qualities of timbre, loudness, and pitch as a function of time (S undberg, 1994). Timbre, also referred to as resonance, adds richness to a tone (Sundberg, 1994; Watts, Barnes-Burroughs, Adrianopoulos, & Carr, 2003). While there is no clear-cut definition of timbre, Sundberg writes, Two tones differ in timbre if they are similar in pitch a nd loudness and still do not sound similar (p. 107). For example, tones perceived from a trumpet and a piano may have the same pitch and loudness, yet th ey are perceived as two different sounds. This difference is the timbre and depe nds on the length and shape of the resonating tract. Loudness is the perceptual co rrelate of the intensity or magnitude of the acoustic stimulus (Gelfand, 1998; Sundberg, 1994). Sounds with low intensity are perceived as
36 soft and sounds with high inte nsity are perceived as loud; however, there is not a oneto-one correlation between loudness and inte nsity. Perceived loudness of complex tones depends on the critical bands of hearing (Sch arf, 1970). Tones of similar amplitudes fall into the same critical band so that the t ones cannot be heard individually; however, as adjacent bandwidths are stimulated, pe rceived loudness increases (Gelfand, 1998; Sundberg, 1994). Perceived loudness is the su mmed loudness of all the critical bands activated by a particular tone. Pitch is the perception related to fre quency and corresponds to the fundamental frequency of the lowest frequency part ial (Gelfand, 1998; Sundberg, 1994). Perceived pitch gets higher as frequency increases; how ever like loudness, there is not a simple oneto-one correlation between pitch and freque ncy. The pitch perception of complex sounds relies on the processing of th e fundamental frequency and it s components or harmonics. Accurate auditory pitch di scrimination across a distribut ed frequency range is a prerequisite for the perception of speech and music (Novitski, Tervaniemi, Huotilainen, & Ntnen, 2004). For the perception of mu sic, pitch differences are said to be perceived categorically. In other words, slig ht variations in fre quency do not affect the perceived tone or note, or the musical interv al (Sundberg, 1994). Within a small range of variation, a change of frequenc y has no effect on our perception or classification of pitch; however, at the border of a frequency ra nge, a minor shift ra dically changes the perception from one category to another. This phenomenon is known as categorical perception. In other words, not all changes in frequency are perceived. In order for sounds to be detected as diffe ring in pitch, the frequency diffe rence must be at least equal
37 to the subjects difference limen; that is, frequency discrimination threshold (Gelfand, 1998). The perception of pitch is fundamentally different when we listen to speech or music. In music, pitch changes are percei ved categorically into a limited number of musical intervals. Most melodies written in Western culture are generally written with small pitch intervals approximately 1/6th to 1/12th of an octave (Ayotte, Peretz, & Hyde, 2002). In speech, the quantity of pitch change is perceived in a continuous fashion and the variations are larger than one-half an octa ve to convey relevant information (Ayotte et al., 2002; Sundberg, 1994). This perhaps expl ains why Seashore (1919) thought that the auditory perception of minute pitch differen ces is a fundamental capacity in musical talent, and upon it rests most of the powers of appreciation and expression in music (p. 42, as cited in Pedersen & Pedersen, 1970). Despite its obvious importance, frequency discrimination is one of the least investigated psychoacoustic abilities in musicians, (Kishon-Rabin, Amir, Vexler, & Zaltz, 2001). Cortical Models of Pitch Perception Modular theory. Researchers have proposed that speech and music may be processed by two distinct systems and refer to this as the theory of modularity (Peretz & Coltheart, 2003; Tervaniemi & Brattico, 2004) or specific domain (Leiberman & Whalen, 2000). According to Fodor (1983, 2001), mental modules have the following characteristic properties: (a) speed of ope ration, (b) automaticity, (c) domain-specificity, (d) information encapsulation, (e) neural specif icity, and (f) innateness. Each property is
38 typical, but not a required feature of a modular system. Information encapsulation (Fodor, 1983) and domain-specificity (Peretz & Coltheart, 2003) are considered two of the more important characteristics. Info rmation encapsulation means that processing within a mental module is immune from influence of the central system (Fodor, 1983). Domain specificity implies that the specifi c operation of a module is restricted to a limited domain of input and output (Peretz & Coltheart, 2003). Peretz & Coltheart (2003) propose a module that is spec ific to the processing of music. This module may contain smaller syst ems specific to different aspects of music, but not necessarily restricted to music. The model is based on the premise that music is an evolutionary and unique cognitive func tion with dedicated and separate neural substrates (Peretz & Coltheart, 2003; Tervaniemi & Brattico, 2004). Selective impairment and sparing of musical abilities ha ve been found in neurologically impaired individuals. For example, there are indi viduals who can no longer recognize musical melodies, but for whom the ability to r ecognize spoken words and environmental sounds are normal; or conversely, spoken words are not recognized, but musical melodies are easily identified (Ayotte, Peretz, & Hyde, 2002; Peretz & Coltheart, 2003). The evidence of a double dissociation supports the modularity theory and points to the existence of separate and dedicated neural circui try for music and speech processing. Peretz and Coltheart (2003) propose that a ll auditory stimuli are first processed in an acoustic analysis module. Within this module, all informa tion is received by all submodules and processed in parallel. It is assumed that activation of the music or the language processing modules is determined by the aspect of the input to which a module
39 is tuned. Thus, speech and music depend on sp ecific and dedicated neural circuitry (Lieberman & Whalen, 2000). The music input modules are thought to be organized in two parallel and independent subsystems: (a) analysis of pitch cont ent, including contour and intervals; and (b) analysis of temporal content, incl uding rhythm and duration. These modules may process in parallel with language processing modules and may be connected by information pathways to other m odules outside of the a uditory cortex, such as memory modules and/or perceptual modules. Parameter theory. The domain-specific theory ma y be challenged by the concept of a parameter-specific lateralization model. This model actually overlaps with the domain-specific paradigm. It proposes neural specializations for processing the acoustic parameters of speech and music based on the premise that auditory hemispheric asymmetries for temporal and spectral pro cessing have evolved as a consequence of functional specialization and not domain-speci ficity (Zatorre et al., 2002). Speech and music stimuli differ in their acoustic structure and thus in their processing requirements. Whereas the analysis of speech requires good temporal resolution to process rapidly changing formants, it can be argued that mu sic processing requires good pitch resolution (Zatorre et al., 2002). Neuroimaging and el ectrophysiological da ta support neural specialization and indicate proc essing distinctions between the left and right auditory cortices; that is, the left hemisphere is favor ed to process fast temporal information, while the right hemisphere is more active for sp ectral resolution (Dalla Bella & Peretz, 1999; Tervaniemi & Hugdahl, 2003; Zatorre et al., 2002).
40 Neural Specialization for Pitch Processing Evidence from neural imaging is accumulati ng to establish structural differences in brain organization between musician s and nonmusicians (Schlaug, 2001) and functional differences such as increased co rtical representation (Elbert et al., 1998; Pantev et al., 2003). Electr ophysiological studies (EEG, MEG) indicate that temporally and spectrally complex sounds are automatically processed by the human auditory cortex and this processing differs between sounds of speech and music and between the cerebral hemispheres (Koelsch, Schrger, & Terv aniemi, 1999; Tervaniemi, 2001). Pitch processing may also be broken into a hierarchy of auditory neural functions from lowlevel activities including the discrimination of two sounds or detecti on of pattern change to high-level activities includi ng discrimination of pitch contours and organization of tones into melodies and harmony. High-level pitch processing. Earlier neuropsychologi cal studies of music processing investigated subjects with brai n lesions and reported that neural pitch processing can be dissociated into different lateralizati on patterns and selectively disrupted by cortical lesions Research designed to assess the contribution of each hemisphere to high-level pitch processing f ound that auditory discrimination of melodic pitch patterns is genera lly more affected by damage to th e right superior temporal area than to the left superior temporal area (P eretz, 1990; Zatorre, 1988). Thus, although there was a substantial contribution of the left hemisphere, results indicated overall right hemisphere superiority for melodic pitch processing.
41 More recent studies have examined highlevel pitch processing using functional magnetic resonance imaging (fMRI). Gaab and Schlaug (2003) examined whether differences in perceptual and/or cognitive strategies alone can explain functional brain difference between musicians and nonmusicians The nonmusicians were selected from a larger sample group and matched with musi cians based on their performance on a pitch memory task. Subjects listened to a sequen ce of 6-7 tones and were asked to decide whether or not the last or sec ond to last tone was the same or different from the first tone. For both groups, fMRI images indicated bila teral activation of th e superior temporal gyrus, supramarginal gyrus, infe rior frontal gyrus, and superior parietal lobe. Despite matching the two groups on a performance scor e of pitch memory, musicians had greater bilateral activation of superior parietal areas, more activ ation of the supramarginal gyrus (SMG), and greater activity in the right inferior frontal lobe These results concur with previous PET scan data (Zatorre, Evans, & Meyer, 1994) that also support the hypothesis of frontal lobe activation for the analysis of higher order pitch processing. Gaab and Schlaug (2003) conclude, Musicians activate a network that includes auditory short-term memory regions (e.g. SMG) and regions im plicated in visual-spatial process (e.g. superior parietal cortex). Nonmusicians s eem to rely more on a network that includes brain regions important for pitch discrimi nation (e.g. Heschls gyrus) and traditional memory regions (e.g. hippocampal gyrus) (p. 2294). Schmithorst and Holland (2003) compared the neural correla tes of auditory processing for melody and harmony between musicians and nonmusicians on a passive listening paradigm. On the melody processi ng task, musicians had significantly greater
42 activation of the inferior pari etal lobes and superior front al gyrus bilaterally, and in contrast to the above findings, the left inferior frontal and superi or temporal gyri. This variation in hemispheric activation of the inferior frontal lobe may be due to the differences among the tasks. In the study by Schmithorst and Holland, subjects listened passively to a popular melody which may have activated semantic memory in the left frontal hemisphere; whereas, Gaab and Schla ug required subjects to make active choices after listening to an unfamiliar sequence of tones. Concurring with earlier reports, Schmithorst and Holland found th at bilateral ac tivation in the anteri or portion of the superior temporal gyrus was robust for bot h musicians and nonmusicians, supporting a contribution of working memory for melodi c processing. For harmonic processing, both subject groups had activation of the occipital lobe suggesting that exposure to harmonic progressions integrates with visual associa tive functions in the br ain. In addition to activation of the occipital lobe, musicians ha d activation of the pari etal-temporal regions near the angular gyrus. Schmithorst and Ho lland (2003) conclude that extensive music training promotes the recruitment of differe nt neural networks to process harmony and melody. Brattico, Ntnen, and Tervaniemi (2001) used electroencep halography (EEG) to investigate context effects on pitch per ception in musicians and nonmusicians by measuring the mismatch negativity (MMN) response. The mismatch negativity is a component of ERPs and reflect s an auditory change detec tion process based on neural representations of acoustic repe titions or regularities indepe ndent of active attention to the task. While quietly reading, subjects were presented with a large pitch change in three
43 contexts: isolated sounds, a sequential pattern with familiar tones from Western culture, and a sequential pattern with intervals unfamiliar to the su bjects. For both groups, the MMN amplitude was greater when the pitch change occurred among sequential patterns within a familiar scale than within an unfamiliar scale and greater when the pitch change occurred within an unfamiliar scale than am ong single tones. Musicians had a faster neural response for pitch changes than the nonm usicians; that is, a shorter latency for the MMN response. Koelsch, Schmidt, and Kansok (2002) inve stigated the influe nce of long-term musical experience on the processing of chords presented within a complex musical context by examining the early right anterior negativity (ERAN) component of ERPs. The ERAN response is triggered by violations of complex musical re gularities and is maximal around 200 ms following stimulus ons et (Koelsch, Gunter, Friederici, & Schrger, 2000). Musicians and nonmusicians listened to harmonica lly appropriate and inappropriate chords. That is, the successi on of chord functions and harmonic relations followed expected patterns of music syntax in classic Western tonal music or did not follow these patterns. ERP data revealed that harmonically inappropria te chords elicited an ERAN in both subject groups; however, a significantly larger ERAN was elicited in the trained musicians. Koelsch, Schmidt, and Kansok conclude that because of their music training, musicians have more exp licit memory representations of harmonic relatedness and therefore, are more sens itive to violations of music syntax. Lopez et al. (2003) compared ERP res ponses of musicians and nonmusicians on five tasks of increasing complexity of musi cal sequences. Stimulus tasks included five
44 oddball paradigms, including single tones in a sequence, 3 simultaneous tones (chord), 3 consecutive tones (arpeggio), a familiar song without words, and a sung song. In a basic oddball paradigm, sometimes referred to as a change-detection paradi gm, infrequent and unpredictable deviant stimuli are presented among frequent standard stimuli. The proportion of stimuli is typically 80% sta ndard and 20% deviant (Sams, Paavilainen, Alho, & Ntnen, 1985). All odd-ball paradigm s elicited (a) a response of the primary auditory cortex, N1; (b) a frontotemporal nega tivity at 200 ms, the mismatch negativity to the deviant stimuli (MMN); and (c) a late positive component be tween 350-450 ms of latency, P3. The P3, also referred to as th e P300, reflects an invol untary attention switch towards the deviant or novel sound in an i gnore and attend conditi on (Novitski et al., 2004). It is typically elicited by an odd-ball paradigm and re lies on active participation of the subject. The P3 is thought to be an i ndicator of memory maintenance or updating (Donchin & Coles, 1988). The N1 peak measurements were distinct for all paradigms but not significantly different between the groups. There were cl ear MMN responses for both groups in all paradigms; however, those with musical ab ility responded with significantly greater amplitudes and shorter latencies. Similarly, this subj ect group had the strongest P3 for the arpeggio task (3 consecutive tones) and the complex melody task, suggesting an influence of musical skill, since the P3 responses for the subject groups did not significantly differ on the simpler paradigms of sinusoidal tones and familiar melody. These studies agree that auditory pitc h perception for high-level activities, including discrimination of pitch contours a nd organization of tones into melodies and
45 harmonies, is influenced by musical context a nd syntax (Brattico et al., 2001; Koelsch et al., 2002). Data from ERP res earch indicate that even nonm usicians have an innate knowledge of musical regularitie s; that is, listeners expect specific musical events to occur (Koelsch et al., 2000). However, musici ans consistently have greater sensitivity to violations of musical syntax secondary to extensive music training. Years of music lessons and practice create a larger number of explicit memory repr esentations, implying that neural mechanisms responsible for pro cessing musical syntax can be influenced by experience and training. Low-level pitch processing. Low-level auditory pitch processing, such as the discrimination of two sounds or the detection of pattern change, have been investigated by a variety of approaches including psycho acoustics, behavioral lesion studies, and electroencephalography (EEG). Spiegel and Watson (1984) compared the performances of orchestral musicians and nonmusicians on tasks of audito ry frequency discrimination in an attempt to relate performance to musical background. Surprisingly, on a task of frequency discrimination for single tones, one-half of the nonmusicians attained thresholds almost as low as the musicians. Th e researchers suggested that innate auditory processing advantages may be present in some persons who have not had music training. This finding is questionable, however, sin ce the subjects in the control group were not screened for previous musical experience or familiarity with psychoacoustic procedures. Kishon-Rabin, Amir, Vexler, and Zaltz (2001) compared frequency discrimination thresholds, also referred to as the difference limen for frequency (DLF) between instrumental musicians and nonmusi cians. The nonmusicians had no previous
46 music training or experience with psychoac oustic studies. Although normative values for DLF in the general population have not been established, average DLF thresholds for pure tones between 500 Hz and 2k Hz using an adaptive two-inte rval forced choice paradigm (2IFC) have approached 1 1.5% ( Moore, 1989). Kishon-Rabin and colleagues compared DLF thresholds using two threshold estimation procedures: twointerval forced-choice method (2IFC) and threeinterval forced-choice paradigm (3IFC). Stimuli consisted of three sets of digitally generated pure to nes. Each set contained one reference tone and 20 different comparison tone s. The comparison tones varied in 0.5 Hz steps for a reference tone of 250 Hz and in 1 Hz steps for 1000 and 1500 Hz. The minimal detectable changes in frequency ( f) were transformed to relative DLF thresholds in percent ( rel DLF%= f/f x 100). Findings indicat ed that both groups had significantly smaller DLFs in the 2IFC paradi gm compared to the 3IFC. Consequently, it was suggested that auditory memory plays a role in frequency discrimination tasks. Musicians had significantly smaller values of rel /DLF % than nonmusicians. The mean DLF for musicians was approximately half the value of the nonmusicians suggesting that auditory pitch discrimination is in fluenced by years of music training. Neuropsychological research has examined the specificity of pitch and temporal discrimination, assuming that damage to a spec ific area of the brain leads to a specific change in behavior which yields clues to the function of the damaged area (Johnsrude, Penhune, & Zatorre; 2000, Ligois-Chauvel, Peretz, Babai, Laguitton, & Chauvel, 1998; Zatorre, 2003). Ligois-C hauvel et al. (1998) compared 65 right-handed patients who had unilateral temporal cortectomies seconda ry to intractable ep ilepsy with a control
47 group. Sequences of simple musical phrases with variations in either pitch or temporal dimensions were presented. Participants j udged whether two phrases were same or different. Results indicated that removal of the posterior superior temporal gyrus (STG) resulted in the greatest deficit in the pitchbased tasks, implying that the posterior regions of the STG may be specialized to compute certa in specific aspects of pitch patterns. The authors could not suggest which STG is specialized since th ey combined all leftand right-sided cases in their analysis. Johnsrude, Penhune and Zatorre (2000) paralleled the above study. Magnetic resonance imaging (MRI) identif ied lesions as extending or not extending into Heschls gyrus. Heschls gyrus is the primary audito ry cortex and a subdivision of the superior temporal gyrus (STG). All participants perf ormed a simple pitch discrimination task, for which the subject decided if two pure tones were the same or different, and a pitch direction judgment task, for which the subjec t decided whether the first tone was higher or lower than the second. On the simple pitch discrimination task, there were no significant threshold differences between the co ntrol subjects and patients with either a left or right temporal lobe excision that encroached upon the superior temporal gyrus (STG). However, thresholds for a pitch direc tion judgment task were significantly higher for patients with excision of the right temporal lobe that encroached upon the lateral portion of the STG. Conversely, normal subj ects and patients with lesions in the left temporal lobe regardless of the extension into the STG were unimpaired on the tasks. Patients with lesions in th e right temporal lobe, but excluding the STG, were also unimpaired on the pitch direction task, thus demonstrating a speci alization of pitch
48 function linked to the right audi tory cortical areas and specifi cally to the right superior temporal gyrus. Using electroencephalography (EEG), neur al sound processing can be probed within a millisecond of accur acy by recording the event rela ted potential (ERP). Event related potentials of the audito ry cortex are also referred to as auditory evoked potentials (AEPs). EEG studies on automatic neural en coding of music prior to conscious attention often examine the following AEP compone nts: P1, N1, P2, or MMN. Shahin, Bosnyak, Trainor, and Robert s (2003) compared auditory evoked potentials, N1 and P2, between musicians and nonmusicians as they pa ssively listened to the presentation of violin tones, piano tone s, and pure tones matched in fundamental frequency and loudness. Effects of group (violinists, pianists and control) and stimulus (violin, piano, pure tones) were evaluated. Compared to the nonmu sician control group, the musicians had larger N1 and P2 amplit udes to all three types of tonal stimuli indicating stronger pre-attentiv e auditory responses. There were no response differences between the violin and piano musicians. Piano tones evoked larger N1 responses for musicians and nonmusicians compared to pure tones and violin tones. As reported previously in this chapte r, Trainor, Shahin and Roberts (2003) examined the P1-N1-P2 complex in musically trained and untraine d 4and 5-year old children. AEPs were compared in response to violin tones, piano tones and pure tones. For all children, ERP responses were most robust to piano tones with clear P1, N1, and P2 components suggesting an increased cortical response to sounds rich in harmonics. In the child musicians, the P2 response was st ronger and the N1 component emerged earlier
49 lending strong support for cortical plasticity as a result of music training; however, the authors cautioned that the influence of genetic factors cannot be dism issed (Trainor et al., 2003). Tervaniemi (1993) compared the pitch disc rimination accuracy of musicians with and without absolute pitch (AP) using a mi smatch negativity (MMN ) paradigm. An MMN response is elicited when a sound is encountered that doesnt match memory representations; that is, it is evoked by a different or deviant stimulus which occurs infrequently among standard or frequent stimuli. It is termed pre-att entive, as it detects aspects of acoustic information that are enco ded without the consci ous attention of the listener, such as pitch, durati on, etc. (Tervaniemi, 2001). A strong MMN refers to large amplitude and short latency. Absolute pitch is characterized as the ability to identify by verbal label (musical note name) the pitch of any sound without re ference to another sound, or by producing a musical tone when gi ven the musical note name (Zatorre, Perry, Beckett, Westbury, & Evans, 1998). Stimuli consisted of pure tones and synthesized piano sounds familiar in Western culture. The deviant stimuli were off-scale (i.e., dissonant). Contrary to expectations, th e MMN responses did not differ significantly between the musician groups, or between onor off-scale tones. Interestingly, both groups had greater amplitudes and earlier late ncy to piano sounds compared to sinusoidal tones suggesting that pitch discrimination is facilitated by the presence of the harmonic partials. Tervaniemi, Ilvonen, Sinkkonen et al. ( 2000) in a series of MMN studies, systematically investigated the facilitating effects of overtones on the accuracy of pitch
50 discrimination. Initially, they found that su bjects pitch-MMN was stronger for complex sounds rather than the pure t ones. The next studys objectiv e was to determine if adding harmonic partials would further enhance the MMN response (Tervaniemi, Schrger, Saher, & Ntnen, 2000). ERP data reveal ed that the MMN amplitude was enhanced for complex sounds when compared to pur e tones, but there were no significant differences between spectrally rich tones having 3 to 5 partia ls. This suggests that few harmonic partials are needed to retrieve the neural representation underlying pitch discrimination. Subsequent to the previous studies, Novitski, Tervaniemi, Huotilainen, and Ntnen (2004) systematically compared the neural and behavioral accuracy of frequency discrimination across a wider ra nge from 250-4000 Hz. The sound structure (pure versus harmonic tones) and magnitude of frequency change were varied. The harmonic tones elicited stronger MMN responses than did pure sinusoidal tones in all frequency bands. The addition of only two part ials to the pure tones caused an increase of MMN amplitude. In the behavioral study, the same subjects indicated whether tones presented in pairs differed in pitch. ERP la tencies and amplitudes correlated with the reaction time and hit rate with the highest correlation occurring between MMN amplitude and hit rate (r = 0.8). Since the MMN detects aspects of acoustic information that are encoded without the conscious attention of the listener. It ma y be used as a means to examine traininginduced changes in auditory neural res ponse (Mody, 2004; Tervaniemi, 2001; Trainor, McDonald, & Alain, 2002). Koelsch, Schrge r, and Tervaniemi (1999) compared pre-
51 attentive auditory responses of musicians a nd nonmusicians to investigate the influence of musical expertise on the brains automa tic pitch change-detec tion (MMN). Major chords and single tones were presented to musicians and nonmusicians under ignore and attend conditions. Slightly disharmonic c hords, presented among perfect major chords elicited a distinct MMN in pr ofessional musicians, but not in nonmusicians. Thus, the musicians automatically detected differen ces in auditory information that was undetectable for nonmusicians. Consensus among published research sugge sts that low-level auditory pitch processing, such as the discrimination of two sounds or the detecti on of pattern change, are linked to the right auditory cortical ar eas and specifically to the right superior temporal gyrus. Pitch discrimination is facilitated by the pres ence of the harmonic partials; however, only a few harmonic partials are necessa ry for retrieval of neural representations. While innate auditory processing advantages may be present in some persons who have not had music training, curren t research indicates superior pre-attentive auditory skills for musicians and supports th e implication that long-term music training modifies pre-attentive neural processing of acoustic input. Vocal Pitch Control Evolution of Singing The evolutionary origin of singing in the human species appears to be an intriguing occurrence since there is no clear-c ut adaptive function for singing (Hauser &
52 McDermott, 2003). Unlike language, which a llows us to communicate our thoughts to others, music has no obvious functiona l outcome. Charles Darwin wrote, As neither the enjoyment nor the ca pacity of producing musical notes are faculties of the least use to man in reference to his daily habits of life, they must be ranked among th e most mysterious with which he is endowed. (1871, p. 878 as cited in Hauser & McDermott, 2003) Darwin and others propose that music evolved as a sexually selected system, designed to attract mates and signal mate quality (Darwin, 1871 as cited in Hauser & McDermott, 2003; Miller, 2000). Thus, shap ing human emotions may be a possible adaptive function of music and a basi s for the evolution of singing. Laryngeal Anatomy for Pitch Control The biological organ for phonation is the la rynx. The larynx is only one part of an anatomical network responsible for voice pr oduction. This network also includes the respiratory system which is comprised of the bronchi, lungs, and trachea, and the supralaryngeal vocal tract which includes th e pharynx and the oral and nasal cavities. Tradition holds that the larynx is innervated by two major branches of the vagus nerve; that is, the superior laryngeal nerve and the recurre nt laryngeal nerve. However, evidence suggests that cranial nerve XI, spinal acce ssory, may also contribute to laryngeal innervation (Webster, 1999; Zemlin, 1999). A dditionally, the muscles that suspend the larynx in the cervical region, in cluding the extrinsic laryngeal muscles, are innervated by several cranial nerves (V tr igeminal, VII facial and XII hypoglossal). The coordination
53 of these muscles tends to e nhance the ability of the lar ynx to manipulate its internal structures. Regardless, the presence of lo wer motor neuron synaptic connections of both spinal accessory and vagus nerves in the jugular nucleus and the nucleus ambiguous indicate that the pure or singl e-nerve innervation theory of the intrinsic laryngeal muscles may need to be reviewed. Gross dissection re veals that all the intrinsic laryngeal muscles are innervated by the recurrent laryngeal branch, except the cricothyroid muscle which is innervated by the superior laryngeal nerve. Most of the laryngeal framework is in the form of cartilage which allows for mobility of the larynx (Titze, 1994). The larynx is comprised of the following cartilages: thyroid, cricoid, a pair of ar ytenoid cartilages, and the epiglottis (Fink & Demarest, 1978). One bone, the hyoid, is a horseshoe shaped structure that floa ts above the larynx and anchors a number of extr insic laryngeal and lingual muscles. The hyoid bone is attached to the thyroid cartila ge by the thyrohyoid membrane a nd to the epiglottis via the hyoepiglottic ligament (Titze, 1994). Laryngeal cartilages. The thyroid cartilage consists of two parts or lamina joined anteriorly at the midline, forming a 90 to 120 angle on the top (T itze, 1994). Adult males usually have a smaller angle which fo rms a laryngeal prominence often referred to as the Adams apple. The posterior third of this horizontal arch is open. There are four projections from the posterior borders; one on each end projecting down (inferior cornus) and one on each end projecting up (superior cornus). Each infe rior cornu articulates in a true joint with the lower side of the post erior portion of the cricoid cartilage and each superior cornu is connected to the poster ior extreme of the hyoid bone via the lateral
54 thyro-hyoid ligament. The thyroid cartilage forms a frontal shield for the airway and a site for attachment of intrinsic laryngeal muscles. The cricoid cartilage lies directly belo w the thyroid cartilage and forms the only complete laryngeal ring surrounding the top of the tracheal airway (Titze, 1994; Zemlin, 1999). This ring is wider a nd taller posteriorly, resembling a signet ring (Fink & Demarest, 1978). The inferior thyroid cornus at tach on either side of the cricoid signet to form the cricothyroid joint. Through action of the intrinsic laryngeal muscles, this joint allows the front of the cricoid to rota te upward toward the thyroid cartilage. The paired arytenoid cartilages are shaped like tetrahedrons (p yramids) with four triangular surfaces. One of the surfaces acts as the base and is attached to the top, posterior portion (the signet) of the cricoid cartilage form ing the cricoarytenoid joint. This joint is very flexible and allows for rotation and gliding of the arytenoid cartilages on top of the cricoid cartilage. The movement of the arytenoid cartilages is responsible for the adduction-abduction of the vocal fo lds (Sawashima & Hirose, 1983; Webster, 1999; Zemlin, 1999). The epiglottis is a cartilage that resemble s the tongue of a shoe. It is attached by connective tissue to the inner su rface of the thyroid cartilage just below the thyroid notch and forms the anterior wall of a chamber (Titze, 1994, p. 10), the laryngeal vestibule above the glottis. Superiorly, the epiglottis is attached to the hyoid bone that also serves as the base of the tongue. During swallowing, the epiglottis retrofle xes over the opening of the larynx to effectively cl ose off and protect the airway.
55 Laryngeal muscles. The muscles of the larynx ar e divided into two groups, extrinsic muscles and intrinsic muscles. The extrinsic laryngeal muscles connect the larynx to other structures in the head, neck, and chest (hyoid bone, sternum, and pharynx). Although there are five extrinsic laryngeal muscles, only the following three significantly contribute to phonation: t hyrohyoid muscle, sternot hyroid muscle, and sternohyoid muscle. These muscles are res ponsible for the basic laryngeal positioning including: (1) suspension and stabilization of the thyroid cartilage within the neck, (2) vertical movement as in elevation or de pression of the larynx as a whole, and (3) imposition of laryngeal stress /tension/rigidity through cervi cal and pectoral tension (Jrgens, 2002; Webster, 1999; Zemlin, 1999). During singing, the external laryngeal muscles proportionally increase their activity as the fundamental frequency increases or decreases from the value of a normal, rela xed speaking voice (Roubeau, Chevri-Muller, & Saint Guily, 1997). The intrinsic laryngeal muscles consist of a pair of thyroarytenoid muscles, a pair of cricothyroid muscles, a pair of lateral cricoarytenoid muscles, a pair of posterior cricoarytenoid muscles and a single interary tenoid muscle. Action of these muscles on the cartilages and joints result in (1) abduction-adduction of the vocal folds, (2) constriction of the supraglottic laryngeal stru ctures, (3) changes in the length and tension of the vocal folds, and (4) vertical moveme nts of the larynx (Sawashima & Hirose, 1983; Webster, 1999; Zemlin, 1999). The intrinsic muscles connect the ca rtilages within the larynx and are primarily responsible for sustai ning or changing pitch by controlling the movements of the cricothyroid and cricoaryteno id joints (Sawashima & Hirose, 1983).
56 The bulk of the vocal fold is made up of the thyroarytenoid muscle. Each muscle extends from the posterior surface of the ante rior arch of the thyr oid cartilage, below the thyroid notch, and inserts into the anterior/vocal angle of the arytenoid cartilage (Sundberg, 1987; Zemlin, 1999). Isotonic contr action of the thyroarytenoid muscles glides the arytenoid cartilages forward, th ereby shortening, thicke ning and stiffening the vocal folds (Titze, 1994). The vocal fold may be described in three layers: (1) the mucosa, which acts as a cover and consists of the epithelium and the superficial layer of the lamina propria; (2) the ligament, which consis ts of the intermediate and deep layers of the lamina propria, and (3) the thyroarytenoid muscle, which is lateral to the ligament (Hirano & Sato, 1993). The vocal folds ar e approximately 3 mm long in the newborn infant and grow to about 9 to 13 mm in females and 15 to 20 mm in adult males (Sundberg, 1987). The longer the vocal fold, the lower the pitch range (Sawashima et al., 1983). The pair of cricothyr oid muscles is the primary muscle for pitch control and the only intrinsic laryngeal mu scle to be innervated by the s uperior laryngeal branch of the vagus nerve. This unique innerv ation pattern allows this one muscle pair to be activated while the others are deactivated. Titze rema rks that this special designation makes the function of the cricothyroid muscle (raising pitch by vocal fold elongation) very select (p. 19, 1994). The cricothyroid muscle consists of two part s that originate on the anterior arch of the cricoid. One part courses vertically and inserts into th e lower border of the thyroid cartilage lamina while the other section runs up and back to insert in the inferior cornu of
57 the thyroid cartilage (Titze, 1994). EMG (e lectromyography) measurements provide evidence that the contraction of the cric othyroid muscle elongates the vocal folds (Sundberg, 1987). When the cricothyroid muscle contracts, it pulls the cricoid arch upward, depresses the thyroid lamina which s hortens the cricothyroid space, and results in lengthening the thyroarytenoid muscle, el ongating the vocal fold. Deactivation of the cricothyroid muscle prevents the production of high pitc h tones (Jrgens, 2002). The actions of the thyroarytenoid and th e cricothyroid muscles oppose each other; that is, the thyroarytenoid muscles shor ten and thicken the vocal folds and the cricothyroid muscles lengthen and thin the vocal folds. Together, these two muscle pairs are responsible for most of the changes in vocal fold length and mass. Muscles that oppose each other comprise an agonist-antagonist pair. It is theorized that because these two muscle pairs are innervated by two separate nerve branches, the effective stiffness of the vocal fold has a wide range of variability. The lateral and posterior cricoarytenoid muscles form bilateral agonist-antagonist pairs. Each lateral cricoarytenoid muscle originates from the superior borders of the cricoid arch and courses upward and posteriorly to insert la terally into the corresponding arytenoid. They adduct (close) the vocal fo lds by rotating the arytenoids forward and medially toward midline on the cricoarytenoid joint. This action is opposed by the action of the posterior cricoarytenoid muscles which originate on the posterior surface of cricoid cartilage and course upward and laterally to insert into the arytenoid cartilages. Their function is to abduct (open) the vocal folds by rotating the vocal proc esses away from the midline.
58 The interarytenoid muscle is the only singl e muscle in the larynx and connects the two arytenoid cartilages posterior ly. This muscle serves as an adductor of the arytenoid cartilages and functions to tightly close the pos terior portion of the gl ottis (Titze, 1994). The interarytenoid cartilages reflexive closure of the vocal cords contributes significantly to airway pr otection. It is believed to be the only intrinsic laryngeal muscle to have muscle spindles. Although controvers ial, recent findings suggest that spindles are sparse or absent in th e thyroarytenoid, lateral cricoa rytenoid, cricothyroid and posterior cricoarytenoid muscles (Brandon et al., 2003; Ludlow, 2005). Physiology of Vocal Pitch Control Pitch is the perceptual correlate of freque ncy which refers to the sounds physical structure (Patel & Balaban, 2001). The pitch produced by a persons voice is measured as the fundamental frequency (F0). Pitch control is an essential feature of voice production. Titze reflects that pitch is one of those dimensions that if correctly adjusted draws little attention to itself, but if incorrect it can reduce the acceptability and intelligibility of the human voice (p. 214, 1994). For singers, accurate control of pitch and intonation patterns is of paramount importance. The biomechanical control of singing follo ws the principle of trade-off or motor equivalence between the activity of mu scles (Hurme, Laukkanen, & Sonninen, 1999, p. 333). A muscle produces moveme nt by contraction that exerts force on levers such as cartilages and bones, never by extension. Musc le contraction causes either the shortening of a muscle between its origin and insertion points (isotonic contraction), or an increase
59 in the inner tension of the mu scle without affecting the leng th (isometric contraction) (Hurme, et al., 1999). Pitch production is a combination of biomechanical activity (intrinsic and extrinsic la ryngeal muscles) and aerodynamics (Larson, 1998). Control of vocal pitch is accomplished th rough a balance of the mass and stiffness of the vocal folds and under the influence of th e subglottic air pressure (Hollien & Hicks, 1979; Hurme & Sonninen, 1998; Jafari, Wong, Behbehani, & Kondraske, 1989; Titze, 1994). Vocal fold stiffness is affected by change s in length and tension of the vocal fold. Mass depends on tension and is also influe nced by vocal fold length and subglottic pressure (Jafari et al., 1989). Mass is defi ned as, the amount of material that is effectively in vibrati on (Titze 1994, p. 193). When the stiffness of the vocal folds is increased, pitch also increases. The longer, thinner, and tenser the vocal fo lds are, the higher the phonation frequency becomes (Sundberg, 1987). An increase in sti ffness and tension of the vocal folds is achieved by contracting the cricothyroid muscle (agonist) while the po sterior and lateral cricoarytenoid muscles stabili ze the cricoarytenoid joints a nd adduct the vocal folds. This concerted effort stretches and tenses the th yroarytenoid (antagonist), and thus lengthens the vocal folds. The F0 is regulated by th e differential control of these muscles (Titze, 1994; Webster, 1999; Zemlin, 1999). In addition to increasing the length of the thyroarytenoid muscles by contracting the cricothyroid, F0 control is also dependent on the amplitude of vibration of the vocal folds. An increase in muscle tension and s ubglottic pressure will increase the amplitude of vibration. If the amplitude of vibration is large enough to involve the muscular layer of
60 the vocal fold, then isometric contraction of the thyroarytenoid occurs increasing the vibrating muscle mass causing an increase in F0 (Hurme et al., 1999; Titze, Luchei, & Hirano, 1989). However, in general, the effect of an increase in subgl ottic pressure is not to increase pitch, but rather to increase the intensity (p erceived loudness) in phonation (Sundberg, 1987). In order to maintain a cons tant F0, the activity of the cricothyroid muscle has an inverse relation with subglottic pr essure; that is, to maintain a constant F0, the activity of the cricothyroid is reduced as the subglottic pressure is increased. Conversely, if the subglottic pres sure is decreased, the activit y of the cricothyroid muscle must be increased to prevent the F0 from lowering (Hurme et al., 1999; Sundberg, 1987). If the intrinsic laryngeal muscle contractions remain constant, then subglottic pressure and F0 are naturally balanced. Vocal Pitch Control An Integration of Systems Physiologically, vocal pitch control requires the integr ation and coordination of numerous neurological system s including neuromuscular, sensorimotor, auditory, limbic, and executive systems. The ability to produ ce a precise pitch and rapid pitch changes for singing relies on the rigorous interaction and control of these motor and sensory systems; however, the manner by which these systems are controlled to execute precise pitch production is inadequately understood. In particular, researchers are exploring the interactions between the aud itory system and the laryngeal system to achieve instant assessments and adjustments for precise pitch production.
61 Phonatory monitoring systems. Based on an analogy to striated muscles in the extremities, Kirchner and Wyke (1965) proposed that the striated muscles in the larynx have a similar proprioceptive reflex system that signals any devi ation from the required condition for accurate pitch production, there by reducing any performance error to a minimum. Previous electromyographic (EMG ) studies documented activity in the cricothyroid and thyroarytenoid muscles prior to the production of sound, implicating voluntary pre-phonatory tuning (Buchthal, 1959 as reporte d in Kirchner & Wyke, 1965). Subsequent research lead to the propos al that singers proceed through a sequence of three precisely controlled neuromuscula r events: (1) pre-phona tory tuning, (2) intraphonatory reflex modulation, and (3) ac oustic monitoring (Wyke, 1967, 1974). Prephonatory tuning refers to the voluntary pos turing of the laryngeal structures for production of an intended pitch (Watts, Murphy, & Barnes-Burroughs, 2003; Wyke, 1967). It is suggested that pre-phonatory tuning of the larynx is the major voluntary contribution to the control of the larynx dur ing singing and speech. The tension pattern of the laryngeal musculature is based on previous experien ce; that is, neuromuscular memories. Because this event is voluntary, its precision may be enhanced with training. Once expiratory air is set into mo tion through the larynx, it is dependent on laryngeal reflex-generating systems to m onitor and modify me chanical action for accurate pitch production; that is, intra-phonatory reflex modulation (Wyke, 1974). These mechanoreceptors are located in the in trinsic laryngeal muscles, the subglottic muscles, and in the joints of the laryngeal ca rtilages (Adzaku & Wyke, 1979; Sundberg, 1987). Stretch receptors in the intrinsic musc les respond to changes in the length of the
62 thyroarytenoid and cricothyroid muscles and re gister relative moveme nts of the laryngeal cartilages against each other (Jrgens, 2002). Mucosal mechanoreceptors in the laryngeal mucosa react to direct puffs of air that elicit reflexive laryngeal adduction (Bhabu, Poletto, Mann, Bielamowicz, & Ludlow, 2003). Pr essure receptors respond to variations in subglottic air pressure; while joint receptors react to rotations or dislocations of joints that connect the laryngeal cartilages (Baken & Noback, 1971; Keene, 1961 as cited in Titze, 1994; Suziki & Sasaki, 1977). Since these reflexes are controlling striated muscular activity, vocal trai ning and practice similar to at hletic training may increase reflex efficiency (Wyke, 1974). However, as the larynx ages, there is an inevitable reduction in reflex efficiency that is refl ected in vocal instab ility even though the voluntary pre-phonatory tuning a nd voluntary respiratory mu scle control may not be impaired (Wyke, 1974). Once the vocalization is audible, Wyke (1974) proposed that acoustic monitoring may provide feedback for readjustments of the laryngeal musculature, but the significance of this ability is unclear. He argued that trained singers could sing accurately even when their own voices are masked and that the onset of hearing impairment in trained singers does not lead to immediate deteri oration of voice. Although the exact relationship between acoustic monitoring and the laryngeal musculature was unable to be determined, Wyke (1974) concluded, there is no doubt that acoustic automonitoring does permit a musi cally talented and well-trained singer to impose on the processes of neuro-muscular control of his laryngeal (and respiratory)
63 muscles a further degree of refinement that cannot be exercised by the untrained subject (p. 303). Auditory monitoring system. Evolving from the work of Kirchner and Wyke (1965), researchers argue that the receptor system in the larynx is continuously influenced by the auditory system and that auditory feedback is crucial for controlling vocal F0 (Amir, Amir, & Kishon-Rabin, 2003; Leydon, Bauer, & Larson, 2003; Perry et al., 1999; Titze, 1994). Previous studies in the literature su pport a closed-loop auditorygovernance system; that is, an auditory perceptual monitori ng system (Davidson, 1959; Elman, 1981; Lane & Tranel, 1971; Perry et al., 1999). A clos ed-loop system is sensitive to errors and uses feedback to make adjust ments whenever error-performance signals are detected. Closed-loop systems, also known as interactive systems, depend on positive or negative feedback to reach or maintain a targeted goal (Hood, 1998) In contrast, an open-loop system completes a task with no influence by external events (Hood, 1998). In 1959, Davidson demonstrated that when auditory feedback is artificially delayed, speakers automatically decrease thei r speaking rate. Lane and Tranel (1971) confirmed that when individuals are subj ected to background noise, vocal intensity immediately increases. This occurrence is know n as the Lombard effect. Investigations on the effects of frequency-sh ifted auditory feedback on the production of vocal pitch have found that the fundamental frequenc y changes when the auditory feedback frequency is altered (Elman, 1981; Larson, 1998). Thus, a concurrence of evidence exists to support a closed-loop system, in which auditory monitoring provides moment-tomoment feedback for the regulation and control of vocal production.
64 Hearing-impaired individual s have a disruption of the auditory feedback loop and consequently, many have difficulty monito ring their vocal produc tion. Inappropriate vocal register, intermittent and unpredictable pitch breaks, and pitch monotony have been reported (Martony, 1968). Jones and Munha ll (2000) report that deterioration of suprasegmental features commonly occurs s oon after the onset of hearing loss; however, accuracy of vowel and consona nt production is maintained much longer. This implies that control of suprasegmental features of voice; that is, inte nsity, pitch, intonation patterns, stress, and rate of speech, may be more sensitive to auditory feedback than control of phoneme production. Jones and Mu nhall (2000) observed that the relationship between auditory perception and vocal production is non-linear. Auditory monitoring may be artificially re stored with a cochlear implant. The cochlear implant is a neuroprosthetic devi ce that converts sound en ergy to electrical energy which then stimulates the auditory nerv e with electrical impulses (Campisi et al., 2005). It is used to provide auditory sens ation to individuals w ith severe to profound deafness. The device does not restore norma l hearing; however, it provides auditory feedback cues in timing, intensity, and fre quency (Campisi et al., 2005). There are numerous reports in the litera ture on vocal changes after cochlear implantation; however, many lack standard methodology and statistical data. Participant se lection includes both pre-lingually and post-lingually de af individuals and spans a wi de age range. Moreover, these individuals have large inter-individual differences in fundamental frequency preand post-cochlear implantation (Langereis Bosman, van Olphen, & Smoorenburg, 1998). Consequently, results are inconsistent a nd conflicting (Campisi et al., 2005; Higgins,
65 McCleary, Carney, & Schulte, 2003; Higgins McCleary, Ide-Helvie, & Carney, 2005; Perrin, Berger-Vachon, Topouzkhanian, Truy, & Morgon, 1999; Schenk, Baumgartner, & Hamzavi, 2003; Seifert et al., 2002). In general, the results (a) support a non-linear relationship between auditory perception and vocal produc tion (Campisi et al., 2005; Higgins, McCleary, Ide-Helvie, & Carney, 2005; Schenk, Baumgartner, & Hamzavi, 2003), and (b) indicate that development of aco ustic speech parameters is better when auditory monitoring is artificia lly restored in pre-lingually de af children before the age of 4 (Higgins et al., 2003; Seifert et al., 2002). Leydon, Bauer, and Larson (2003) and others investigated the degree to which singers rely on auditory input to regulate ch aracteristics of vibrato (Dejonckere, 1995; Jones & Munhall, 2000). Vocal vi brato is characterized by sma ll periodic fluctuations in fundamental frequency and intensity in the si nging voice. These pulse s typically occur at a rate of 47 Hz with a fundamental freque ncy fluctuation of 1 semitone (Shipp & Izdebski, 1982). A semitone is the smallest musical interv al, but relatively a gross pitch difference (Bentley, 1966, p. 36). The semitone corresponds to a frequency difference of approximately 6% in Western music. In ot her cultures, such as Arabic, Indian, and Chinese, the tuning interval is smaller (T ervaniemi & Brattico, 2004). A performer must be able to distinguish between much smaller pitch differences than a semitone to achieve unison, good intonation, and artistry (Bentley, 1966, p. 36). Thus, the semitone is further divided into cents; one semitone equa ls 100 cents. Vocal vibrato is a desirable characteristic for singers as it lends ric hness to a tone (Seasho re, 1939) and helps to distinguish a singers voice from the orchestra (Sundberg, 1987).
66 Leydon and colleagues (2003) proposed that a closed loop negative feedback reflex within the auditory system contribu tes to sustaining funda mental frequency and intensity modulations in singers during vocal vibrato. They refer to this component of the auditory system as the pitch-shift reflex ( PSR). Like a stretch re flex, the PSR is a bidirectional closed loop negative feedback re flex that is triggered in response to discrepancies between the intended and percei ved pitch with a latency of approximately 100 ms. Compensatory reflexive responses l ead to oscillations in pitch approximately every 200 ms resulting in ~ 5 Hz modulation of the fundamental frequency; thus, the PSR contributes to the production of vocal vi brato (Leydon et al., 2003). PSRs were experimentally elicited from nonsingers by in troducing sinusoidal pitch-modulations in auditory feedback at discrete integer fr equencies from 1 to 10 Hz with 25 cents amplitude modulation, resulting in a peak-to-peak pitch modulation of 50 cents (0.5 semitones). Modulated audito ry feedback induced pitch fl uctuations consistent with vocal vibrato with peak energy gains between 47 Hz with an average of 5 Hz in the F0 of all subjects, demonstrating the existence of a pitch-shift reflex within the auditoryvocal system. Kinesthetic feedback system. It is well-known that those who have lost their hearing after speech and language acquisi tion rely on proprioceptive memory and knowledge of previous experience to compute the motor-sequence for the desired vocal production in the absence of auditory feedb ack (Amir et al., 2003, Waldstein, 1990). This implies not only an auditory loop for vo cal production, but a complex and dynamic internal proprioceptive monitoring system.
67 For professional vocal musicians, audito ry feedback alone cannot explain their ability to accurately control pitch when they cannot hear their own voices, such as in choir singing or with orchestral accompanim ent (DiCarlo, 1994; Mrbe, Pabst, Hofman, & Sundberg, 2002). This occurrence provides su pport for an additional feedback system such as an internal kinesthetic model (D iCarlo, 1994; Jones & Munhall, 2000; Mrbe, Pabst, Hofman, & Sundberg, 2004; Mrbe et al., 2002; Murry, 1990; Ward & Burns, 1978). For speech production, auditory feedback ma y be the primary control for variation in pitch that constitutes intonation patterns. For singers, auditory feedback has the same function, but by its very nature real-time auditory feedback does not provide a reliable means for controlling the prec ision of vocal characteristics for singing. Pitch extraction is fundamental in the percepti on of speech intonation, but it is crucial to the processing of music. Ward and Burns (1978) demonstrated that trained singers rely more on kinesthetic feedback for vocal pitch contro l, while untrained singers rely more on auditory feedback. Trained and untrained singers sang rising and falling scales with and without a masking noise. There was no significant difference between the singers and nonsingers control of pitch as long as the subjects could hear the sound of their ow n voices. When the auditory feedback was masked, all subjects sang out of tune; however, the nonsingers were significantly less accurate that the traine d singers. This implies that vocal training may enhance proprioceptive memory and/or shar pen proprioceptive refl exes of laryngeal joints and muscles for su perior pitch control.
68 Murry (1990) compared laryngeal accur acy of nonsingers and singers and the effect of repeated trials wh en matching pitch to a pre-set tone. Results revealed the trained singers to be more accurate in both of the pitch matching tasks. The performance of the nonsingers was highly variable and less accurate. Murry concluded that singers are able to adjust the physiological parameters of F0 control (vocal fold mass, stiffness, tension and subglottic pressure) more ra pidly and precisely than the nonsingers. DiCarlo (1994) describes internal vo ice sensitivities as a proprioceptive feedback system used by profe ssional singers. She writes th at internal voice sensitivities rely on proprioceptive feedback resulting from the transmission of laryngeal vibrations to the skeletal framework of the thorax and cran iofacial structure by me ans of the extrinsic laryngeal muscles. This de scription agrees with Wyke (1974) who proposed that the kinesthetic feedback of the intrinsic larynge al muscle system may be supplemented by discharges from peripheral mechanoreceptors in the thorax, the abdominal wall, and the vocal tract. DiCarlo descri bes a reflex conditioning tech nique which allows students over a learning period between two and si x years to develop the kinesthetic proprioceptive memory required for the contro l of voice. She states, The teacher must guide the students solely by ear, teaching them to associate an auditory image with an internal sensation (p. 84). For the professional singer, intern al voice sensitivities may be a means of control more reliable than auditory feedback alone. Similar to DiCarlos description of inte rnal sensitivities, Jones and Munhall (2000) explored the use of acoustic feedback in calibrating an intern al feedback system for the control of speaking pitch. Jones a nd Munhall state that although internal models
69 might reduce the need for closed loop control, auditory feedback is necessary for the acquisition and maintenan ce of internal models. The inve stigators tested the extent to which an individuals habitual speaking pitch is controlled by an internal F0 target. Subjects produced the vowel /a/ under a cont rol condition (normal F0 feedback) and two experimental feedback conditions: (1) F0 sh ifted up and (2) F0 sh ifted down. Results indicated two related effects of altered feedback on F0. Duri ng the trials of shifted pitch (either up or down), subjects compensated for th e pitch shifts in an apparent attempt to maintain habitual pitch; that is, if the p itch was shifted up, subject s lowered their pitch relative to a control conditi on and conversely, if the pitc h was shifted down, subjects raised their pitch relative to a control condition. The subjec ts also showed evidence of sensorimotor adaptation; that is, after normal feedback was returned for the last 10 test trials, the mean pitch of subjects for the shift-up condition increas ed significantly to maintain the test stimuli frequency; c onversely, the shift-down condition showed a significant decrease when feedback was retu rned. The results suggest that F0 may be controlled using auditory feedback and an internal pitch representation. In similar studies of pitch-shifted aud itory feedback, Hain and colleagues (2000) noted two responses of the lar yngeal mechanism to altered pitc h feedback. The first is an automatic response that corrects for sma ll unplanned changes in pitch. The second reaction is a slower and voluntary response th at modifies pitch to meet a target or reference pitch, such as in speaking or si nging. The latency of the first response is approximately 100-150 ms, while the second re sponse varies between 250-600 ms (Hain et al., 2000; Larson, 1998). It has been suggested that the initial automati c reaction may
70 be part of a laryngeal-specific reflex pa th (Sapir, McClean, & Larson, 1983) or an automatic function of the audio-vocal reflex system (Hain et al., 2000) or part of a dual auditory-mediated feedback system that wo rks in parallel and is controlled by the cerebellum (automatic reaction) and the cortex (voluntary response) (Kawahara & Williams, 1996). The effect of music traini ng on either of these responses is unknown. Mrbe et al. (2002) conducted a three-y ear longitudinal study investigating the significance of auditory and ki nesthetic feedback for pitch control in vocal students and the effects of training on pitch control. At the beginning of their professional singing education, the students sang an ascending and descending triad pattern with and without masking noise. The effect on pitch control wa s investigated in four tasks: (1) legato slow, (2) staccato slow, (3) legato fast, and (4 ) staccato fast. Mrbe et al. (2002) noted a significant difference between unmasked and masked regardless of the technique and tempo. The singers intonation a ccuracy was reduced in the ab sence of auditory feedback and the singers relied on ki nesthetic feedback to perfor m the tasks with masking. The same students were re-examined three years later (Mrbe et al., 2004). The same measurements, procedures, and equipm ent were used. The contribution of the auditory feedback to pitch control was not sign ificantly different after education; that is, intonation accuracy did not signi ficantly improve after three years of vocal training. However, a significant improvement of pitc h accuracy was found for all of the slow singing conditions. Mrbe and colleagues c onclude that although education did not improve auditory feedback skills, kinesthe tic feedback improved for the slow singing tasks. They report that this improvement indicates that the accura cy of the absolute
71 neuromuscular memory of pitch increased after education (p. 241). The authors speculate that three years of professional training may not be enough time to improve kinesthetic pitch control in demanding vocal task s such as fast legato or fast staccato. As DiCarlo (1994) reported, it may take 2 to 6 y ears for kinesthetic feedback to develop for accurate pitch control. Summary. Control of pitch is a complex biomechanical and aerodynamic system. It appears that researchers ag ree that the ability to rapidl y produce a precise pitch is essential for the professional vocal musicia n. Evidence indicates that accurate pitch control depends on auditory perceptual m onitoring, proprioceptive feedback of the laryngeal system and phonatory reflex sy stems (Amir et al., 2003; Jones & Munhall, 2000; Kirchner & Wyke, 1965; Mrbe et al., 2004; Murry, 1990; T itze, 1994; Ward & Burns, 1978; Wyke, 1974). Trai ned singers consistently co ntrol fundamental frequency and maintain targeted pitch better than unt rained singers. The evidence implies that professional vocal training enhances kinest hetic feedback of the laryngeal system, improves neuromuscular memory for pitch co ntrol and increases reflex efficiency. Research suggests that neuroanatomical conne ctions may exist between the auditory and vocal systems to regulate vocal pitch (Hain et al., 2000); however, the exact relationship between the laryngeal and auditory systems is unknown. Development of Pitch Control Sequence of development. Welch (1994) described the activity of singing as a complex web of interacting factors embracing perception, cognition, physical
72 development, maturation, society, culture, hi story, and intentionality (p. 3). In the literature of music research, investigators have suggested a sequence for the development of singing accuracy in children (Bentle y, 1966; Boardman, 1964; Davidson, McKernon, & Gardner, 1981; Goetze, Cooper, & Brown, 1990). There is general agreement that infant s pass through a babbli ng song phase after playing with groups of musical pitches and phrases in a repe titive fashion (Davidson et al., 1981; Ries, 1987; Welch, 1994). Under the age of 12 months, Ries (1987) found that infants singing is characteri zed by a restricted pitch range focused around a central tone, but with no evidence of rhythm As children mature from toddlers to preschoolers, a sense of rhythm develops first. Children ar e observed to move their bodies with the pulse of the music. Rhythmic memory is establishe d first and is more hi ghly developed at all ages of childhood than pitch memory (Bentley, 1966). Follo wing the establishment of rhythm, is the addition of melody which is defined by Bentley as a succession of pitch sounds within a rhythmic framework (p. 26, 1966). When children recognize a previously heard tune, melodic memory is es tablished. By age 5, children normally gain mastery of words and rhythm of the song before the pitch components (Welch, 1994). Pitch discrimination for the specific per ception of melody, followe d by the ability to analyze harmony (chords), are th e last skills to develop. In 1966, Bentley published Musical Abilities in Children based on his study of 2000 school children between the ages of 7 and 14 from state primary and secondary schools. His conclusions are still quoted in contemporary literature (Davidson et al., 1981; Goetze et al., 1990; Phillips & Aitc hison, 1997; Yarbrough et al., 1991). Bentley
73 found no significant gender differences in mu sical ability, nor did he find a strong correlation between intelligence and musical abilities. Auditory pitch discrimination appears to be more accurate on sounds near the middle of the vocal range than on sounds at the extremes of the vocal range. By age 7, most children can accurately discriminate pitch differences of a quarter tone (~ 12 Hz difference). By age 12, most can discriminate between one-eighth tones (~ 6 Hz). Although yearly increases are small, pitch discrimination and singing accuracy im prove with chronological age throughout childhood. Socio-cultural influence. Singing is an ancient and complex form of human behavior richly influenced by ones cultural familiarity, aesthetic preferences, and artistic intent. Deutsch (1992) reports that musical perception is linked to the patterning of speech and both are developmental in nature and related to membership in particular socio-cultural groups. Observations of child ren singing provide evidence of patterning, repetition, and transformation, implying a sens e of organization that is reflective of familiarity, and an unconscious knowledge of rules significant to the childrens particular school and home environments (Welch, 1994). Thus, singing not only follows a developmental pattern, but is strongly in fluenced by social norms and culture. Potential Factors Related to Vocal Pitch Control The ability to sing with accu rate pitch control is the mo st basic skill of singing and is considered the featur e that distinguishes singers from nonsingers (Murry, 1990; Titze, 1994; Watts, Barnes-B urroughs, Adrianopoulos, & Carr, 2003). Much of the
74 existing research has focused on children who cannot match pitc h (Apfelstadt, 1984; Geringer, 1983; Goetze et al., 1990; Gree n, 1990; Howle, 1992; Joyner 1969; Moore, 1994; Pedersen & Pedersen, 1970; Porter, 1977; Yarbrough et al., 1991). The ability to vocally match a pitch or sing a melody varies among individuals and the prospect that some children may not learn to sing accurate ly is a challenge for music educators (Yarbrough et al., 1991). There are those within our society who have exceptional abilities to sing with breathtaking resonance and vocal precision, while other individuals are unable to discriminate between musical tones or to voc ally match a pitch. A complete explanation and clarification of such inter-ind ividual variability is lacking. Potential functional variables. Research on childrens vocal pitch matching abilities has focused on training models, char acteristics of the singing task, singing with text, singing environment (group versus indi vidual), accompaniment, age, gender, and other contributing factors (Moore, 1994). After reviewing 25 years of research literature on the singing abilities of elementary children, Goetze, Cooper, and Brown (1990) summarized the factors affecting childrens success as follows: 1. As children grow older, singing accuracy improves. 2. The relationship between pitch discri mination ability a nd vocal production accuracy remains unclear. 3. Perhaps due to a childs natural tend ency to imitate, the presence and quality of the model pitch can inhibit or improve a childs vocal accuracy and vocal range.
75 4. Simple melodic material and descendi ng as opposed to ascending intervals appear to be most conducive to vocal accuracy. 5. Evidence was inconclusive to indi cate whether children sing more accurately with or without text. 6. Some children may sing more accurately alone than in a chorus. It is speculated that children may not hear themselves in a group; therefore, auditory feedback is not sufficient for accurate singing. 7. Children who receive feedback and reinforcement following their singing performance improve significantly. In general, positive personal characteristics such as motivation, concentration, and musical interest influence singing abil ities. In addition, en vironmental factors including exposure to music within the home environment, education, and training contribute to singing ac curacy (Howle, 1992). Potential physical variables. The physical act of singi ng requires muscular action and fine sensorimotor control and coordination. For the perfor mer, it is an athletic event; an aerobic exercise focused on the respiratory system and the larynx in particular. As in any athletic event, there are those who are more skilled or talented than others. Joyner (1969) examined children who were characterized as monotone to gain insight into the problem and to provid e suggestions for re medial training. A true monotone is described by Joyner as one who c onsistently fails to reproduce the tonal configuration of a melody in a recognizable manner (p. 115). Joyner proposed that those characterized as monotones are tone-dumb rather than tone-d eaf. He suggested
76 that the laryngeal mechanism lacks flexib ility in those who are monotone and is physically unable to adjust the length and tension for accurate pitch production. Consequently, kinesthetic feedback is dimi nished and accurate tonal memories do not develop leading to deficiencies in pitch discrimination. Joyn er (1969) concluded that in order to produce accurate pitch, a person must be able to do three things: (a) tell one pitch from another; that is, discriminate pitch, (b) recall successions of pitches organized into melodic patterns; in other words, remember pitch, and (c) have a vocal instrument physically capable of reproducing or matching an immediate pitch. After an extensive review of music l iterature, Goetz and colleagues (1990) came to very similar conclusions. They found th e following skills necessary for singing talent: (a) the ability to discriminate between pitches, (b) the ability to vocalize over a wide range of pitches, (c) the ability to monito r vocal pitch, and (d) the desire to sing. In a more recent quest to identify factors associated with natural singing talent, Watts, Barnes-Burroughs, Adrianopoulos, a nd Carr (2003) conducted a national survey among a homogenous group of professional sing ing pedagogues. Questions addressed three areas: (a) perception of singing talent, (b) physiological variables that distinguish between individuals with singing talent a nd those without, and (c) factors affecting untrained or natural sing ing talent. Pitch intonation or s inging in tune was identified as the most relevant factor for the perception of singing talent followed by timbre or richness of the tone and stylistic appropriateness. Moreover, pitch-matching ability and pitch discrimination were ranked as the t op two physiological factors having the most
77 influence on natural talent. Su rvey data predicted that natu ral singing talent is a product of genetics and environment rather than practice. Summary. In the development of pitch contro l, researchers agree that potential factors for singing accuracy are a combination of innate characteristics and a stimulating environment. Vocal pitch control and audito ry pitch discrimination are consistently identified as factors for singing accurac y. Although many researchers acknowledge a relationship between these two skills, the st rength and nature of this relationship is unclear. Relationship between Pitch Di scrimination and Pitch Control The role of auditory governance on vo cal production is evident from early infancy. By the end of the first year of lif e, infants with normal hearing produce complex vocal patterns that match the vocal language patterns in their environment (Boone, 1996). For those children who have a severe hearing impairment, voice onset time and control of fundamental frequency appear to be the mo st sensitive to the effects of diminished auditory feedback (Hig gins et al., 2005). Intuitively, it seems that vocal production and auditory input are directly related. Vocal pitch production is unequivocally a ffected by acoustic cues and may be manipulated by alterations in auditory F0 feedback (Campisi et al., 2005; Elman, 1981; Haines et al., 2000; Higgins et al., 2003; Higgins et al., 2005; Jones & Munhall, 2000; Larson, 1998; Leydon, et al., 2003; Schenk, Baum gartner, & Hamzavi, 2003; Seifert et al., 2002; Ward & Burns, 1978). The results of these studies imply that the auditory
78 system plays a substantial role in vocal produ ction; however, in these investigations, the perceptual dimensions were manipulated. Thus, the implications are not directly applicable to the role of auditory perception in no rmal vocal pitch production. Specifically, auditory frequency discrimi nation and vocal pitch production reflect abilities to accurately integrate sensory pe rception with motor pl anning and execution. Although a relationship is acknowledged between vocal pitch accuracy and auditory pitch discrimination, research is sparse and the natu re of this relationship remains uncertain. Auditory Discrimination and Voca l Pitch Control in Children In the literature of music research, prev ious research has focused predominantly on children. Several studies have found signifi cant correlations between auditory pitch discrimination and singing accuracy among school -aged children, while others have not. The earliest investigation of the relations hip between auditory abilities and vocal production was conducted by Seashore in 1919 (as cited in Amir et al., 2003). He asked a group of singing teachers to evaluate thei r students singing accuracy. This was done subjectively with no reported reliability. S eashore then tested the students auditory discrimination using a series of tuning forks. He concluded that there is a slight tendency toward relationship (1919, p. 58). In an early descriptive study by Gould (1969), elementary school teachers of music we re surveyed and asked to list the reasons for inaccurate singing. Inattention to pitch, inabil ity to hear pitch changes, and inability to coordinate the vocal mechanism were among the reasons given.
79 Porter (1977) investigated th e training effects on audito ry pitch discrimination and vocal pitch matching in randomized groups of elementary children. He questioned whether inaccurate singing results from inadequate vocal control or from an inability to discriminate stimuli accura tely (p. 68). Porter found no evidence demonstrating that inaccurate singing results from inaccur ate auditory pitch discrimination. Geringer (1983) examined the relationship between auditory pitch discrimination and vocal pitch-matching abilities among ra ndomly selected preschoolers and fourthgrade children. The older children performe d significantly better than the preschoolers on the vocal pitch matching tasks. No co rrelation was found betw een auditory pitch discrimination and vocal pitch matching abilit ies. Geringer proposed that a degree of auditory pitch discrimination may be a pre-re quisite skill to vocal pitch matching. He concluded, It is possible that pitch discri mination and pitch matching are simply two independent abilities, or that maturation and training are necessary to develop an interrelationship (p. 98). More recently, Philips and Aitchison ( 1997) investigated th e relationship of singing accuracy to auditory pitch discrimi nation and tonal aptitude among third-grade students. Vocal pitch-matching skills were evaluated by the investig ators and each child was labeled as accurate or i naccurate. Auditory pitch di scrimination was tested as the ability to judge the difference between higher and lower tones. Tonal aptitude was the ability to judge two musical patte rns as same or different. Responses on the auditory pitch discrimination task did not differ signi ficantly between the accurate and inaccurate singers; however, on the tonal ap titude task, the accurate si ngers performed significantly
80 better. Philips and Aitchison suggested that t onal aptitude may be more related to singing accuracy than auditory pitch discrimination a nd concluded that the relationship between auditory and vocal skills is uncertain and may be the case of a lagging development in both aural and vocal skills (p. 19). In contrast to the previous studies, i nvestigations by Pedersen and Pedersen (1970), Zwissler (1972), Goetze et al. (1990), and Yarbrough et al. (1991) have found a positive relationship between vocal pitchmatching abilities and auditory pitch discrimination. As in the studies reviewed above, the subjects were elementary children and the procedures and test measures varied. Pedersen and Pedersen (1970) studied th e relationship between auditory pitch discrimination and vocal pitch production of 6th graders using a rating system for pitch accuracy. They agree with Joyner (1969) that those who cannot vocally match pitch are deficient in auditory pitch discriminati on. They found a fairly strong relationship between pitch discrimination and vocal production (p. 271). Zwissler (1972, as cited in Phillips & Aitchison, 1997) examined the difference in the auditory pitch discrimination skills of 100 first graders. The childrens vocal pitch accuracy for singing was evaluated and the children were divided into two groups, accurate singers (n = 50) and inaccurate singe rs (n = 50). A test of auditory pitch discrimination was designed that asked the children to choose if the second tone of a pair was higher or lower than the first. The pitche s were presented in the childs vocal range, one octave higher, and one octave lower. Results indicated that the first graders who were judged to be accurate singers performe d significantly better on the auditory pitch
81 discrimination task than the inaccurate singers. Thus, these data support a positive relationship between auditory pitch discrimi nation abilities and accurate singing. Zwissler also noted that subjects identified pitch di fferences more accurately when the stimuli were within their singing range rather than an octave above or below. Yarbrough et al. (1991) examined the eff ects of different vocal models on vocal pitch-matching accuracy and the effects of different response modes including hand signals and pitch syllables. Children in grades K-3 and 7-8 were sele cted on the basis of their failure to accurately ma tch pitch (n = 163). They were randomly assigned to one of three different response modes. In general, results demonstrated the following: (1) no significant differences among correct responses due to response mode (e.g. hand signals, solfege syllables [do, re, mi], or la-la-la syll able), (2) a significant difference in response to female versus male vocal model suggesting an effect of timbre, and (3) significantly better vocal pitch-matching accuracy for eight h grade students versus kindergarteners, suggesting a maturational effect. Summary. The relationship between auditory pi tch discrimination and vocal pitch control as factors of accurate singing in childr en remains controversial. Researchers have proposed that vocal pitch control and audito ry pitch discrimination may be independent abilities between which a relationship deve lops with training or developmental maturation of the child (Geringer, 1983; Goetze et al., 1990; Yarbrough et al., 1991). Goetze et al. (1990) concluded that (a) children who sing accurately are likely to demonstrate accurate pitch discrimination, and (b) children who demonstrate inaccurate pitch discrimination are likely to sing inaccurately (p. 30). Goetze and colleagues
82 express concern over those chil dren who have accurate a uditory pitch discrimination skills, but do not control vocal pitch accurate ly for singing. This inconsistency implies that accurate singing may have a positive relationship with accurate auditory pitch discrimination skills; however, this relationshi p is not reciprocal. It is possible that children with accurate auditory discrimi nation but inaccurate vocal skills have undeveloped laryngeal control. In addition, there may be poor kinesthetic feedback from the larynx or delayed internal auditory monitoring. Auditory Discrimination and Vocal Pitch Control in Adults Although auditory pitch disc rimination (perception) and vocal pitch control (production) have been identified as relate d abilities and essen tial skills for vocal musicians, few studies have investigated and compared these variables in formally trained adult singers. For whatever reas ons, research has focused predominantly on children or formally trained instrumental mu sicians, not vocal musicians. Two earlier studies of professional adult si ngers were reviewed previous ly in this chapter (Murry, 1990; Ward & Burns, 1978). Ward and Burn s (1978) demonstrated that under masking conditions trained singers rely more on kinesthetic feedback ra ther than auditory input for vocal pitch control compared to untrained si ngers, implying that voc al training enhances proprioceptive memory. Murry (1990) sugge sted that singers adjust the physiological parameters for pitch production faster and more accurately than nonsingers. A more contemporary study examined th e relationship of auditory pitch discrimination and vocal pitch matching abilit ies in adult professional singers compared
83 to adult subjects who expr ess natural singing talent. Watts, Murphy, and BarnesBurroughs (2003) examined the vocal pitch-matc hing abilities of fifteen female subjects divided equally into three groups: trained si ngers, untrained subj ects with expressed singing talent, and untrained subjects with nontalented singing voices. A talented singing voice is defined as a special natural ability w here the sounds vary over a wide range of frequencies and are in tune with each other, or where such sounds are melodious (Watts et al., 2003, p. 185). The placement into a particular group was determined by professional voice teachers. The purpose of this study was to assess the abilities of th ese three groups to control fundamental frequency (F0) during a pitch-matching task using targeted pure tones, and to investigate whethe r these abilities were affected differentially when internal auditory feedback was and was not available. It was also questioned whether trained singers are able to pre-tune (i.e., pre-phonatory set) their vocal mechanism to more accurately match pitch when compared to those without training. As expected, the untrained singers with singing talent and the trained singers demonstrated significantly greater pitch-matching accuracy on all measur ed conditions compared to the nontalented singers. Watts and colleagues (2003) conclude that the ability to accurately match and produce vocal pitch is a prerequi site for singing talent and re quires accurate perception of pitch and coordination of that perception with moto r planning, programming, and execution. Moreover, the ability to accurate ly position the laryngeal structures for production of an intended frequency may be a nother variable relate d to singing talent.
84 While the separation of untrain ed talented singers from th e control group is unique to address issues of natural ta lent, the authors caution that results should not be overgeneralized since sample size was small a nd method of voice training was unknown. In addition, placement into a particular gr oup was based on subjective judgment. Amir, Amir, and Kishon-Rabin (2003) studi ed the relationship between auditory perception and vocal producti on between professional musicians and nonmusicians. The musicians played musical instruments (an aver age of 13 years) and none of the subjects had previous vocal or singing training. In a previous study (Kishon-Rabin, Amir, Vexler, & Zaltz, 2001), these authors compared the frequency discrimination thresholds, also referred to as the difference limen for freque ncy (DLF), between the same professional musicians and nonmusicians. Difference lime n for frequency (DLF) is the smallest detectable frequency difference. They conc luded that the professional musicians had superior auditory skills. Am ir et al. (2003) questioned whet her individuals with superior auditory abilities would also demonstrate better-than-normal performance on vocal production accuracy. Results indicated that the musician group produced the tones approximately three times more accurately than the nonmusician group. A significant correlation (r = 0.67, p < 0.001) was found between auditory discrimina tion and vocal production. The analysis of data suggested that 43% of the varian ce of the production data can be explained by auditory perception. When the data was c onverted to semitones, the musicians had average production errors no more than of a semitone for each frequency. In contrast, the nonmusicians had mean errors of approxima tely 1.3 semitones. On a musical scale,
85 inaccuracies greater than one semitone are perceived as a melody change. Thus, 0.5 semitone may be viewed by a musician as crossing a categorical boundary and creating an error of music syntax. The authors conclu ded that individuals w ith superior frequency discrimination abilities were able to vocally imitate pure tones with greater accuracy. However, frequency discrimination thresh olds could not be predicted from vocal production accuracy. The author s speculate that music trai ning enhances a musicians auditory perception of acoustic parameters in vocal productions that are otherwise ignored by nonmusicians. Summary Despite variable strategies and limitati ons of past researc h, investigators and educators consistently identify auditory pitc h discrimination (percep tion) and vocal pitch control (production) as related abilities and fundamental skills for vocal musicians. It is essential that singers accurately integrate se nsory perception with ne uromotor planning to precisely execute vocal producti on. Intuitively, it seems thes e two abilities are directly related; however, there is a consensus in the literature suggesting the relationship between auditory perception and vocal producti on may be indirect and complex. Reliable evidence supports the ex istence of a kinesthetic feedback loop between the auditory and laryngeal systems for accurate voice producti on. Longitudinal data indicate that professional vocal training enhances this pr oprioceptive reflex sy stem and alludes to neuromuscular pitch memory. The specific in teractions and relationships among these neurophysiological processes have yet to be defined.
86 Chapter Three Methods and Procedures Introduction Neuroanatomical differences between mu sicians and nonmusicians support the premise that intense music training and skill acquisition effect functional and structural change in the auditory system (Gaser & Schlaug, 2003; Pantev, Engelien, Candia, & Elbert, 2001; PascualLeone, 2001; Sc hlaug, 2001; Schn, Magne, & Besson, 2004; Zatorre, 2003). Current research data from electroencephalography (EEG) and magnetoencephalography (MEG) suggest that mu sic training influences pitch processing by refining the auditory neural frequencyprocessing network (K oelsch, Schrger, & Tervaniemi, 1999; Pantev et al., 2001; Schn et al., 2004; Shahin, Bosnyak, Trainor, & Roberts, 2003; Tervaniemi, 1993; Trainor, Sh ahin, & Roberts, 2003) Following intense music training, auditory neural responses of instrumental musi cians have shorter latencies (faster responses) and larger amplitudes (stronger responses) for pitch changes than nonmusicians (Koelsch, Schmidt, & Kansok, 2002; Shahin et al., 2003). Psychoacoustic studies of auditory frequenc y discrimination indicate that formally trained instrumental musician s have superior auditory p itch discrimination skills (KishonRabin, Amir, Vexler, & Zaltz, 2001; Spei gel & Watson, 1984). Auditory pitch discrimination and the ability to sing with accurate pitch control are regarded by music
87 educators as fundamental abilities for musical talent and essential sk ills of a successful singer (Bentley, 1966; Geringer, 1983; Murr y, 1990; Seashore, 1919; Titze, 1994; Watts, Barnes-Burroughs, Adrianopoulos, & Carr, 2003). Accurate vocal production of targeted pitch is a complex biomechanical system. This skill depends on precise neuromuscular control of the larynx, accur ate auditory pitch discrimination, and continuous proprioceptive feedback (Kirchner & Wyke, 1965; Wyke, 1974). L ongitudinal research of vocalists in training indicates that accuracy for the absolute neuromuscular memory of pitch increases with music education (M rbe, Pabst, Hofman, & Sundberg, 2003, 2004). Auditory pitch discrimination and vocal pitch accuracy are two identified processes that demonstrate the integration of sensory perception with motor planning for the execution of vocal production. An accumulation of evidence acknowledges a relationship between these two skills; however, there is an inadequa te understanding of the nature of this relationship particularly in formally trained vocal musicians (Amir, Amir, & Kishon-Rabin, 2003; Geringer, 1983; Goetze, Cooper, & Brown, 1990; Watts, Murphy, & Barnes-Burroughs, 2003; Yarb rough, Green, Benson, & Bowers, 1991). Previous psychoacoustic and neurophysiologica l research has focused on formally trained instrumental musicians rather than formally trained vocal musicians (Mnte et al., 2003; Schlaug, 2001; Zatorre, 2003). Moreover, ther e is a prevalence of evidence to support training-induced neural changes of the aud itory system in instrumentally trained musicians; however, auditory neuroplasticit y in formally trained vocal musicians has been studied to a much lesser extent.
88 Purpose of the Study This study is a beginning step of inquir y into the effects of long-term professional music training on the auditory neural functi on of vocal musicians. Reliable evidence suggests that formally traine d instrumental musicians expe rience neural changes in the auditory system following skill acquisition and sensory stimulation. Vocal musicians undergo similar intensive training; howeve r, it is unknown whethe r this class of musicians also experiences neural changes of the auditory system. Prior EEG data indicates that formally trai ned instrumental musicians compared to nonmusicians have superior pre-attentive auditory discrimi nation; however, it is unknown whether this superior ability is also present in vocal musicians. Relationships among the following variab les for perception and production of musical stimuli were tested: vocal pitc h matching accuracy, active auditory pitch discrimination, and pre-attentiv e auditory pitch discriminati on. In general, this study sought to determine relationships among these variables betw een the untrained population and formally trained musicians. Moreover, it was questioned whether differences exist between subc lasses of formally trained musicians, such as vocal musicians and instrumental musicians. Sp ecifically, the purpose of this study was to assess, compare and correlate three identified vari ables of perception and production that contribute to the performan ce of the singing voice.
89 Research Questions Hypotheses Based on a review of previous investigati ons and on theories of neural plasticity, it was hypothesized that an association exists between perception and production abilities for musical stimuli and that th is association would be str onger for the formally trained musicians and strongest for the formally trained vocal musician s. Moreover, due to the requirements for precise pitch control, such as auditory perceptual monitoring and proprioceptive feedback of the laryngeal syst em, it was predicted that those with formal vocal training would perform best on the pe rception and performance tasks chosen for this study. It was also hypot hesized that intensive music training affects pre-attentive neurophysiological function. Consequently, the formally trained musicians were predicted to respond to small deviances in p itch with a faster and stronger neurological response than the nonmusicians on a pre-atten tive auditory discrimination task measured by electroencephalography (EEG). Questions The relationship between perception and production abilities for musical stimuli between formally trained musicians and musi cally untrained subjects was examined. Specifically, relationships between vocal pi tch matching accuracy, active auditory pitch discrimination, and pre-atten tive auditory pitch discrimi nation among formally trained vocal musicians, formally trained instrument al musicians and a matched control group of
90 musically untrained participants was investig ated. This study was de signed to answer the following questions: 1. Is there a difference in vocal pitch matching accuracy between musicians and nonmusician control subjects and furtherm ore, is there a difference between the instrumental and vocal musician groups? 2. Is there a difference in active auditory frequency discrimination ability between musicians and the control subjects? Mo reover, is there a difference between the instrumental and vocal musician groups? 3. Is there a difference in pre-attentive aud itory neural responses to pitch change (i.e., pre-attentive auditory pitch disc rimination for musical stimuli) between musicians and the control subjects and pa rticularly between the instrumental and vocal musician groups? 4. Is there an overall correlation betwee n perception and production variables across the groups? 5. Is there a correlation between percepti on and production variables within each subject group (i.e., controls, instrument al musicians, and vocal musicians)? Null Hypotheses Based on the above research questions, the following were the null hypotheses: 1. There is no difference of vocal pitch ma tching accuracy between musicians and control subjects, or between the tw o sub-classes of musicians (i.e., instrumentalists and vocalists).
91 2. There is no difference of auditory disc rimination ability between musicians and control subjects, or between the tw o sub-classes of musicians (i.e., instrumentalists and vocalists). 3. There is no difference of the pre-atten tive auditory neural response (i.e., preattentive auditory pitch discrimination) for musical stimuli between musicians and control subjects or between the instru mental and vocal musician groups. 4. There is no correlation of perception and production abilities among the groups. 5. There is no correlation of perception and production abilities w ithin each subject group (i.e., controls, instrumental musicians, and vocal musicians). Research Design A quantitative research design following a causal-comparative (ex post facto) format was proposed. Three performance meas ures were analyzed: two psychoacoustic and one neurophysiological. The two psychoacoustic tasks were designed to assess subjects vocal pitch matching a ccuracy and active auditory pitch discrimination ability. The neurophysiological task was used to meas ure subjects pre-atte ntive auditory pitch discrimination ability by mean s of electroencephalography (EEG ). The results of these three tasks were assessed and compared w ithin and between formally trained vocal musicians, formally trained instrumental musicians, and a matched control group of musically untrained subjects. Administration of the neurophysiological task followed the psychoacoustic measures by approximately one month to avoid a short-term memory effect of stimuli on
92 the participants responses. Presenta tion of the two psychoacoustic tasks was counterbalanced to control for a possible primi ng effect. Within each task, the order of stimulus conditions was randomized to prevent an order effect. Variables Independent variable. There was one independent variable, subject group which was subdivided into formally trained vocal musicians, formally trained instrumental musicians, and a matched control group of subj ects with less than 12 months of formal music training. This variable is considered an attribute or an assigned variable since it was not actively manipulated. Dependent variables. There were three dependent variables measured: vocal pitch matching accuracy, active auditory pitch discrimination, and pre-attentive auditory pitch discrimination. These dependent va riables were measured and reported respectively as: (1) relative accuracy for vocal pitch production in percentage ( rel PPA%), (2) relative difference limen for frequency in percentage ( rel DLF%), and (3) latency and amplitude of the mismatch negativity (MMN ) and other auditory evoked potentials (AEPs) associated with audito ry discrimination and perceptio n. An operational definition for each dependent variable is described in the section corresponding to the task.
93 Procedures Participant Selection Sixty-one females participated in this study; that is, 19 formally trained vocal musicians, 21 formally trained instrumental musicians, and 21 nonmusicians. Three groups of subjects were selected to partic ipate in this study us ing a nonrandom purposive sampling. All participant volunteers were as ked to initially complete a Participant Screening Questionnaire to determine if the volunteer met the qualific ations for this study (Appendix A). Exclusion criteria. To control for extraneous variables that may affect the larynx and therefore influence vocal quality and/or pitch production, exclus ionary prerequisites for all participants included : (a) no history of laryng eal pathology or neurological impairments, (b) no history of drug or alcohol abuse, (c) no history of habitual cigarette smoking, (d) no current allergies or respirat ory illnesses, and (e) no voice problems at time of testing. To control for factors th at may influence aud itory pitch perception, exclusionary criteria also included: (f ) no hearing impairment, (g) no previous participation in psychoacoustic studies, and (h ) no history of abso lute pitch ability. Previous participation in psychoacoustic experiments has been noted as a possible confounding variable (Ari-Evan Roth, Amir, Alaluf, Buchsenspanner, & Kishon-Rabin, 2003; Spiegel & Watson, 1984). Since prior ex perience with psycho acoustic testing may have a learning effect on responses of fre quency discrimination, subjects with this previous experience were excluded.
94 Absolute pitch (AP), also referred to as perfect pitch, is the ability to identify by musical note name the pitch of any sound without reference to another sound, or by producing a musical tone when given the mu sical note name (Zatorre, Perry, Beckett, Westbury, & Evans, 1998). Since the exact e tiology and neural char acteristics for this ability are controversial, th is special ability is a possi ble confounding variable. Thus, unless AP is the topic of the study, it is usuall y an exclusionary criterion for studies of music perception. Musicians know if they po sses this ability and were simply asked during the subject selection pro cess. An explanation of th e concept was given to those control subjects who were not familiar with AP. Any who claimed this ability were excluded from this study. Inclusion criteria. To be included in this study, al l participants passed an airconduction hearing screening at 25 dB HL for the frequencies of 250, 500, 1000, 2000, and 4000 Hz bilaterally. Hearing was screen ed with a GSI 17 (Grason-Stadler, Inc. Model 1717) portable screening audiometer ( ANSI, 1996). In addition, all participants successfully imitated a vocal sweep of frequenc y stimuli to ensure that the experimental stimuli were within their dynamic vocal range The vocal sweep was cued by a chromatic pitch instrument and modeled one octave from A3 (220 Hz) to A4 (440 Hz) inclusive of the pitch stimuli for this study. The three groups were approximately matched for age and education. In addition to these criteria, other inclus ion criteria for all subjects included: (a) gender, (b) language background, an d (c) handedness. To control for gender effects, all subjects were female. Gender effects on br ain symmetry have been detected by voxel-
95 based methods and fMRI (Good et al., 2001; Schlaug, 2001). Gender has also been found to influence the MMN latency. MMN la tency was found to be significantly longer for females than males for automatic discrimination of complex stimuli (Aaltonen et al., 1994). Thus, since the effects of gender ar e uncertain, mixed gender may be considered a confounding variable. Due to implicit learni ng of culture-specific musical intervals and pitch inflections (e.g., tone languages), subjects were from Western cultures with English as their native language (Hauser & McDe rmott, 2003; Tillman, Bharucha, & Bigand, 2000; Welch, 1994). Volunteers who were fluent in tone languages (e.g. Vietnamese and Mandarin) were excluded. Re search indicates that speaker s of these languages have superior pitch skills compared to speakers of nontone languages (Deutsch, Henthorn, Marvin, & Xu, 2004). Since neur al processes were being inve stigated, all subjects were right-handed so that conf ounding arguments regarding hemispheric dominance for the task were excluded. Once a volunteer qualified to participate in the study, she was given a Participant Information Questionnaire to complete (Appendix B). Questions pertained to the subjects education and music training. Although not controlled criteria, the age at which music and/or vocal training began and the num ber of years of training were noted. In addition, the incidence of any immediate fam ily members who had 5 years or more of formal music training was also recorded. Documented informed consent was obtained from all subjects in accordance with the ethical guidelines established by the University of South Floridas Instituti onal Review Board (Appendix C).
96 Formally trained vocal musicians. For the purpose of this investigation, formal music training referred to the participation in professionally dire cted and implemented music instruction and technical exercises provided by a professional musician and/or music educator (McNamara, 2005). A formally trained vocalist was defined as one who has had a minimum of five years of formal vo cal training and is either a performing artist, full-time music teacher, or full-time conserva tory student (Amir et al., 2003; Gaab & Schlaug, 2003; Gaser & Schlaug, 2003; Shahin et al., 2003). Partic ipants were drawn from the student population in the University of South Floridas (USF) School of Music in the College of Visual and Performing Arts. Formally trained instrumental musicians For the purpose of this investigation, a formally trained instrumental musician wa s one who had received a minimum of five years of formal music training to play a musi cal instrument within any of the following instrument categories: brass, wind, or string s. Those volunteers whose formal training focused mainly on percussion instruments were excluded since training for that category of instruments places greater emphasis on rhythm and tempo rather than pitch. Participants were either a performing ar tist, full-time music teacher, or full-time conservatory student (Amir et al., 2003; G aab & Schlaug, 2003; Gaser & Schlaug, 2003; Shahin et al., 2003). Participants were drawn from the student population in the University of South Floridas (USF) School of Music in the College of Visual and Performing Arts. Control subjects. A group of female students ap proximately matched to the other two groups for age and education were recr uited from USF. In addition to the
97 inclusion/exclusion criteria discussed above, the subjects in the control group had less than twelve months of formal musical or vocal training and di d not play a musical instrument. Exposure to general music educ ation in school curricu lum is common to the general student population and was not considered formal training. Stimuli Considerations Harmonic versus pure tone selection. The large majority of pitch perception experiments in psychoacoustics and neurosci ence use sinusoidal tones (pure tones) consisting of only the fundamental freque ncy (Novitski, Tervaniemi, Huotilainen, & Ntnen, 2004). However, pur e tones are artificial and not a part of our natural acoustic environment. All naturally occurring peri odic sounds have a sound spectrum consisting of a time-varying pattern of multiple harmonic partials across a wide frequency range (Novitski et al., 2004). Musica l sounds, in particular, ar e temporally, spectrally, and structurally complex. Event related potentials (ERPs) derived from electroencephalography (EEG) have consisten tly demonstrated greater sensitivity of the auditory processing sy stem for harmonic complexes rather than pure tones (Novitski et al., 2004; Pantev et al., 1998; Sh ahin et al., 2003). It is be lieved that the addition of acoustical information facilitates neural encoding (Tervaniemi & Brattico, 2004). Tervaniemi et al. (2000) systematically inve stigated the facilitati ng effects of overtones on the accuracy of pitch discrimination. ERP da ta revealed that the amplitude for the mismatch negativity task (MMN) was enhan ced for complex sounds when compared to pure tones, but there were no significant diffe rences between spectrally rich tones having
98 3 to 5 partials. Thus, stimuli for each task in this study consisted of harmonic tone complexes that approximated the physical char acteristics of piano tones. Each stimulus contained a fundamental frequency (F0) and th e first three harmonics. The amplitude of each harmonic was divided by its harmonic number to create a natural amplitude contour in the frequency spectrum. Frequency selection. Since the assumption underlying th is investigation was that pitch production accuracy is related to a uditory frequency perception, the frequency stimuli for all three tasks were chosen from the mid-frequency range of the untrained female vocal register and extended from mu sic tones C4 to G4 (F0 = 261.63 Hz to F0 = 392 Hz) (Hirano, 1981). In order to better co mpare and correlate vocal pitch production accuracy to the two auditory discriminati on tasks, the following harmonic complexes were digitally generated using Tucker-Davis Technologies (TDT) hardware at a sampling rate of 50,000 Hz: C4 (F0 = 261.63 Hz), 3% increase from C4 (F0 = 269.48 Hz), 6% increase from C4 (F0 = 277.32 Hz or C4#), E4 (F0 = 329.63 Hz), 3% decrease from E4 (F0 = 320.03 Hz), 6% decrease from E4 (F0 = 310.97 Hz or D4#), G4 (F0 = 392 Hz), 3% decrease from G4 (F0 = 380.58 Hz), and 6% decrease from G4 (F0 = 369.81 Hz or F4#). Thus, the stimuli included three whole tone s (C4, E4, and G4), three semitones (C4#, D4#, and F4#), and three additional synthesized quarter tones. The frequency difference between two adjacent whole tones is approximately 12%. For example, a 12% frequency increase added to C4 (F0 = 261.63 Hz) creates D4 (F0 = 293.03 Hz). The difference in frequency of the adjacent semitone (e.g., C4#) is half; that is, 6%. Thus, the frequency
99 difference between two contiguous quarter t ones equals 3%; the frequency difference between eighth tones is 1.5% and so on. Equipment Common to All Tasks For all tasks, stimuli were digitally ge nerated, controlled, and presented using a Tucker-Davis Technologies (TDT) RP2 R eal-Time Processor with model HB 7 headphone buffer. Stimuli were generated at a sampling rate of 50,000 Hz. Locally written software using Microsoft Visual Basi c 6.0 was developed to present the stimuli, record subject responses to the stimuli, a nd to calculate difference limens of frequency for each condition. Harmonic complexes were calibrated to 75 dB SPL using a Brel and Kjaer Type I Precision Sound Level Meter (T ype 2235) and a Brel and Kjaer inch condenser microphone (Type 4134) with a 2cc coupler (model DB 0138). The harmonic complexes were presented via Etymotic Rese arch (ER2) insert earphones at 75 dB SPL. Vocal Pitch Matching Pitch is the perceptual correlate of frequency. A vocal pitch is measured as the fundamental frequency (F0) (Patel & Balaban, 2001). For vocal musici ans, the ability to rapidly change and produce an accurate pitch is essential. Procedure for vocal pitch matching. In addition to insert earphones, subjects wore a head-set microphone (Parrott Tran slator VXI) positioned at a constant microphone-to-mouth distance of one inch and pl aced off-center at the right corner of the
100 mouth. Each participant was instructed to list en to each stimulus until it ended and then to reproduce it on the vowel /a/ at the same pitch as accurately as possible for 3 seconds. Each stimulus tone was one second in durati on and was randomly pr esented three times, totaling 27 stimuli. Inter-stimulus interval (ISI) was manually controlled by the examiner allowing time for the subject to respond prior to the next presentation. To verify comprehension of the task, each subject had two practice opportunities using a synthesized piano tone other than the te st stimuli (e.g., B3, F0 = 246.94 Hz). The productions were directly recorded into a Dell Inspiron (model 2650) laptop computer using a sampling rate of 22,050 Hz and 16 bits per sample. Data analysis. An autocorrelation analysis of fundamental frequency for each production was performed using Praat software (Version 4.4). The middle 50% of each production was selected as a representati ve portion to measure the fundamental frequency. For each subject, the mean funda mental for each target harmonic complex was calculated from this portion of each of the three productions. A group average and standard deviation for each target stimu lus production were calculated from the individual mean fundamentals. Relative pitch production accuracy in percent ( rel PPA%) was calculated as the absolute difference between the produced F0 and the targeted freque ncy relative to the targeted frequency in percent ( rel PPA%= f/f x 100) and a mean rel PPA% and standard deviation were determined for each group and each harmonic complex. The rel PPA% value decreases as the difference between the pr oduced F0 and the targeted F0 decreases. Rel PPA% is assumed to reflect the accuracy of pitch production and can be compared
101 and correlated with measurements of rel DLF%. Group data for vocal pitch production was also converted to semitones to co mpare to the Western musical scale. Active Auditory Frequency Discrimination The basic auditory skill of frequency discri mination refers to the ability to detect the smallest change in the frequency of two successive tones (Turner & Nelson, 1982). Not all changes in frequency are perceived. In order for sounds to be detected as differing in pitch, the frequenc y difference must be at l east equal to the subjects frequency discrimination threshold; that is the difference limen for frequency (DLF) (Gelfand, 1998). Frequency discrimination is one of the l east investigated psychoacoustic abilities in mu sicians, especially vocal musicians (Kishon-Rabin et al., 2001). Procedure for frequency difference limen (DLF). DLF is one of the most common and efficient methods for measuring auditory discrimination for frequencies below 2k Hz (Sek & Moore, 1995). An adaptiv e three-interval, three-alternative forcedchoice (3I/3AFC) paradigm was used in c onjunction with a thr ee-down, one-up stepping rule to estimate the frequency discrimi nation threshold yielding a 79.4% performance level for each subject and condition (Levitt, 1971; Moore & Peters, 1992). The 79% 3I/3AFC paradigm has been shown to be more efficient with less threshold bias than the 2I/2AFC method (Leek, 2001). The audito ry discrimination threshold may be underestimated using a 2I/2AFC procedure due to chance performance and the effects of guessing (e.g., 50% for 2I/2AFC, 33.3% fo r 3I/3AFC). Difference limens were
102 determined for three conditions, that is, th ree harmonic complexes: F0 = 261.63 Hz (C4), F0 = 329.63 Hz (E4) and F0 = 392 Hz (G4). Each harmonic complex was 200 ms in duration including a 10 ms rise and fall time with an interstimulus interval (ISI) of 300 ms (Moore & Peters, 1992). Interstimulus interv al refers to the time from the end of the previous stimulus to the beginning of the next stimulus. For this experiment, subjects were seated in front of a computer monitor in a sound treated booth. There were three obser vation intervals for each trial. These intervals were represented by three boxes that appeared on th e monitor screen. Subjects were instructed to select the box representi ng the sound that was different from the other two stimuli. The choice was made using a mouse click. Visual feedback was provided by a light under the co rrect selection box. The frequency difference between the harmonic complexes began with a 6% frequency difference (i.e., one semitone) be tween a target harmoni c stimulus (i.e., C4, E4, or G4) and a comparison stimulus. Pilot data indicated that 6% (a semitone difference) was easily detected by most subjects with no prior music training. For the first three reversals, the frequency differe nce changed by 3% or one quarter tone. Thereafter, the step-size was 0.375%. The di rection of frequency difference was above C4 and below E4 and G4 in order to correspond with piano keys and to remain within the mid-frequency range for the female voice. Following 10 reversals of response, the frequency difference of the final six reversal s was averaged and reported as the frequency difference limen (DL) for that run of that condition. A minimum of three runs of each condition was completed. When necessary, ad ditional runs were completed until three
103 consistent threshold estimates were obtained. Runs were cons idered consistent when DLs were within a factor of two. The final frequency discrimination threshold for each condition was determined by averaging the three most consistent runs. Data analysis. Average difference limen for fr equencies (DLFs) for harmonic complexes C4 (F0 = 261.53 Hz), E4 (F0 = 329.63 Hz), and G4 (F0 = 392 Hz) were determined for each subject. A group averag e and standard deviation for each harmonic complex was calculated from the individual av erages. In an effort to effectively organize, describe and compare data, relative values for frequency discrimination of synthesized piano stimuli (i.e., harmonic tone complexes) were determined relative to each targeted frequency (Amir et al., 2003; Kishon-Rabin et al., 2001). The minimal detectable changes in frequency ( f) were transformed to relative DLF thresholds in percent ( rel DLF%= f/f x 100) and a mean rel DLF% and standard deviation were determined for each group and each condition. Rel DLF% is assumed to reflect the accuracy of pitch discrimination and can be compared and correlated with measurements of rel PPA%. The minimal detectable change in frequency was also converted to semitones. Pre-Attentive Auditory Pitch Discrimination Neural imaging technology provides noninvasive methods to examine the structure and function of the brain and has become a mainstay of neuroscience. In contrast to the high spatial resolution a nd accuracy of functional magnetic resonance imaging (fMRI) and positron emission tomography (PET), the strength of
104 electroencephalography (EEG ) and its magnetic counterpart, magnetoencephalography (MEG), is in the time domain. EEG and MEG have excellent temporal resolution and allow for the study of processes in the brain within a millisecond of precision (Mody, 2004). EEG and MEG are non-invasive methods to record derived potentials that reflect the activity of a group of neurons in the human cortex (Pantev, Engelien, Candia, & Elbert, 2001). These procedures delineate the time course of neural activity associated with a stimulus and may localize the sour ce of the electromagnetic signal (George, Vikingstad, & Cao, 1998; Mody, 2004). EEG and MEG are particularly well suited for studying the neural representation of sound a nd speech at the acoustic level (Friesen & Tremblay, 2003). Electroencephalography is th e recording of spontaneous bio-electric activity generated by the central nervous system. EE G is often used in medical facilities to identify sleep and seizure disorders; howev er, electrophysiologic activity can also be measured in response to various sensory stimuli, including auditory, visual, and somatosensory stimuli (Friesen & Trembl ay, 2003). EEG is a noninvasive method of recording continuous electrical activity and ch anges in real-time cognitive processing in the brain from electrodes placed on the s calp (Mody, 2004; Pantev et al., 2001). This electrical activity is tracked by eliciting event-related potenti als (ERPs). An ERP reflects this electrical brain activity in waveforms consisting of positive and negative deflections or peaks (Sams, Paavilainen, Alho, & Ntnen, 1985). The ERPs are time-locked to stimulus events and provide information about electrical activity of cortical neurons as a
105 response to experimental stimuli (Koelsch & Friederici, 2003; Mody, 2004). ERPs in response to auditory stimuli are referred to as auditory evoked potentials (AEPs). Typically, AEP waveforms are described in terms of latency (time of occurrence after stimulus onset), polarity (positive or negative wavefo rm), amplitude (height) and amplitude distribution across electrode recording sites (F riesen & Tremblay, 2003; Mody, 2004). Latency is measured in milliseconds and refers to the time at which an evoked response wave component occurs after the pr esentation of a stimul us. The latencies and amplitudes of particular peaks have been a ssociated with specific cognitive operations (Mody, 2004). For the perception of auditory stimuli, early latency responses occur from 1-10 ms and measure the integrity of the cochlea and eighth cran ial nerve. Middle latency responses occur from 10-50 ms; the origin of these potentials is unclear. The late latency responses occur between 50 and 700 ms after the stimulus onset and reflect neural processing in the cortex (Friesen & Tr emblay, 2003; Goldstein & Aldrich, 1999). Common AEP components for the study of the auditory cortex include the P1-N1-P2 complex, mismatch negativity (MMN), early right anterior negativity (ERAN) and the P300. Mismatch negativity (MMN). The AEP component of prime interest for this study was mismatch negativity (MMN) which is typi cally maximal in adults at fronto-central electrode sites and reflects a pre-attentive ac oustic discrimination pr ocess based on neural representations of acoustic repetitions or regularities (Mody, 2004; Morr, Shafer, Kreuzer, & Kurtzberg, 2002; N tnen, Gaillard, & Mntys alo, 1978; Opitz, Mecklinger, Friederici, & von Cramon, 1999; Schrge r & Winkler, 1995; Tervaniemi, 2001;
106 Tervaniemi, Medvedev, Alho, et al., 2000). It is a negative wave that occurs approximately 100-250 ms post-stimulus (Pettig rew et al., 2004). The timing and location of the neural activity underly ing the MMN suggest that this neurophysiological response reflects access to an acoustic memory early in central auditory information processing (Nousak, Deacon, Ritter, & Vaughan, 1996). MMN provides an objective index of i ndividual discrimination ability for different acoustic features (Ntnen, 1995; Ntnen, Pakarinen, Rinne, & Takegata, 2004). It is elicited by infreque nt deviations in simple acoustic parameters including frequency, duration, intensity, or relations between sounds such as intervals or melody contour (Nager, Teder-Slejrvi, Kunze, & Mnte, 2003; Pettigrew et al., 2004; Schrger & Winkler, 1995; Trainor, McDonald, & Alain, 2002). A sequence of standard auditory input establishes a memory trace; thus, a de viation from this memory trace generates a mismatch response reflecting the automatic dete ction of change in a stream of auditory information (Koelsch & Friederici, 2003; Menning et al., 2000). The MMN can be elicited by very small changes in stimuli th at approximate the difference limen (i.e., perceptual discrimination th reshold) (Sams et al., 1985). An MMN response is termed pre-attentiv e as it represents aspects of acoustic information that are encoded without the cons cious attention of the subject (Tervaniemi, 2001). The MMN is a highly automatic, passive ly elicited (not requiring subject participation) neurophysiologica l response to an acoustically different (deviant) stimulus when presented in a series of homogeneous (standard) stimuli (Nt nen et al., 1978). Since the MMN response can be elicited independent of a ttention, it is free from
107 attentional activities that may contamin ate behavioral and attention-dependent physiological measures (Ntnen, 1995). Because of its excellent temporal re solution, the MMN provides a tool for determining the speed of sound-change discrimi nation. This automatic neural response to stimulus change has been found to co-vary re liably with perceptual and higher cognitive functions and is believed to reflect the neurophysiological processes that underlie auditory discrimination (Sams et al., 1985; Te rvaniemi et al., 2005). For example, Lang and colleagues verified a strong relationshi p between the MMN amplitude recorded in non-attended conditions and behavioral measures of pitch discrimination ability (Lang et al., 1995). Studies on automatic neural encoding of music prior to conscious attention have employed the mismatch negativity (MMN) para digm (Tervaniemi, 2001). The value of the change-detection paradigm in the neurosci ence of music is its su itability for the study of neural memory function among musi cians and nonmusicians without being contaminated by attention, mo tivation, or demands of the task (Novitski, Tervaniemi, Huotilainen, & Ntnen, 2004; Tervaniemi & Huotilainen, 2003). The MMN provides an objective featureand stimul us-specific measure of audito ry discrimination ability as well as an objective index of training-indu ced functional neural plasticity (Ntnen, 1995). It provides a method to study the indi vidual components of musical talent (e.g., pitch, interval, duration, rhythm perception) and the effects of music training on these components (Ntnen, 1995). The correlation between MMN parameters and behavioral measures, such as reaction time and hit rate, imply that pre-attentive neural functions
108 shape the accuracy of subsequent attentiv e processes (Tervaniemi, 2001), thus rendering the MMN an optimal tool for studying the ne ural bases of human auditory perception (Menning, Roberts, & Pantev, 2000; Novitski et al., 2004; Pantev et al., 2003). P100-N100-P200. Other AEP components to be m easured in this study included the P100-N100-P200 (P1-N1-P2) complex. The P1 -N1-P2 complex is the earliest pattern of negative and positive waveforms to occur in the Late Latency Response (LLR) group. These components primarily represent sensoryperceptual stages of processing and may be elicited without the subject s attention to the stimuli (B ertoli et al., 2004). The P1 component, referred to as the P100, is the fi rst positive wave peak and is recorded approximately 30-80 ms after the onset of an auditory stimulus (Ber toli et al., 2004). The P1 is thought to be a neurophysiological inde x of preferential a ttention and is often associated with auditory inhibition and s uppression of unattended information (Key, Dove, & Maguire, 2005). P1 is associated with sensory gati ng and thus may reflect the ability of the brain to modulate its sensitivity to incoming stimuli (Braff & Geyer, 1990). Sources of the auditory P1 have been lo calized in the superior temporal gyrus. The N1 component, also referred to as N100, is a negative wa ve peak recorded approximately 80-100 ms after stimulus onset (Mody, 2004; Wood & Wolpaw, 1982). It reflects activation of the large neuronal populat ion in regions of the auditory cortex on the superior surface of the temporal lobe a nd can be reliably measured in individual subjects (Menning, Roberts, & Pantev, 2000; Mody, 2004). Sources for the N1 response include the primary auditory cortex, the s upratemporal plane ante rior to the primary auditory cortex and the temporal cortex (Scherg, Vajsar, & Picton, 1989). The recorded
109 N1 varies as a function of stimulus intensit y, presentation rate, and attention (Ntnen & Picton, 1987). The N1 may be augmented by plastici ty that takes place ei ther cortically or at subcortical sites that project to the aud itory cortex (Menning et al., 2000; Trainor, Shahin, & Roberts, 2003). The robustness of the N1 response holds enormous potential for studying the physiology of the a uditory cortex (Mody, 2004). The P2, known as P200, is the second positive peak and is recorded at approximately 180-250 ms after stimulu s onset (Wood & Wolpaw, 1982). The generators of P2 appear to lie within the audito ry cortex. It appears to be sensitive to the physical parameters of the stimulus, such as pitch and intens ity (Novak, Ritter, & Vaughan, 1992). The P2 has been found to di ffer between musicians and nonmusicians for musical stimuli and has been enhanced in nonmusician adults by auditory training (Bosnyak, Eaton, & Roberts, 2002 as c ited in Trainor et al., 2003). Thus, the P1-N1-P2 components are sensit ive to physical changes in auditory stimuli and reflect a pre-attentive neural re sponse to auditory stimuli. These components were compared among the subject groups prior to measurement of the MMN to establish that encoding of auditory stimuli at the level of the auditory cortex was similar for all groups. A correlation between individual pitch discrimination performance and the amplitude or latency of the P1-N1-P2 components has not been found (Sams et al., 1985). Electroencephalographic recording. The MMN and P1-N1-P2 responses were recorded and analyzed using a Compumedics Neuroscan EEG system with a SynAmps 2 amplifier and Neuroscan Scan 4.3 acquisition so ftware. As described earlier, TuckerDavis Technologies (TDT) hardware and locally written software were used to generate,
110 control, and present harmonic tone comple xes. TDT hardware and locally written software were used to send triggers to th e Neuroscan system to mark the time of the stimulus presentation. A cap of 32 sintered electrodes was placed on the subjects head according to the International 10/20 recordi ng system (Jasper, 1958) using a conductive water-soluble paste applied between electr odes and the subject s scalp. Additional electrodes were placed above and below the left eye and at the outer canthus of each eye to monitor eyeblink activity. An electrode on the nose served as reference and an electrode on the forehead served as ground. After the electrodes were applied, the subject was seated in a reclining chair in a sound tr eated booth. Etymotic research (ER2) insert earphones were placed in the subjects ear canals for binaural presentation of the stimuli at 75 dB SPL. Electrode impedance was kept below 20k and was monitored periodically through-out the da ta acquisition and the continuo us EEG data were stored on the computer for off-line averaging. The acquisition of EEG data was by continuous sampling and was recorded at an AD (analog to digital) sampling rate of 1000 Hz. The raw signal was amplified within a frequency band of 0.05100 Hz. Procedure. To minimize any attentive aud itory behavior, the subject was instructed to rest comfortably and watch a closed captioned video of her choice as harmonic complexes were presented. The use of a primary task, such as reading a book or watching a closed captione d video is reported to si gnificantly reduce movement artifact while having no attenuating/enhanc ing effect on the MMN or the P1-N1-P2 response (Pettigrew et al., 2004; Sinkkonen & Tervaniemi, 2000). Moreover, the MMN
111 to frequency change seems to be unaffect ed and strongly indepe ndent of attention (Ntnen, 1995; Ntnen, Paavilai nen, Tiitinen, Jiang, & Alho, 1993). The listening task was structured as a multiple deviant oddball paradigm in which the subject was presented with a standard st imulus (70%) and three deviant stimuli (10% each). Multiple deviances may be embedded into the same sequence without significantly affecting the si ze of the response (Ntnen et al., 2004; Nousak et al., 1996). Not only is the multiple deviant met hod more time efficient (Ntnen et al., 2004; Nousak et al., 1996; Pettigrew et al., 2004), it is also a mo re ecologically valid paradigm. That is to say, our natural envi ronment is comprised of a wide variety of sound sources that the auditory system pre-a ttentively monitors simultaneously (Nager et al., 2003). This is particularly true for musicians performing in an orchestra or vocalists performing in a choir and/or w ith musical accompaniment. Previous research suggests that the amplit ude of the mismatch response is directly proportional to the logarithm of the stimulus probability (Sinkkonen et al., 1996). Rarely occurring deviant stimuli produce large MMN responses; however, too few deviant stimuli result in a poor signa l-to-noise ratio (Sinkkonen & Tervaniemi, 2000). A deviantstimulus probability between 0.05 and 0.2 has been demonstrated to yield reliable results (Sinkkonen & Tervaniemi, 2000). Thus, deviantstimulus probability for this study was established at 0.1 for each deviant target. An MMN response border is sometimes overlapped by N1 activity. This is often seen when there are large frequency differe nces between the standard stimulus and the deviants; however, deviances up to 10% are considered to produce relatively pure MMN
112 responses (Sinkkonen & Tervaniemi, 2000). Freq uency deviances for this study were 6% or smaller. The standard tone was a ha rmonic tone complex that approximated the physical characteristics of the piano tone G4 (F0 = 392 Hz). Based on the best and poorest DLFs obtained from the psychoacousti c task, the following three deviant tones were selected for the electrophysiological testing: 386.21 Hz (Deviant #1), 1.5% difference between the target (an eighth t one); 380.58 Hz (Deviant #2), 3% difference between the target (a quarte r tone); and 369.81 Hz (Deviant #3), 6% difference between the target (a semitone). Deviant #1 (1.5% difference) was slightly above the mean frequency discrimination threshold obtained fo r the musicians in the psychoacoustic task (i.e., 1.35%) and Deviant #2 (3% difference) wa s slightly below the average DLF for the nonmusicians (3.19%). Thus, the three sele cted deviant harmonic complexes were musically meaningful and represented a c ontinuum of behavioral performance. The stimuli were presented in a pseu dorandom sequence with at least three standard stimuli separating presentations of de viant stimuli; thus, tw o deviant stimuli did not occur in succession. Stimulus duration wa s 200 ms (including a 10 ms rise and fall time) with an interstimulus interval (ISI) of 500 ms pres ented at 75 dB SPL (Bertoli, Smurzynski, & Probst, 2005). The standard tone occurred on 70% (minimum of 2000) of the trials and each deviant occurred on 10% (minimum of 200) of the trials for a minimum of 2600 stimuli. With the excepti on of only two particip ants, the averaged responses to each deviant condition in the multi-paradigm contained a minimum of 185 accepted trials. Due to subtraction of consid erable movement artifact, the two exceptions had a minimum of 120 accepted trials for each deviant condition.
113 Each of the three deviant stimuli was pr esented alone in a single block of 300 stimuli with the same presentation time and IS I to establish a baseline response to each deviant stimulus when presented as a sta ndard. ERP responses to the deviant alone conditions were used to cal culate difference waveforms and to ascertain common waveform conditions (e.g., P1-N1-P2) for the subject groups (Lang et al., 1995). Presentation order of the deviant stimuli was randomized. Total time including electrode setup for the neurophysiological ta sk was approximately 65 minutes. Data analysis. All electrophysiological measurements were made with the use of Compumedics Neuroscan SynAmps 2 amplifier and Scan v. 4.3 acquisition and edit software. Following data collection, con tinuous EEG waveforms were examined and areas of large muscular ar tifact were rejected by hand. As a precaution for data analysis, the first 10 ERP responses of each stimulus block were omitted from the averaging process to exclude the variation of the N1 am plitude (i.e., the refractoriness) associated with the start of the stimulation se quence (Pekkonen, Rinne, & Ntnen, 1995; Pettigrew et al., 2002; Sinkkonen & Tervaniemi 2000). EEG epochs of 350 ms, starting 50 ms prior to stimulus onset were obtained, baseline correc ted (-50-0 ms), and averaged separately for the standard (deviant alone) and deviant/target stimuli. To eliminate ocular movement contamination, epochs containing artifacts exceeding 80 V in the HEOG and VEOG channels were rejected from averag ing. ERP waves were digitally band-pass filtered at 1-30 Hz. Arithmetic re-referencing of the contributing ERP waveforms to the average of the mastoids has been shown to maximize the MMN amplitudes at the frontal and central electrodes (Fpz, F4, Fz, F3, C4, Cz or C3) (Ntnen, 1995; Pettigrew et al.,
114 2002; Sinkkonen & Tervaniemi, 2000). In order to maximize signal-tonoise ratio, all of the processed average files were indivi dually re-referenced to the mastoids. The mismatch negativity response (MMN) is traditionally measured as a difference waveform obtained by subtracti ng the grand average ERP response to the standard stimulus from the grand average ER P response to a deviant stimulus (Morr et al., 2002; Novitski et al., 2004; Schrger & Winkler, 1995). The difference waveform is considered to reflect the differential neurona l processing of a deviant stimulus compared to the standard stimulus (Tervaniemi, 2001). A number of studies have demonstrated that the first standard following a deviant in an odd-ball paradigm may be perceived as a new standard and thus, the MMN response in the difference waveform is attenuated (Nousak et al., 1996; Sams, Alho, & Ntn en, 1983). To account for possible contamination and/or attenuation of the MMN response, an acceptable alternative method was chosen to calculate the difference wave forms. Using this method, the averaged response to a deviant when it is presented a nd recorded alone is subtracted from the averaged response to the deviant when it is presented in an oddball paradigm (Pettigrew et al., 2004; Picton, 1995). In addition to determining the MMN difference waveforms, the averaged group ERP waveforms for each deviant alone condition were used to examine the P1-N1-P2 complex.
115 Chapter Four Results Participant Demographics Participants were 61 fema le students from the Univ ersity of South Florida including 21 nonmusician controls, 21 instru mental musicians, and 19 vocal musicians. Ages of those in the control group ranged fr om 20 to 34 with a mean of 23.4. The musicians were closely matched with a mean age of 21.8 and a range of 18 to 33. None of the women in the control group had more th an 12 months of formal music training. By comparison, the musicians began music traini ng between the ages of 3 and 15 with a beginning mean age of 8.5 years. More than half of the musicians (21 of 40) began music training at age 9 or younger. On averag e, the instrumental musicians had 9.8 years of music training; 11 of the in strumentalists played a string instrument, while 10 played a wind instrument. The vocal musicians had an av erage of 11.3 years of music training. In addition to formal vocal training, almost half of the singers also received training on a musical instrument (9 of 19). Only one part icipant in the control group reported having a parent who had received 5 or more years of music training; while one-third of the musicians (13 of 40) had at leas t one musically trained parent. Self-reported Scholastic Assessment Test (SAT) scores indicated there was no difference between nonmusicians and musicians for either math or verbal abilities.
116 Subtest medians for both subject groups were in the 500-599 range. Participant profile data may be found in Appendix D. All 61 participants completed the two ps ychoacoustic tasks; however, one control subject and one instrumentalist did not complete the electrophysiological task. Consequently, data for the electrophysiol ogical task is based on 20 controls, 20 instrumental musicians, and 19 vocal musicians. Psychoacoustic Measures Vocal Pitch Matching Every participant attempted to vocally match the pitch of three presentations of nine harmonic complexes presented in random order for a total of 27 responses. The harmonic complexes represent the Western musical scale from C4 to G4 (F0 = 261.63 Hz, F0 = 269.48 Hz, F0 = 277.32 Hz, F0 = 329.63 Hz, F0 = 320.03 Hz, F0 = 310.97 Hz, F0 = 392 Hz, F0 = 380.58 Hz, and F0 = 369.81 Hz). Individual subject data for the averaged productions of the targeted ha rmonic complexes are shown in Appendix E. The descriptive group data for pitc h production accuracy (PPA) is displayed in Table 1. Table 1. Pitch Production Accuracy Group Data Group N Hz SD rel % Semitone Control 21 25.5 31.3 7.83% 1.30 Musician 40 4.1 5.4 1.28% 0.21 Instrumental 21 4.9 7.0 1.50% 0.25 Vocal 19 3.4 2.6 1.05% 0.17
117 On average, pitch production accuracy ( rel PPA%) for those without music training deviated from the reference pitch by 7.83% or 25.5 Hz. In other words, the average pitch difference between the target and the producti on for the nonmusicians was greater than one semitone (6% frequency di fference). By comparison, the musicians mean pitch production accuracy was within 1.28 % (4.1 Hz) of the target or within oneeighth of the reference tone. Interestingl y, both subgroups of musicians vocally matched pitch with comparable accuracy. Mean rel PPA% for the vocal musicians was 1.05% (SD = 2.6 Hz), while the instrumental musicians with no vocal training matched the reference pitch with an average rel PPA% of 1.50% (SD = 7.0 Hz). Statistical analysis. Measurements of psychoacoustic variables are reported in terms of relative accuracy in percent ( rel PPA% and rel DLF%) and thus represent a rank ordering of observations ra ther than precise measurements. Consequently, nonparametric statistics were deemed more a ppropriate for analysis of the psychoacoustic variables. Analysis was completed using SPSS software (version 11.0). An alpha level of .05 was used for all statistical tests. Agreement between the harmonic tone complex stimuli for the pitch production (PPA) task was measured using Kendalls W coefficient of concordanc e. The W ranges from 0 to 1 with 1 indicating complete agreement and 0 indicating complete disagreement (Barry, n.d.). PPA perfor mance did not differ across the harmonic complexes for nonmusicians (W = 0.08, p = .10) and instrumentalists (W = 0.04, p = .57); however, a difference existed among the stim uli for performance by the vocalists (W = 0.13, p = .02).
118 Because there was not complete conc ordance among the stimuli, the MannWhitney U test was used to compare differences between the musicians and nonmusicians and each harmonic complex for an effect of harmonic tone on pitch production accuracy (PPA). The Mann-Whitney U Test is a non-parametric procedure used to evaluate the differences between two independent samples. It is not dependent on the assumption of normal distri bution and is appropriate for sample sizes having less than 100 observations. Measures are placed into a composite distribution then ranked from the highest to the lowest scores to determin e if the ranks tend to be higher for one group (Glass & Hopkins, 1996). Overall, the musicians were more acc urate than the nonmusicians for all 9 stimulus tones, implying that the musicians rel PPA% was significantly smaller. However, when the Bonferroni correction was ap plied to compensate for Type I errors (p = .006), one of the nine harmonic complexes wa s not significantly different between the musicians and controls (369.81/F4#, U = 274, p = .02). Considering the total pitch production data between the musicians and the controls, the effect of this one harmonic complex on pitch production accuracy was fe lt to be inconseque ntial to the overall results. Within the musician group, pitch production accuracy was not significantly different for any of the nine harmonic co mplexes confirming no si gnificant difference between the vocal musicians and the instru mental musicians for vocal pitch matching accuracy.
119 Although the rel PPA% means of the two musician groups were not significantly different (PPA, U = 199, p = .99), the vocalists appeared to have minimal variability in production compared to the instrumental musi cians (SD = 2.6 Hz for vocalists and SD = 7 Hz for instrumentalists). The shapes of variance distribution between the instrumentalists and vocalists were compared with the Wald-Wolfowitz Runs test (Barry, n.d.). Distribution of variance was significan tly smaller for the vocalists (PPA, Z = 2.39, p < .001) compared to the instrumentalists, re presenting a less variab le, more consistent and uniform pattern of response for the vocal musicians. Active Auditory Frequency Discrimination A difference limen for frequency (DLF) was determined for each individual for three harmonic complexes C4 (F0 = 261.53 Hz), E4 (F0 = 329.63 Hz), and G4 (F0 = 392 Hz) using an adaptive 79% 3I/3AFC paradigm Individual data are shown in Appendix F. The descriptive group data for DLF is displayed in Table 2. Table 2. Difference Limen for Frequency Group Data Group N Hz SD rel % Semitone Control 21 10.3 11.7 3.19% 0.53 Musician 40 4.4 1.4 1.35% 0.23 Instrumental 21 4.5 1.6 1.40% 0.23 Vocal 19 4.2 1.1 1.30% 0.22 The just noticeable difference between two harmonic complexes for the nonmusicians was greater than one quarter of a musical tone ( rel DLF% = 3.19%). By
120 contrast, trained musicians discriminated be tween two pitches with only one-eighth of a difference in tone frequency ( rel DLF% = 1.35%). As a group, the musicians average rel DLF% was at least 50% smaller compared to the control group. The instrumentalists and the vocalists had comparable difference limens for frequency ( rel DLF% = 1.4% and 1.3%, respectively). Statistical analysis. Comparison of group performance using the Mann-Whitney U test indicated that measures of differen ce limen for frequency (DLF) were significantly smaller for the musicians compared to the controls (DLF, U = 146, p < .001). Auditory pitch discrimination for harmonic complexes di d not differ between the instrumentalists and the vocalists (DLF, U = 170, p = .42). The vocalists and the instrumentalists perceived all harmonic complexes equally well (Kendalls W = 0.003, p = .95; W = 0.05, p = .37, respectively); however, the nonmusicians did not demonstrate perceptu al agreement among th e three tones (W = 0.25, p = .005). Consequently, comparisons between subject group and each harmonic complex were completed using the Mann-Whitn ey U to evaluate for an effect of harmonic tone on the DLF. Musicians and nonmusicians differed significantly for each tone and for the tones overall indicating no effect of stimul us on pitch discrimination and confirming a smaller DLF for the musicians than for the nonmusicians for each harmonic complex (261.63/C4, U = 93.5, p = .001; 329.63/E4, U = 97, p = .002; 392.00/G4, U = 118, p = .01). Between the two subclasses of musi cians (vocalists and instrumentalists), difference limens for frequency (DLF) we re equivalent for all of the harmonic complexes.
121 Comparison of DLF and PPA The means of the DLF and PPA tasks were compared with dependent t-tests to assess whether these abilities were equivale nt within each group. The nonmusicians auditory skills were significantly more accurate than their vocal pitch matching ability, t(20) = 2.46, p = .02. Although the nonmusicia ns just noticeable difference (jnd) between harmonic complexe s averaged 3.19%, their rel PPA deviated 7.83% from the reference tone. By contrast, as a group, the musicians auditory discrimination and pitch production skills were comparable, t(39) = 0.22, p = .83. Interestingly, separate analysis of the two musician genres re vealed distinct differences. DLF and PPA abilities did not significantly differ for the instrumental musici ans, t(20) = 0.69, p = .50. However, for the vocalists, pitch production skills were more accu rate than auditory discrimination, t(19) = 3.17, p = .005. The vocal musicians produced a musical tone within 1.05% of a given reference, while their jnd averaged 1.30%. Th is difference reflects an influence of vocal training on laryngeal reflexes and sugges ts that long-term practice develops neuromuscular memory for accurate pitch pr oduction. Within all groups, auditory pitch discrimination tended to be less variable than pitch production accuracy. Correlation Analysis of DLF and PPA Group data was combined to examine th e overall relationship between auditory pitch discrimination ( rel DLF%) and pitch production accuracy ( rel PPA%). Individual pitch production averages as a function of i ndividual DLF averages are displayed in a scatter plot in Figure 1. Spearmans rho co rrelation was used to examine the overall
122 relationship between auditory pitch discrimination and pitc h production accuracy for all groups combined. This nonparametric measure assumes that the individual observations can be ranked into two ordered series (Cri chton, 1999). The rho coefficient values are between -1 and +1 with a positive correlation indicating that the ranks of both variables increase together. A negative correlation i ndicates that as the rank of one variable increases, the other decreases. 0.10% 1.00% 10.00% 100.00% 0.10%1.00%10.00%100.00% DLF (%)PPA (%) CONTROL INSTRUMENTAL VOCAL Figure 1. Scatter plot of DLF and PPA. Individual pitch production data ( rel PPA%) as a function of individual discrimination data ( rel DLF%) plotted on a logarithmic scale.
123 A large positive correlation based on ra nk order of all individuals was found between the two measures (rs = 0.61, p < .001). Visual insp ection of the scatter plot reveals that the vocalists are tightly clus tered between 1% and 2% for PPA and DLF. This illustrates the vocalists very small di stribution of variance. Because there was minimal difference in the DLF and PPA data between subjects, sufficient evidence for a relationship was not found for the vocalists (rs = -0.13, p = .60). Analysis also revealed no correlation between DLF and PPA data for the nonmusicians (rs = 0.38, p = .09). The scatter plot shows greater performance variab ility for this group, but no clear relational pattern. Thus, while there is a positive re lationship between DLF and PPA overall, the only group to actually have a si gnificant positive co rrelation between th ese two variables was the instrumental musicians (rs = 0.49, p = .03). Correlations of DLF and PPA with Questionnaire Variables Individual measures of DL F and PPA of all participants were correlated with the following Information Questionnaire variables: age of training onset (before age 9 and after age 9), years of music training (instr umental and vocal), number of immediate family members with music training (inc luding mother, father, and siblings), and Scholastic Assessment Test (SAT) scores (mat h and verbal). Ther e were no significant correlations within groups be tween any of the demographic variables and psychoacoustic task performance.
124 Electrophysiological Measures Pre-Attentive Auditory Pitch Discrimination Sensory perception P1-N1-P2 complex. For electrophysiological testing, the methods for determining the mismatch nega tivity (MMN) were based on the assumption that participants would demonstrate equivalent sensory perception of auditory stimuli. As described in Chapters 2 and 3, the P1-N 1-P2 components are sensitive to physical changes in auditory stimuli a nd reflect a pre-attentive neural response to auditory stimuli. Thus, the ERP group grand average waveforms for the deviant-alone conditions (standard stimuli) were visually inspected for similar ity of the P1-N1-P2 complex. Inspection of the waveforms revealed that mean latencies of the P1-N1-P2 comple x for all conditions fell within the range of 50-200 ms after stim ulus onset and there were no remarkable differences in average latency of ERP responses between the groups. However, P1 amplitude was consistently greater for th e control group (C-Dev) compared to the musicians (M-Dev) for all deviant-alone waveform s. Figures 2-A, 2-B, and 2-C illustrate these similarities and differences for each c ondition. The P1 amplitude is visualized as the first positive peak (up) occurring approxima tely 50 ms after the onset of the auditory stimulus. The N1 component is the firs t downward dip following the P1. The P2 component is the second positive peak.
125 Figure 2-A. Comparison of Group Average Waveforms of P1-N1-P2 Complex for Deviant 1 Alone Condition (386.21 Hz, 1.5% devian ce) at Fz between Controls (C-Dev1) and Musicians (M-Dev1) Figure 2-B. Comparison of Group Average Waveforms of P1-N1-P2 Complex for Deviant 2 Alone Condition (380.58 Hz, 3% devian ce) at Fz between Controls (C-Dev2) and Musicians (M-Dev2)
126 Figure 2-C. Comparison of Group Averag e Waveforms of P1-N1-P2 Complex for Deviant 3 Alone Condition (369.81Hz, 6% devian ce) at Fz between Controls (C-Dev3) and Musicians (M-Dev3) Individual peak amplitudes and latencies for P1 were determined for each deviant alone condition and group averages were cal culated for analysis (Appendices G and H). Analysis of variance (ANOVA) on P1 amplitude revealed a main effect of subject group, F(2,168) = 6.68, p = .002. Post-hoc group-wi se comparison (Tukey HSD) confirmed larger P1 amplitude for the control group (nonmusicians) compared to the instrumentalists (p = .001) and vocal musician s (p = .04). As discussed in Chapters 2 and 3, the P1 component is consider ed a neurophysiological indicato r of preferential attention to sensory stimuli and reflects differences in regulating excitatory and inhibitory processes. Thus, larger P1 amplitude sugge sts reduced sensory gating of stimuli (Key, Dove, & Maguire, 2005). There was no significa nt P1 amplitude difference between the instrumentalists and the vocalists (p = .53) in dicating that encoding of the stimuli at the level of the auditory cortex was similar for both groups of musicians. There was no
127 significant effect of stim uli, F(2,168) = 0.039, p = .96, nor was there an interaction between groups and stimuli for P1 amplitude, F(4,168) = 0.357, p = .84. Thus, participants demonstrated equivalent sensory perception of auditory stimuli as indicated by P1-N1-P2 latency and N1 and P2 amplitude. Group differences in P1 amplitude suggests that sensory gating may differ be tween musicians and nonmusicians. Mismatch negativity-MMN. Visual inspection of the grand average difference waveforms for all electrode sites was used to determine the electrode for subsequent analyses. The MMN response was strongest at electrodes Fz and Cz with the largest amplitudes measured at the fronto-central Fz electrode in the 10-20 system (Figure 3). Thus, reported measures and statistical anal ysis are based on ERP responses measured from Fz. MMN has been shown to invert in polarity at electrodes below the level of the Sylvian fissure (Ntnen, 1995). Polarity in version at the mastoids is an accepted method to verify the MMN response to tonal ch anges as illustrated in Figure 3 (Morr et al., 2002; Sinkkonen & Tervaniemi, 2000).
128 Figure 3. Grand Average Waveform Exampl es of Fz, Cz, and M2. Grand average waveforms show greater MMN amplitude for the Fz electrode and an inversion of polarity at the right mastoid site (M2). Musician example of response to Dev 3. Individual and group grand av erages were determined fo r each deviant contrast as an oddball and each deviant-alone condition (sta ndard). For each group, the average ERP response wave of the deviant alone conditi on (standard) was subtracted from the ERP response of the deviant as the odd-ball stimul us at Fz (Figures 4-A, 4-B, & 4-C). There were three deviant contrasts: 386.21 Hz (1.5%, Deviant #1), 380.58 Hz (3%, Deviant #2), and 369.81 Hz (6%, Deviant #3). Gra nd average group difference waveforms for each deviant were derived for the controls and the musicians as well as the musicians divided into their respective genr e, instrumental and vocal (Figures 5-A, 5-B, and 5-C).
129 Figure 4-A. Control Groups Mismatch Res ponse (MMN). Shaded area represents the average mismatch response of control group to 1.5% change (Dev 1, 386.21 Hz) from the standard tone (392 Hz), 3% (Dev 2, 380.58 Hz), and 6% (Dev 3, 369.81 Hz) change.
130 Figure 4-B. Instrumentalists Mismatch Re sponse (MMN). Shaded area represents average mismatch response of instrument al musicians to 1.5% change (Dev 1, 386.21 Hz) from standard tone (392 Hz), 3% (Dev 2, 380.58 Hz), and 6% (Dev 3, 369.81 Hz).
131 Figure 4-C. Vocal Musicians Mismatch Res ponse (MMN). Shaded area represents the average mismatch response of vocal musi cians to1.5% change (Dev 1, 386.21 Hz) from the standard tone (392 Hz), 3% (Dev 2, 380.58 Hz), and 6% (Dev 3, 369.81 Hz) change.
132 Figure 5-A. Grand Average Difference Waveforms for Deviant 1 (386.21 Hz, 1.5%) comparing control group (C), instru mentalists (IN) and vocalists (V). Figure 5-B. Grand Average Difference Waveforms for Deviant 2 (380.58 Hz, 3%) comparing control group (C), instru mentalists (IN) and vocalists (V).
133 Figure 5-C. Grand Average Difference Waveforms for Deviant 3 (369.81Hz, 6%) comparing control group (C), instru mentalists (IN) and vocalists (V). MMN responses may be quantified by the latency and amplitude of a negative peak in the difference waveform. A strong MMN response refers to large amplitude and short latency. Thus, a latency (in millis econds) and amplitude (in microvolts) were measured for each subject, group, and stimulus condition. Because the MMN amplitude and latency reflect two indepe ndent factors influencing the MMN, it is recommended that they be measured and analyzed separate ly (Lang et al., 1995; Ntnen, 1992). The grand average difference waveform s for each group and deviant stimulus were used to visually determine the latenc y region of the MMN peak amplitude. MMN peak latency is typically measured as the la rgest negative peak occu rring between 150 to
134 300 ms post-stimulus period for each group and frequency condition. In the present study, the MMN latency peaked between 177 and 227 ms depending on the magnitude of the deviance and the subject group (Table 3). Overall average peak latency was 201 ms. Latency windows of 40 ms around the peak for a given group and condition were determined. The average amplitude at Fz within these latency windows was calculated for each individual subject and condition (Appendix I). Within the pre-determined latency windows, MMN peak latency was de termined for each subject and condition (Appendix J). Table 3. Average Peak Latencies Deri ved from Grand Average Waveforms Statistical Analysis of the Mismatch Negativity MMN Amplitude For the control group, MMN amplitude increased as the magnitude of the frequency deviance became la rger. Specifically, as the frequency deviance increased from 1.5% to 3% to 6%, the average amp litude of MMN response for the nonmusicians Deviant MagnitudeControlMusicians Instrumental Vocal 1.5% (Deviant #1) 386.21 Hz 227 ms 212 ms 216 ms 207 ms 3% (Deviant #2) 380.58 Hz 199 ms 190 ms 195 ms 182 ms 6% (Deviant #3) 369.81 Hz 190 ms 185 ms 185 ms 177 ms
135 increased from -1.5 V to -2.0 to -2.5 V. This neural respon se pattern was not consistent for the musicians (Figure 6-A). -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 Controls MusiciansAmplitude (microvolts) Dev1 Dev2 Dev3 Figure 6-A. MMN Response Amplitude s by Group and Deviant Condition. Deviant magnitudes: Deviant 1 = 1. 5%, Deviant 2 = 3%, Deviant 3 = 6%. As seen in Figure 6-B, the instrumentalists responded with minima l variability between the three deviant stimuli (i.e., 0.3 V); wh ile the vocalists had the largest response amplitude for the smallest deviance magnitude (-2.4 V /1.5%).
136 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 Vocalists InstrumentalistsAmplitude (microvolts) Dev1 Dev2 Dev3 Figure 6-B. MMN Response Amplitude by Musician Genre and Deviant Condition. Deviant magnitudes: Deviant 1 = 1. 5%, Deviant 2 = 3%, Deviant 3 = 6%. To evaluate for the effect of group and deviant stimulus condition on the MMN amplitude, a two-way factorial analysis of variance (ANOVA) was conducted on the individual mean amplitudes. There was no main effect of group, F(2, 168) = 0.35, p = .70 or deviant condition, F(2, 168) = .71, p = .49 on the MMN amplitudes, nor was there an interaction between the magnitude of the deviance and subject group, F(4, 168) = 1.89, p = .12. MMN Latency For both musicians and nonmusicians, as the magnitude of the frequency deviance increased, the response latency beca me shorter as shown in Figure 7.
137 0 50 100 150 200 250 300 Controls MusiciansLatency (ms) Dev1 Dev2 Dev3 Figure 7. MMN Response Latency by Group and Deviant Condition. Deviant magnitudes: Deviant 1 = 1.5%, Deviant 2 = 3%, Deviant 3 = 6%. Overall, the neural responses of the musi cians occurred earlier than the controls. An ANOVA was conducted on the individual peak latencies to evaluate for an effect of group or deviant stimulus condition on the MM N latency. There was a main effect for group, F(2, 168) = 4.93, p = .008, as well as deviant condition, F(2, 168) = 35.38, p < .001. There was no significant interaction between group membership and deviant magnitude, F(4, 168) = 0.24, p = .91. Post-hoc analysis for group-wise co mparisons (Tukey HSD) confirmed that differences in MMN latency were significant between the vocalists a nd the controls. The
138 vocal musicians auditory neural responses to frequency differences occurred earlier than the nonmusicians for all deviant magnitudes rega rdless of the size (p = .009). However, the differences in response la tency between the instrumentalists and the nonmusicians did not quite meet significance (p = .06). This implies that the auditory neural responses of the vocal musicians occurred earlier than th e instrumental musicians; yet, the MMN latency difference between the two subclasses of musicians did not meet significance (p = .73). Post-hoc testing identified significant differences between all three deviant conditions (p < .05) and confirmed that the larg est frequency deviance elicited the earliest auditory neural MMN response for musician s and nonmusicians. Conversely, as the magnitude of the frequency difference decrease d, latency of response increased; that is, auditory neural response occurred later. Correlation Analysis with Electrophysiological Variables Psychoacoustic and Electrophysiological Variables Overall correlations between psychoac oustic (i.e., difference limen for frequency and pitch production accuracy) and electrophys iological variables (i.e., MMN amplitude and latency) were not supported by the evidence In addition, indi vidual DLF measures by subject group for the stimulus condition G4/392 Hz were compared to the corresponding MMN latency and amplitude data for 386.21 Hz (Deviant #1), 380.58 Hz (Deviant #2), and 369.81 Hz (Deviant #3). No significant correla tions were found among the variables for any of the subject groups Individual averages of pitch production
139 accuracy (PPA) were compared to corre sponding measures of MMN latency and amplitude for each deviant condition by subjec t group. Again, sufficient evidence to support significant correlations among the variables for any subject group was not observed. As a point of interest, it was questio ned whether those vocal musicians who had additional instrumental traini ng (9 of 19) would perform di fferently than those who did not have this additional musical expe rience. A comparison of means for electrophysiological and psychoacoustic variab les using a series of two-tailed t-tests revealed no significant differences between t hose singers with instrumental training and those without [Dev3latency: t(17) = 0.34, p = .74; Dev3amplitude: t(17) = 0.6, p = .56; difference limen frequency: t(17) = 1.2, p = .25; pitch production accu racy: t(17) = 0.93, p = .37]. Years of Music Training and El ectrophysiological Variables On average, the musicians had 10.5 years of music training ranging from 6 to 19 years. The Pearson product moment correl ation coefficient was used to determine if there was any relationship between the total number of years of music training received by the musicians and the elec trophysiological variables. Th e total number of years of music training for each individual musician were compared to the corresponding MMN latency and amplitude data for responses to 386.21 Hz (Deviant #1), 380.58 Hz (Deviant #2), and 369.81 Hz (Deviant #3). Evid ence was lacking to support significant correlations between response amplitude and year s of training. However, for latency of
140 response, there was a signifi cant negative correlation for Deviant #3, (r = -0.34, p = .03) and Deviant #1, (r = -0.31, p = .05), but no corr elation for Deviant #2, (r = -0.08, p = .62). For the smallest and largest deviances in pitch magnitude, as the number of years of music training increased, the response latenc y decreased. Thus, th e number of years of music training may have influenced the timi ng of the pre-attentive auditory neural response to pitch deviance. Age Training Initiated and Electrophysiological Variables The average age at which musicians in this study began music training was 8.5 years. More than half of the musicians (21 of 40) be gan music training by age 9 or younger. Using the Pearson product moment correlation coefficient, the age at which training began for each musician was comp ared to the corresponding MMN latency and amplitude data for responses to 386.21 Hz (Deviant #1), 380.58 Hz (Deviant #2), and 369.81 Hz (Deviant #3). The evidence did not support any signifi cant correlations between the age at which music training be gan and the electrophys iological variables (MMN amplitude or latency). As a point of interest, the musician group was divided between those who began training at age 9 or earlier, and those w ho began music training after age 9. No significant differences were found between the musician groups for PPA, t(39) = 0.49, p = .62; DLF, t(39) = 1.13, p = .26; or MMN latency, t(39) = 0.92, p = .36.
141 Summary of Findings 1. The ability to vocally match a single pitc h to a reference pitc h was more accurate for musicians than nonmusicians; however, for the particular vocal task in this study, mean pitch production accuracy did not differ between vocal musicians and instrumental musicians. 2. Pitch production accuracy across all fre quencies was most consistent for the vocalists. 3. Difference limens for frequency were sm aller for musicians than nonmusicians; while DLFs for vocalists and instrumentalists were similar. 4. The musicians were superior to th e nonmusicians on both psychoacoustic variables; however, (a) the nonmusicians had better auditory pitch discrimination than vocal pitch matching ability, (b) th e instrumental musicians demonstrated equal ability of the two skills, and (c) the vocal musicians were more accurate at vocal pitch matching than a uditory pitch discrimination. 5. When all individual data were combin ed, there appeared to be a positive correlation between rel DLF and rel PPA. Closer inspection of the data indicated a relationship between these variables only for the instrumentalists. There was minimal performance variability among th e vocal musicians and only a tendency towards a relationship for the nonmusicians.
142 6. All participants demonstrated auditory neural sensory perception to harmonic complexes. Neural responses were different for the musicians compared to the nonmusicians as early as 50 ms after pr esentation of a harmonic stimulus. 7. All groups demonstrated preattentive auditory neural responses to three pitch deviances (1.5%, 3%, and 6%). Interestingly, nonmusicians responded preattentively to pitch deviances as sma ll as 1.5%; even though on the behavioral auditory pitch discrimination task, their just noticeable difference was two times greater (3.19%). 8. Amplitude differences in event-related pot entials (ERPs) did not differentiate the musicians from the nonmusicians. For the mu sicians, the strength of the auditory neural response did not depend on the magnit ude of pitch deviation; that is, there was no predictable response pattern. On the other hand, the control group tended to have stronger responses to larger pitch changes. 9. For all groups, as the magnitude of pitch deviance increased, pre-attentive auditory neural response to the p itch change occurred earlier. 10. Overall, auditory change detection wa s faster for the musicians than the nonmusicians. 11. The vocal musicians responded faster to the pitch changes than the nonmusicians, while response latency did not significantly differ between the instrumentalists and the nonmusicians. Latency was shorter for the vocalists than the instrumentalists; however, the difference was not significant.
143 12. No relationships were found between ps ychoacoustic variables (DLF and PPA) and electrophysiological vari ables (MMN amplitude and latency) for any of the groups. 13. The number of years of music training app ears to influence pre-attentive auditory neural responses. Those musicians who had more year s of music training tended to respond faster to pitch deviances. There were no relationships between the number of years of music training and measures of DLF or PPA. 14. No relationships were found between th e age that music training began and psychoacoustic or electrophysiological variables.
144 Chapter Five Discussion This study was a beginning step of inqui ry into the effects of intensive music training on the auditory neural processe s of musicians and subsequently, on the relationship between audito ry perception and vocal pr oduction. Reliable evidence suggests that instrumental musicians experien ce changes in the auditory system following skill acquisition and sensory stimulation and ha ve superior auditory pitch discrimination and vocal pitch production compared to nonmus icians; yet little is known about neural changes in the auditory system in formally trained vocal musicians. Auditory pitch perception and laryngeal control are consid ered essential skills for accurate pitch production; however, the relationship be tween neurophysiological processes and perception-production abilities is unclear. Electrophysio logic and psychoacoustic measures were used to examine relationshi ps between pitch production accuracy, active pitch discrimination, and pre-attentive pitc h discrimination between two genres of musicians (vocalists and instrumentalists ) and a musically untrained control group.
145 Discussion of Findings in Relations hip to the Research Questions Vocal Pitch Production Accuracy The first hypothesis predicted a differe nce of vocal pitch matching accuracy between musicians and nonmusicians and betwee n the instrumental and vocal musicians. Although the exact relationship be tween auditory feedback and laryngeal control is yet to be determined, evidence indicates that accurate pitch production is influenced by auditory monitoring and proprioceptive feedback of the laryngeal system (Amir et al., 2003; Campisi et al., 2005; Kirchner & Wyke, 1965; Leydon et al., 2003; Mrbe et al., 2004; Ward & Burns, 1978; Wyke, 1974; Titze, 1994 ). Amir et al. (2003) reported that instrumental musicians with no previous vocal training, vocally matched pitch approximately three times more accurately than nonmusicians (i.e., relative accuracy based on grand means was 2.88% and 8.94%, re spectively). Results of this study reinforce the implication that musicians who have superior auditory pitch perception may also have enhanced pitch production abilities In the present st udy, the overall target pitch production was six times more accurate for the musicians than the nonmusician controls (i.e., rel PPA of the grand means was 1.28% and 7.83%, respectively). However, it should be remembered that one-half of the musicians were formally trained singers. When the vocal musician data was removed, th e instrumental musicians were still five times more accurate (i.e., mean rel PPA = 1.5%) than the nonmusicians. In both studies, musically untrained participants had mean pitch production errors greater than one semitone.
146 It is suggested that music training nurtur es a musicians sensitivity to acoustic parameters (e.g., pitch, rhythm, loudness). During music instruction, whether for an instrument or voice, the accuracy between a re ference pitch and the actual produced pitch is developed and proprioceptive memory for pi tch is enhanced by continued training and practice (Mrbe et al., 2004) These experiences build and strengthen a sensory perception-motor production re lationship that may explain why instrumental musicians not only have exceptional auditory percepti on skills, but also gr eater pitch production accuracy than nonmusicians. It was also questioned whether a differe nce of vocal pitch production accuracy (PPA) existed between instrume ntal and vocal musicians. Research suggests that professional vocal training shar pens proprioceptive reflexes of laryngeal joints, refines neuromuscular control of laryngeal and resp iratory muscles (Ward & Burns, 1978; Wyke, 1974), and improves neuromuscular memory for pitch control and accuracy (DiCarlo, 1994; Mrbe et al., 2004, 2002). For these reas ons, it was anticipated that the formally trained singers would match vocal pitch to a reference pitch more accurately than the instrumental musicians. Surprisingly, this hypothesis was not supported by the data; that is, average rel PPA% did not significantly differ be tween the two groups of musicians (i.e., 1.5% for the instrumentalists and 1.05% for the vocalists). Based on the grand means, both groups produced pitch within 1/ 8 of the reference pitch. Although average rel PPA% was comparable, the distribution of the musicians responses differed significantly. It is re levant to note that within the vo cal musician group there was very minimal inter-subject variability. Nearly ever y vocalist produced each pitch within 0% to
147 3% of the reference; while re sponses of the inst rumental musicians varied between 0% and 18% of the reference. The precision a nd consistency of performance demonstrated by the trained singers is believe d to be a reflection of their vocal training. A plausible explanation for the lack of dis tinction between the vocalists and the instrumentalists is the simplicity of the vocal task. The single pi tch imitation task was designed to determine whether a difference existed between musician s and nonmusicians even for a very basic pitch imitation task. The harmonic tone co mplexes were purposefully chosen to be within the mid-female vocal range. Imitation was performed for a single tone stimulus at a pace determined by the participant. While even this task appeared to be challenging for some of the nonmusicians, it was simplistic for the musicians, especially for the vocalists. It is likely that a more cha llenging pitch production task (e .g., sequence of tones, variable pitch range, or faster stimulu s presentation) would delineat e differences between the two musician groups. The pronounced uniformity of the vocalis ts responses implies that intense vocal training has a pos itive affect on laryngeal cont rol such that pitch production is not only precise, but consistent. Active Auditory Pitch Discrimination The second hypothesis proposed that musi cians and nonmusicians would differ in terms of active auditory frequency discrimina tion ability (DLF). Furthermore, it was questioned whether a difference in this abil ity exists between instrumental and vocal musicians. Consensus among published research has suggested that formally trained musicians have smaller frequency discrimina tion thresholds (DLFs) than nonmusicians
148 (Kishon-Rabin et al., 2001; Spiegel & Wa tson, 1984). Whether pitch discrimination differs between different genres of musicians is uncertain. Auditory pitch discrimination of harmoni c complexes (musical tones) was more precise for the musicians than the nonmusicians. On average, the musicians just noticeable difference of pitch change was less than one-half that of the control group (i.e., 1.35% compared to 3.19%). These results are compatible with the approximate 2:1 ratio between nonmusicians and musi cians found by Kishon-Rabin et al. (2001) and Spiegel and Watson (1984). It is reasonable to su rmise that intense practice and musical instruction hones a musicians categorical perc eption of pitch. As discussed in Chapter 2, the Western musical scale is based on interval s of tones and semitones. A change in one semitone is perceived by musicians as an alteration in the musical melody. For the musicians in this study, whose jnd averaged 1.3 5%, a pitch change of plus or minus onequarter of a semitone (i.e., 1.5%) crosses a musical boundary and signals a significant change similar to crossing a critical bandwid th in the cochlea. As a group, the formally trained musicians clearly demonstrated s uperior pitch discrimi nation and on average detected pitch changes within one-eighth of a difference in tone frequencies. The rel DLF values for musicians reported by Spiegel and Watson (1984) and Kishon-Rabin et al. (2001) were much smaller than those in this study (e.g., approximately 0.01 for instrumental musicians in both studies). This difference may be explained by choices of stimuli and paradi gm procedure. Although Spiegel and Watson (1984) used both pure tones and complex (squa re wave) sounds, component frequencies were above 440 Hz and presented in a 71% 2I /2AFC design; while Kishon-Rabin et al.,
149 (2001) used a 71% 3I/3AFC paradigm and s timuli consisted of pur e tones at 250 Hz, 1k Hz and 2k Hz. In the present study, stim uli were harmonic tone complexes with fundamental frequencies within the mid-fe male vocal range between 261.63 and 392 Hz. DLF was established by a 79% 3I/3AFC paradi gm. Each of the selected conditions (harmonic complexes, low fundamental fre quencies, and the 79% three-alternative forced-choice design) has been shown to el icit larger DLFs in the normal population compared to pure tone stimuli at frequencies between 500 and 2k Hz and a 2I/2AFC paradigm (Leek, 2001; Moore, 1989; M oore & Peters, 1992). Consequently, a combination of all three of these conditions is expected to elicit higher DLFs than the previous studies. Interestingly, KishonRabin and colleagues reported that for all subjects the largest rel DLF% occurred for 250 Hz. Examina tion of their data reveals that if the DLFs of the 3 runs for the 250 Hz condition are averaged together, the rel DLF% for the musicians and nonmusicians are remarkab ly similar to those in the present study. Namely, the average rel DLF% would be 1.42% for the musicians and 2.86% for the nonmusicians, compared to 1.35% and 3.19% respectively in the present study. It has been suggested that auditory skills may differ between musicians of distinct musical genres (Kishon-Rabin et al., 2001; Spiegel & Watson, 1984; Tervaniemi, Castaneda, Knoll, & Uther, 2006). Spiege l and Watson (1984) reported that musicians who tuned their own instrument (e.g., bra ss, string, wind instruments) had smaller discrimination thresholds, half the size of t hose who did not tune their own instrument (e.g., piano). These findings were not replicat ed by Kishon-Rabin et al. (2001); however, their data indicated that musicians of classi cal training had significantly smaller threshold
150 estimates than those with a contemporary ba ckground. Most recently, Tervaniemi et al., (2006) provided evidence that musicians se lectively encode acoustic parameters most relevant to their musical genre. Since precise pitch discrimination is crucial for a singers performance and because a singers instrument is endogenous to the body (i.e., the larynx), it was speculated that the just noticeable differe nce (jnd) may be even smaller for vocal musicians than for instrumentalists. Th is hypothesis was not supported by the data; difference limens for frequency were almost id entical (i.e., 1.3% for the singers and 1.4% for the instrumentalists). While the evid ence did not support a difference between these two musical genres on this partic ular psychoacoustic task, it is the first study to show that formally trained vocal musicians, like inst rumental musicians, have superior pitch discrimination abilities. The musicians superior and comparable auditory discrimination skills for harmonic tone complexes may be due to the instruction received by student musicians enrolled in the School of Music at the Univer sity of South Florida. Musicians receive similar instruction in aural theory including lessons in musical syntax, sight-reading, and pitch perception. Moreover, it is well docum ented that auditory pitch perception is a fundamental capacity in musical talent (S eashore, 1919, p. 42 as cited in Pedersen & Pedersen) and a vital skill for all musicians. Whether this exceptional pitch perception is an inherent advantage, a skill developed from years of intense training or a combination of these two factors, is yet unknown.
151 Present findings indicate that formally trained musicians, even those whose have never participated in vocal training, have superior auditory perception ability and enhanced pitch production accuracy (Table 4) This finding suggests that instrumental music training influences the integration of the bodys motor and sensory systems. This dual effect provides further evidence to suppor t the interaction betw een the auditory and laryngeal systems. If musi c training facilitates sensor y perception and motor production, then perhaps similar training techniques may be incorporated into treatment strategies for populations facing sensory perception and mo tor production challenges, including those with dyslexia or hearing impairment and/or pers ons who are neurologically impaired. Table 4. Rel DLF% and Rel PPA% Group Comparison Group N rel % Semitone Control DLF 21 3.19% 0.53 PPA 21 7.83% 1.30 Musician DLF 40 1.35% 0.23 PPA 40 1.28% 0.21 Instrumental DLF 21 1.40% 0.23 PPA 21 1.50% 0.25 Vocal DLF 19 1.30% 0.22 PPA 19 1.05% 0.17
152 Electrophysiological Measures of Pre-At tentive Auditory Pitch Discrimination This study proposed an electrophysiol ogical investigati on to examine preattentive sensory processing of auditory pitch stimuli by formally trained musicians. It was anticipated that small changes in pitch w ould be detected prior to the participants attention to the auditory stim uli and that these pre-attentive auditory neural responses would be faster and stronger for musici ans than nonmusicians. Moreover, it was hypothesized that auditory neur al responses may differ betw een instrumental and vocal musicians. Measures of latency and amplit ude to deviances in harmonic tone complexes were examined for the following ERP compone nts: P1-N1-P2 complex and mismatch negativity (MMN). Sensory perception P1-N1-P2 complex. As described in Chapters 2 and 4, the P1N1-P2 components are sensitive to physical ch anges in auditory stimuli. The complex occurs approximately 50 200 ms after the onset of an auditory stimulus and is interpreted as an indicator of preferential attention (Key et al., 2005). For the purposes of this study, P1-N1-P2 was examined to establish that all participants had comparable basic pre-attentive sensory percepti on, similar to verifying heari ng acuity prior to behavioral testing. No significant differences at this le vel of auditory processing were expected. The P1-N1-P2 response latencies and amplit udes for harmonic tone complexes did not appear to differ between the two subclasses of musicians, nor did the P1-N1-P2 latencies or the N1-P2 amplitudes appear to differ between the musicians and nonmusicians. Surprisingly, the nonmusicians had larger P1 amplitudes for each harmonic complex.
153 The P1 is associated with sensory gati ng which has been described as the brains ability to modulate its sens itivity to incoming irrelevant sensory stimuli (Boutros, Torello, Barker, Tueting & Wu, 1995; Braff & Geyer, 1990). The P1 amplitude may be interpreted as a neurophysiologi cal indicator of pr eferential attention to sensory input (Key et al., 2005). It is cons idered to reflect differences in regulating excitatory and inhibitory processes; that is, gating in or gating out auditory information. Failure of this process is thought to be a possible underlying contribut or to the development of psychotic states, such as schiz ophrenia (Boutros et al., 1995). An increase in the P1 amplitude has been shown to reflect pre-attentive recognition of novel stimuli in the normal population (Boutros et al., 1995). As a consequence of a musicians training, the central auditory system is familiar with musical tones and thus, probably less sensitive to the presentation of harmonic complexes. Moreover, during the electroencephalographic re cording, all participants were instructed to watch a closed captioned video and ignor e the presence of sound. By the inherent nature of their training, musicians may be be tter prepared to ignore or to attend to competing musical stimuli. It is possible that music training influences the brains sensory gating mechanism. Ther efore, it is reasonable to ex pect that musicians may have smaller P1 amplitudes for familiar musical stimuli than nonmusicians. Findings of this study are the first to suggest that musical expertise influences neural responses as early as 50 ms after the onset of a mu sically relevant stimulus. The P1-N1-P2 complex has not been specifically examined in adult musicians and warrants further research.
154 Mismatch negativity (MMN). The MMN is a neurophysiological response that reflects access to a neural memory of an acoustic parameter (Nousak, Deacon, Ritter, & Vaughan, 1996) and provides a neurological i ndex of individual discrimination ability (Ntnen, 1995; Ntnen, Pakarinen, Rinne, & Takegata, 2004). It may be described in terms of latency (time of occu rrence after stimulus onset) a nd amplitude (height of epoch representing strength of re sponse). Previous EEG evid ence indicates superior preattentive auditory processing abilities for mu sicians and suggests that musical expertise influences pitch processing by refining th e neural frequency-processing network (Koelsch et al., 1999; Pantev et al., 2001; Schn et al., 2004; Shahin et al., 2003; Tervaniemi, Just, Koelsch, Widmann, & Schrger, 2005; van Zuijen et al., 2004). Because of the extensive music training and practice received by musicians, the present study predicted that musicians would have la rger amplitudes and earlier latencies than nonmusicians for detection of small pitch de viances approaching behavioral DLFs. It was also questioned whether these neur ophysiological responses differed between instrumental and vocal musicians. MMN amplitude. Auditory neural responses re flecting sensory memory for pitch were anticipated to be stronger for musician s than nonmusicians. Surprisingly, amplitude values for pre-attentive auditory neural respon ses to changes in pitch did not significantly differ between controls and musicians. Wh ile all groups responded with an amplitude mismatch to the deviant stim uli, the MMN amplitude values did not differentiate the musicians from the nonmusicians nor did they va ry between the subcla sses of musicians. There was a non-significant trend for the control group to have stronger sensory
155 responses as the magnitude of deviance increas ed; that is, as the pitch difference became larger, so did the response amplitude. There was no such pattern or tendency for the musicians. Musicians have been shown to have larger MMN amplitudes in response to multidimensional deviances, such as harmonically inappropriate chords (Koelsch et al., 2002), pitch change within a familiar scale (Bra ttico et al., 2001), and note change within a complex melody (Lopez et al., 2003). Koelsch et al. (1999) suggested that the auditory sensory memory traces of musicians contain more acoustic parameter information than the memory traces of musically untrained in dividuals. This implies that musicians may require less neural effort to extract certain acoustic informa tion. Given this perspective, perhaps the auditory discrimi nation task in the present st udy was not complex enough to distinguish the MMN amplitude responses betw een the musicians and the nonmusicians. This explanation concurs with similar fi ndings reported by Tervaniemi and colleagues (2005) who also questioned whether a similar auditory perceptual task was too easy for musicians. P1 amplitude findings in the present study provide further evidence for traininginduced changes in auditory neural processing. Musicians had smaller P1 amplitudes in response to harmonic complexes compared to nonmusicians implying that the brains familiarity with musical stimuli moderated th e reaction of the sensory gating mechanism. Current evidence reinforces the suggestion that musicians have superior trainingenhanced sensory memory representations fo r acoustic parameters of harmonic stimuli.
156 Thus, as a consequence of their training, perhaps less neural energy is required for musicians to process simple acoustic para meters of musically relevant stimuli. MMN latency. Electrophysiological evidence in dicates that music training modifies neural processing of acoustic input revealing that instrume ntal musicians have faster neural responses for pitch changes than nonmusicians (Koelsch et al., 2002; Koelsch et al., 1999; Shahin et al., 2003; Tervaniemi et al., 2005) It was anticipated that all subject groups would have a pre-attentiv e response to changes in pitch deviance approaching behavioral DLFs. It was furt her hypothesized that musicians would have earlier latencies than nonmusicians and that within the musician group a difference would exist between instrumental and vocal musicians. Pre-attentive auditory neural responses to changes in pitch frequency were present for all groups and for all three conditions of pitch deviance (1.5%, 3%, and 6%). As the frequency difference between the deviant stim ulus and the standard stimulus increased, the auditory neural response to the pitch ch ange occurred faster. For example, the 6% pitch deviance elicited the earliest MMN re sponse for all groups, followed by the 3% change and then the 1.5% pitch change. As predicted, the musicians responded to all pitch changes faster than those without musi c training and those musicians who had more years of music training tended to respond fa stest. These findings provide further evidence that fundamental auditory proces sing abilities can be f acilitated by music training. Closer inspection of the two musician groups noted that only the vocalists responded earlier to pitch changes than the co ntrol group; the instrumentalists did not.
157 This suggests that latency was shortest for the vocalists; however, th is inference was not supported by the analysis. Although MMN latency for detection of pitch deviance occurred earlier for the vocal musicians, the difference did not reach significance. Nevertheless, ERP evidence indicates that form ally trained vocal musicians, similar to instrumental musicians, have superior preattentive neural frequency processing. Electrophysiological data from the pres ent study clearly reinforce the philosophy that MMN amplitude and latency reflect two independent factors influencing the mismatch response and should be measured a nd analyzed separately (Lang et al., 1995; Ntnen, 1992). Moreover, findings support th eories that music tr aining and experience facilitate modification of neural processing and enhance sensory me mory representations of acoustic parameters. The exact effect of music training an d expertise on the neural frequencyprocessing network remains unknown. Intens e music training has been shown to stimulate microstructural neurological change s (Pantev et al., 2001; Pascual-Leone et al., 1995). Considering the mechanisms by which mi crostructural plasticity occurs (Calford, 2002), perhaps rigorous musical practice amplifies the comm unication between neurons by strengthening the synapses and thus increasi ng the efficiency of neural transmission at a cellular level. This idea is compatible with the propos al that acoustic training enhances the tuning processes of neurons in the auditory cortex (Shahi n et al., 2003). Taking this supposition one step further, perhaps modifica tion of the frequency-tuning process affects categorical perception. As described in Chapter 2, pitch differences in music are perceived categorically (Sundberg, 1994). Slight variations in fre quency have no effect
158 on the classification of a pitc h or tone; however, at the bo rder of a frequency range, a minor shift radically changes the perception from one category to another. It is speculated that formal music training may sharpen the borde rs of categorical pitch perception so that slight changes in pitch are detected auto matically with greater precision prior to a cognitive decision. This is a plausible explan ation for the musicians earlier reactions to pitch deviances and supports the argument for indepe ndent processes underlying amplitude and latency responses. Relationships between Auditory Perception and Pitch Production Across Groups Auditory pitch discrimination (perception) and vocal pitch control (production) have been identified as related abilities and essential skills for musicians. Based on a review of previous investiga tions and on theories of neural plasticity, it was hypothesized that a positive correlation exists between pe rception and production abilities for musical stimuli and that this correlation would be st ronger for the formally trained musicians and strongest for the formally trained vocal musicians. Specifically, relationships between pitch production accuracy, active auditory pitch discrimination, and pre-attentive auditory neural responses were examined among forma lly trained musicians and nonmusicians as well as between vocal and instrumental musicians. Correlations between DLF and PPA. A positive relationship between perception and production was expected. At first glance, it appears that this hypothesis is true. A combination of all individual data yielded a significant positive correlation between auditory discrimination and vo cal pitch production accuracy (rs = 0.61, p < .001). These
159 results are similar to those of Amir et al. (2003) who reported a positive relationship between vocal production and a uditory perception when in strumental musician and nonmusician data was merged (r = 0.67). However, in the present study, when the relationship between perception and production is viewed separately for each group, the outcome is not as clear and points to differences between the two musical genres. Neither the control group nor the vocal musicians demonstrated a significant relationship between perception and producti on. Perception and production abilities varied greatly within the cont rol group. By severe contrast the vocalists performed both tasks accurately with minimal response variab ility reflecting their tr aining and expertise in both areas. Only the instrumental musicians had a positive correlation between auditory discrimination and pitch production ac curacy, reinforcing the implication that instrumental music instruction facilita tes a sensory perception-motor production relationship. Correlations between psychoacoustic and electrophysiological variables. Others have investigated relationships between ne ural and behavioral responses to pitch deviances by comparing MMN amplitude and la tency to hit rate (HR) and reaction time (RT) for auditory discrimination between two frequencies (Novitski et al., 2004; Lang et al., 1995). There tends to be a positive co rrelation between HR and MMN amplitude; however, correlations between HR, RT, and MMN latency are inconsistent (Novitski et al., 2004; Lang et al., 1995). Evidence from the present study indicat es that formal music training affects psychoacoustic abilities (DLF and PPA) as well as electrophysiol ogical responses (P1
160 amplitude, MMN latency); yet, evidence did no t support significant re lationships between these measures for any subject group. Difference limens for frequency were determined rather than response speed (RT) or accuracy (HR) for pitch discrimination. Since the behavioral variables were not the same as those measured in the prior studies, the discrepancy between correlati on outcomes may be due to ta sk measurement differences. Moreover, the variables in this study may not correlate simply because they represent different processes. ERPs reflect automa tic neurophysiological memory-based sensory responses to change detection. By contra st, difference limen for frequency and pitch production accuracy are measures of active behavioral choi ces influenced by attention and short-term memory in addition to s ubjective motivation and cooperation. Thus, the electrophysiologic variables may represent on ly a subset of the processes underlying behavioral discrimination and production of frequency. Summary of relationships among groups. Musicians had faster neural responses to pitch deviances and demonstrated superior active (attentive) auditory discrimination and vocal pitch production compared to nonmus icians. A significant correlation between perception and production abilities was appare nt for only the instrumental musicians. Sufficient evidence to suppor t relationships between th e electrophysiological and psychoacoustic variab les was not observed. Relationships between Auditory Perc eption and Pitch Production Within Groups The final hypothesis questioned relations hips between perception and production variables within each group. It was anticipated that within a subject group a pattern of
161 abilities may exist between pitch production ac curacy, auditory pitc h discrimination and pre-attentive pitch discrimination. Co mparison of DLF and PPA. For those without music training, auditory discrimination was almost 3 times more accura te than pitch production. Whereas the just noticeable difference (jnd) between harmonic complexes was within of a semitone, vocal pitch production deviated almost 1 sem itones from the target pitch. Studies of singing development in children attribute this deviance of pitch production from discrimination to inadequate laryngeal muscle control, poor kinesthetic feedback from the larynx and/or delayed internal auditory mon itoring (Goetze et al., 1990). By comparison, the instrumental musicians discriminate d and vocally produced harmonic complexes equally well reinforcing the supposition that instrumental music tr aining facilitates the integration of the bodys motor and sensory systems. In contrast to the other two groups, vocal musicians were more accurate for pitch control than for pitch percep tion. Although their jnd was approximately one-quarter of a semitone, their PPA was within one-sixth of a semitone. This implies that the vocalists internally discriminated between a target and a produced pitch with greater precision than they distinguished between two externally presented harmonic tones. The present findings support previous research (DiCarl o, 1994; Kirchner & Wyke 1965; Mrbe et al., 2002) and concur that formally trained vocal musicians develop explicit sensory memory representations and enhanced laryngeal prop rioceptive reflexes secondary to traininginduced neural changes. Perhaps the li nk between auditory perception and pitch production is not exclusively between active pi tch discrimination and laryngeal control.
162 Rather, training also enhances a relationshi p between auditory sensory memories and laryngeal reflexes. Consensus among published research sugge sts that a degree of auditory pitch discrimination may serve as a pre-requisite to vocal pitch matching ab ility and that these two skills may be two indepe ndent abilities between which a relationship strengthens with training and development (Geringer, 1983; Goetze et al., 1990; Yarbrough et al., 1991). Present findings show that the rela tionship between PPA and DLF can occur on a continuum and agree that this relationship is influenced by music tr aining. Specifically, PPA was poorer than DLF for those with no music training; PPA wa s equal to DLF for those who received only instrumental tr aining; and PPA was better than DLF for vocalists who were specifically tr ained for pitch production accuracy. Comparison of pre-attentive and active pitch discrimination. It was not the purpose of this investigation to establish pre-attentive difference limens for frequency (i.e., the just noticeable difference between two frequencies). Pr evious MMN data has shown that musicians respond to pitch devian ces as small as 0.8% (Tervaniemi et al., 2005). The harmonic complexes chosen for th is electrophysiologic task were based on the best and poorest DLFs obtained from the psychoacoustic task. The smallest pitch deviation (1.5%) from the standard tone (392 Hz) was slightly above the mean frequency discrimination threshold obtained for the musi cians (1.35%). As expected, the musicians responded pre-attentively to the smallest pitch deviance. Surprisingly, the control group also res ponded pre-attentively to the 1.5% pitch deviance even though attentively, just noticeab le difference for pitch deviance was two
163 times greater ( rel DLF% = 3.19%). This discovery ha s several implicat ions. While the influence of genetic factors can not be dismi ssed, it may be argued that superior auditory discrimination is not an inherent ability, but one that may be shaped by specific training. Electrophysiological evidence suggests that mu sic training enhances the neural encoding of memory representations and facilitates retr ieval of these sensory memory traces. This implies that nonmusicians, who have not experi enced this explicit tr aining, may have less efficient access to neural memories for making attentive behavioral decisions. Furthermore, this discovery reinforces specula tion that automatic memory traces underlie subsequent attentive processe s (Sams et al., 1985; Tervaniemi, 2001; Tervaniemi et al., 2005) and lends further support to theories of training-induced cort ical plasticity. Summary of within group relationships. It appears that those with no music training have better pre-attentive neural pitch discrimination than active (attentive) discrimination and more accurate auditory pitch discrimination than vocal pitch production. The vocal musicians had superi or pitch production skill compared to auditory pitch discrimination, while the instru mental musicians demonstrated equivalent abilities for all three tasks. Formal music training appears to facilitate underlying auditory neural processes that in turn in fluence attentive auditory discrimination and laryngeal control.
164 Limitations of the Study The research design and methods of the present study were successful in addressing the research questions However, while every effort was made to control for validity and reliability, the following limitations are considered: 1. The sample population of musicians volunt eered from the Univ ersity of South Floridas School of Music and was not ra ndomly selected. Subjects had a mean age of 22 and an average of 10.5 years of music training. Care should be given when generalizing results to all musicians. The type of music instruction and the extent of training, practice, and perfor mance vary greatly among musicians and any one of these factors may influence task performance. 2. There are no formally established procedur es or standards of protocol for eliciting and/or analyzing auditory evoked poten tials (AEPs) of mi smatch negativity (MMN) of the human brain. Methods for procedures and analysis were based on extensive review of curr ent published literature and personal consultation with leading experts in the field. Caution s hould be exercised when comparing results between studies since protocols may vary. 3. While the pitch perception and producti on tasks designed for this study were appropriate for nonmusicians, they may have been too simplistic to delineate significant differences between the vocal and instrumental musicians. Thus, performance measures may not be the best representation of the musicians pitch perception or production abilities.
165 4. While present findings indicate that musici ans have superior pitch perception and production skills secondary to training-i nduced neural changes in auditory processes, it is not possible to rule out an influence of genetic coding and innate abilities. The musicians in this study reported a much higher incidence of musicians in the immediate family (13 of 40) than the nonmusicians (1 of 21). Directions for Future Research Previous neurophysiological research ha s identified anatomical and physiological differences between musicians and nonmusi cians. Electrophys iological evidence, including the present st udy, strengthens the pr emise that auditory neural changes occur following skill acquisition and sensory stimulation. This study was the first to explore and compare the effects of long-term music training on variables of pitch perception and production between vocal and inst rumental musicians. Furthe r research is warranted to differentiate the effects of training on ne ural processes between these two groups. Findings suggest that more complex per ception and production tasks may tease out distinctive abilities. The mismatch negativity (MMN) br ain response provides an objective noninvasive index of auditory discrimination and an excellent means to study training effects on auditory neural plas ticity. Because of their rigor ous training, musicians are an exceptional population in which to examine the influence of expertise on acoustic parameters of perception such as frequenc y, duration, intensity, timbre, and rhythm. Moreover, there are many diverse genres of musicians with which to compare the
166 influence of training and practice on the br ain. A longitudinal study or comparative analysis of young and old musicians may also provide distinctiv e information on the effects of aging on their superior per ception and production skills, not to mention differentiating age of onset of music traini ng versus number of years of training which necessarily covary in a population of uniform age. Native singers of tone languages (e.g. Vietnamese and Mandarin) offer a unique popul ation in whom to study pitch production and auditory neural responses to pitch perception. Thei r language specific pitch perception and production skills may provide additional insight in to the debate over inherent abilities and genetic codi ng versus effects of training. Findings from the present study suggest that music training facilitates pitch perception and production rega rdless of musical genre. Moreover, published research concurs that musicians have s uperior pitch detection skills not only for music stimuli, but also for language suggesting that music trai ning enhances pitch pro cessing for both music and language (Schn, Magne, & Besson, 2004). Previous electrophysiologic (Friederici, Pfeifer, & Hahne, 1993) and neural imagi ng (Koelsch et al., 2000) studies report considerable overlap of neural structures a nd similar neural systems responsible for the integration of pitch processing of both musi c and language. The therapeutic effects of music training on pitch perception and producti on for clinical populations who present with impairments of pitch perception and production, such as those with cochlear implants, dyslexia, and/or Parkinsons disease, warrants further research. Another area of electrophysiologic resear ch that merits further investigation is cerebral lateralization of activity. Is ther e hemispheric dominance for the processing of
167 certain acoustic parameters and does this hemispheric activity differ by population, task or training? Much remains to be discovered about th e diverse effects of long-term training on electrophysiological responses. Th e present study identified an incidental effect of music training on the P1 AEP component. The P1 ampl itude, an index of the brains ability to modulate sensitivity to sensory input, diffe red between musicians and nonmusicians and warrants additional study. Visual inspection of ERP data also noted that the P3a response differed between the subject groups. The P3a component is thought to reflect a passive attention switch to stimuli and often follows an MMN response. Cursory examination of this component suggested a possible training e ffect and merits additional investigation. Conclusions Electrophysiologic and psychoacoustic m easures were used to examine preattentive and active pitch disc rimination as well as pitch production accuracy between nonmusicians, formally trained instrumental musicians, and formally trained vocal musicians. This study was a beginning step of inquiry to compare the effects of longterm music training on the auditory neural function of nonmusicians compared to musicians from two discrete musi cal genres. The overall objec tive was to take an initial step to contribute to the body of basic rese arch regarding the pe rception and production abilities of formally trained vocal musicians. All musicians, regardless of specialt y, demonstrated superior auditory pitch perception (DLF) and vocal pitch production accu racy (PPA) compared to nonmusicians.
168 Vocal musicians and instrumental mu sicians performed equally well on the psychoacoustic tasks. Furthermore, pitch production accuracy across all frequencies was most consistent for the vocalists. Evidence supports the implicati on that music training facilitates both auditory per ception and vocal produc tion regardless of mu sic specialty. Findings suggest that audito ry pitch discrimination may se rve as a pre-requisite to pitch production accuracy. Namely, the data reflect that the rela tionship between PPA and DLF occurs on a continuum. Pitch pr oduction accuracy was poorer than auditory pitch discrimination for those with no previ ous music training. PPA was equal to DLF for those who had only instrumental music tr aining; while pitch production accuracy was superior to auditory pitch discrimination fo r vocal musicians, reflecting their specialty training. The two psychoacoustic variables were significantly correlated only for the instrumentalists. Vocalists demonstrated mi nimal inter-subject va riability so that a correlation was not detected. Electrophysiological evidence from th e present study indi cates that vocal musicians, like instrumental musicians, expe rience neural changes in the auditory system following skill acquisition and sensory training and demonstrate superior pre-attentive auditory discrimination. This st udy is the first to report an in fluence of musical expertise on auditory neural responses as early as 50 ms (P1) afte r onset of musical stimuli. MMN responses indicate that vocal musicians, as well as instrumental musician, have superior sensory memory representations for acoustic parameters of harmonic stimuli and imply that auditory neural changes are faci litated by long-term musi c training. Overall, auditory neural detection of pitch deviance wa s faster for musicians than nonmusicians.
169 In other words, the musicians recognized a change in the acoustic parameter of the musical stimuli sooner than the nonmusicians Based on P1 and MMN amplitude data, it is suggested that perhaps musicians require less neural energy to extract simple acoustic parameters of musically relevant stimuli. Interestingly, nonmusicians responded pre-a ttentively to pitch deviances that were their attentive DLF. This discovery supports the theory th at automatic memory traces may underlie subsequent attentive processes and are enhanced and facilitated by music training. The present findings reinforce the hypothesis that plasticity in the neuroanatomical system is reflected in neurol ogical change in the auditory system as a result of long-term music training. The exact relationship among physiological variables, perceptu al abilities, and pitch production remains elusive; however, auditory pitch perception and vocal pitch production appear to be independent abilities between which a relationship develops with training. Perhaps the elusive link is not to be found between attent ive cortical processes. Rather, a complex connection, as yet to be discovered, may lie within the neural substrates of auditory sensory memories and laryngeal reflexes.
170 List of References Adzaku, F., & Wyke, B. (1979). Innervation of the subglottic mucosa of the larynx, and its significance. Folia Phoniatrica 31, 271-283. Alho, K., Connolly, J., Cheour, M., Lehtokoski, A., Huotilainen, M., Virtanen, J., et al. (1998). Hemispheric lateralizat ion in pre-attentive pro cessing of speech sounds. Neuroscience Letters, 258, 9-12. American National Standards Institute. (1996). Specification for audiometers (ANSI S3.6-1996). New York: ANSI. Amir, O., Amir, N., & Kishon-Rabin, L. (2003). The effect of superior auditory skills on vocal accuracy. Journal of the Acoustical Society of America 133 (2), 1102-1108. Andersen, B., Korbo, L., & Pakkenberg, B. (1992). A quantitative study of the human cerebellum with unbiased st ereological techniques. Journal of Comprehensive Neurology 326, 519-560. Anderson, K., Brown, C., & Tallal, P. ( 1993). Developmental language disorders: Evidence for a basic processing deficit. Current Opinion in Neurology and Neurosurgery, 6, 98-106. Apfelstadt, H. (1984). Effects of melodic pe rception instruction on pitch discrimination and vocal accuracy of kindergarten children. Journal of Research in Music Education 32, 15-24. Ari-Evan Roth, D., Amir, O., Alaluf, L., Bu chsenspanner, S., & Kishon-Rabin, L. (2003). The effect of training on frequency disc rimination: generalization to untrained frequency and to the untrained ear. Journal of Basic and C linical Physiology and Pharmacology, 14 (2), 137-150. Ayotte, J., Peretz, I., & Hyde, K. (2002). Congenital amusia: A group study of adults afflicted with a music-specific disorder. Brain, 125, 238-251. Baken, R., & Noback, C. (1971). Neuromuscular spindles in intrinsi c muscles of a human larynx. Journal of Speech and Hearing Research 14, 513-518.
171 Barry, N. H. (n.d.). Nonparametric statistics. Retrieved May 10, 2006, from http://faculty-staff.ou.edu/B/Nancy.H.Barry-1/nonpar.html Bentley, A. (1966). Musical abilities in children. London: George G. Harrap and Co. Ltd. Bertoli, S., Smurzynski, J., & Probst, R. (2005) Effects of age, agerelated hearing loss, and contralateral cafeteria noise on the discrimination of small frequency changes: Psychoacoustic and electr ophysiological measures. Journal of the Association for Research in Otolaryngology 6, 207-222. Bhabu, P., Poletto, C., Mann, E., Bielamowicz, S., & Ludlow, C. (2003). Thyroarytenoid muscle responses to air pressure stimula tion of the laryngeal mucosa in humans. Annals of Otology, Rhinology, and Laryngology 112, 834-840. Boardman, E., (1964). An investigation of the effect of preschool training on the development of vocal accuracy in young children. (Doctoral dissertation, University of Illinois). Dissertation Abstracts International 25, 1245. Boone, D. (1996). Clinical relevance of c ontrolling chaos and complexity: Implications for the speech pathologist. In P. Davis & N. Fletcher (Eds.), Vocal fold physiology: Controlling complexity and chaos (pp. 347-358). San Diego: Singular Publishing Group, Inc. Boutros, N., Torello, M., Barker, B., Tue ting, P., & Wu, S. (1995). The P50 evoked potential component and mismatch detec tion in normal volunteers: Implications for the study of sensory gating. Psychiatry Research 57 (1), 83-88. Bosnyak, D., Eaton, R., & Roberts, L. (2002) Modification of di stributed auditory cortical representations in non-musicians by training for pitch discrimination with 40-Hz modulated tones. Proceedings of the 13th International Conference on Biomagnetism, Jena, Germany. Braff, D., & Geyer, M. (1990). Sens orimotor gating and schizophrenia. Archives of General Psychiatry 47, 181-188. Brandon, C., Rosen, C., Georgelis, G., Horton, M., Mooney, M., & Sciote, J. (1991). Staining of human thyroaryte noid muscle with myosin antibodies reveals some unique extrafusal fibers, but no muscle spindles. Journal of Voice 17, 245-254. Brattico, E., Ntnen, R., & Tervaniemi, M. (2001). Context effects on pitch perception in musicians and non-musicians: Evidence from ERP recordings. Music Perception, 19 (2), 199-222.
172 Buchthal, F. (1959). Electromyography of intrinsic laryngeal muscles. Quarterly Journal of Experimental Physiology 44, 137-148. Buonomano, D., & Merzenich, N. (1998). Cortic al plasticity: From synapses to maps. Annual Review of Neuroscience, 21, 149-186. Calford, M.B. (2002). Dynamic representa tional plasticity in sensory cortex. Neuroscience 111 (4), 709-738. Campisi, P., Low, A., Papsin, B., Mount, R., Cohen-Kerem, R., & Harrison, R. (2005). Acoustic analysis of the voice in pedi atric cochlear implant recipients: A longitudinal study. The Laryngoscope 115, 1046-1050. Crichton, N.J. (1999). Information point: Spearmans rank correlation. Journal of Clinical Nursing 8, 763. Dalla Bella, S., & Peretz, I. (1999). Music agnosias: sele ctive impairments of music recognition after brain damage. Journal of Neuro-Music Research, 28, 209-216. Darwin, C. (1871). The descent of man and selection in relation to sex London: John Murray. Davidson, I. (1959). Sidetone delay, r eading rate, articulation, and pitch. Journal of Speech and Hearing Research 2, 266-270. Davidson, L., McKernon, P., & Gardner, H. (1981). The acquisition of song: A developmental approach. Documentary Report of the Ann Arbor Symposium Reston, VA: Music Educators National Conference. Dejonckere, P. (1995). Fascinating and intri guing vibrato. In P. Dejonckere, M. Hirano, and J. Sundberg (Eds.), Vibrato London: Singular. Deutsch, D. (1992). Paradoxes of musical pitch. Scientific American, 267 (2), 88-95. Deutsch, D., Henthorn, T., Marvin, E., & Xu, H. (2004). Perfect pitch in tone language speakers carries over to music. Acoustical Society of America 148th Meeting Lay Language papers. Retrieved on May 4, 2005 from http://www.aip.org/148th/deutsc.html DiCarlo, N. (1994). Internal voice sensitivities in opera singers. Folia Phoniatrica et Logopaedica, 46, 79-85. Donchin, E., & Coles, M. (1988). Is the P 300 component a manifestation of context updating? Behavioral & Brain Sciences 11, 357-427.
173 Elbert, T., Candia, V., Altenmuller, E., Rau, H., Sterr, A., Rockst roh, B., et al. (1998). Alteration of digital representations in somatosensory cortex in focal hand dystonia. NeuroReport, 9, 3571-3575. Elbert, T., Pantev, C., Wienbruch, C., Ro ckstroh, B., & Taub, E. (1995). Increased cortical representation of the fingers of the left hand in string players. Science 13, 305-307. Elman, J. (1981). Effects of frequency-shifte d feedback on the pitc h of vocal production. Journal of the Acoustical Society of America 70, 45-50. Fink, B., & Demarest, R. (1978). Laryngeal biomechanics. Cambridge, MA: Harvard U.P. Fodor, J. (1983). The modularity of mind. Cambridge, MA: MIT Press. Fodor, J. (2001). The mind doesnt work that way. Cambridge, MA: MIT Press. Foxton, J., Dean, J., Gee, R., Peretz, I., & Gri ffiths, T. (2004). Characterization of deficits in pitch perception underlying tone deafness. Brain 127 (4), 801-810. Friesen, L., & Tremblay, K. (2003). Electrophys iological measures of hearing, speech and language. Perspectives on Neurophysiol ogy and Neurogenic Speech and Language Disorders, 13, 3-9. Gaab, N., & Schlaug, G. (2003). The effect of musicianship on pitch memory in performance matched groups. NeuroReport, 14 (18), 2281-2295. Gaser, C., & Schlaug, G. (2003). Brain st ructures differ between musicians and nonmusicians. The Journal of Neuroscience, 23 (27), 9240-9245. Gelfand, S. (1998). Hearing: An introduction to psychological and physiological acoustics. New York: Marcel Dekker, Inc. Geringer, J. (1983). The relationship of pitc h-matching and pitch discrimination abilities of pre-school and four th grade students. Journal of Research in Music Education 31 (2), 93-99. Glass, G., & Hopkins, K. (1996). Statistical methods in education and psychology (3rd Ed.). Needham Heights, MA: Allyn & Bacon.
174 Goetze, M., Cooper, W., & Brown, C. (1990). R ecent research in singing in the general music classroom. Bulletin of the Council of Re search in Music Education, 104, 16-37. Good, C., Johnsrude, I., Ashburner, J., Henson, R., Friston, K., & Frackowiak, R. (2001). Cerebral asymmetry and the effects of sex and handedness on grain structure: a voxel-based morphometric analysis of 465 normal adult human brains. NeuroImage, 14, 685-700. Gould, A. (1969). Developing specialized progr ams for singing in the elementary school. Bulletin of the Council of Re search in Music Education, 17, 9-22. Green, G. (1990). The effect of vocal modeli ng on pitch-matching accuracy of children in grades one through six. Journal of Research in Music Education 38, 225-231. Hain, T., Burnett, T., Kiran, S., Larson, C ., Singh, S., & Kenney, M. (2000). Instructing subjects to make a voluntary response re veals the presence of two components to the audio-vocal reflex. Experimental Brain Research 130, 133-141. Hauser, M., & McDermott, J. (2003). The evol ution of the music faculty: a comparative perspective. Nature Neuroscience, 6 (7), 663-668). Higgins, M., McCleary, E., Carney, A., & Sc hulte, L. (2003). Longitudinal changes in childrens speech and voice physiology after cochlear implantation. Ear & Hearing 24 (1), 48-70. Higgins, M., McCleary, E., Ide-Helvie, D., & Carney, A. (2005). Speech and voice physiology of children who are hard of hearing. Ear & Hearing 26 (6), 546-558. Hirano, M. (1981). Clinical examination of voice. New York: Springer-Verlag Wien. Hirano, M., & Sato, K. (1993). Histological color atlas of the human larynx. San Diego: Singular Publishing. Hollien, H. (1993). That golden voice talent or training? Journal of Voice 7, 195-205. Hood, L. (1998). An overview of neural func tion and feedback control in human communication. Journal of Communication Disorders 31, 461-470. Howle, M. (1992). An examination of sel ected aspects of pitch-matching problems among children. Research Perspectives in Music Education, 3, 22-27.
175 Hurme, P., Laukkanen, A., & Sonninen, A. ( 1999). The external frame function in the control of pitch, register, and singing mode: Radiographic observations of a female singer. Journal of Voice 13 (3), 319-340. Hurme, P., & Sonninen, A. (1998). Vocal fo ld strain and vocal pitch in singing: Radiographic observations of singers and non-singers. Journal of Voice 12 (3), 274-286. Jasper, H. (1958). The ten-twenty electrode system of the International Federation. Electroencephalography and Clinical Neurophysiology 10, 371-375. Jafari, M., Wong, K., Behbehani, K., & Kondraske, G. (1989). Performance characterization of human pitch cont rol system: An acoustic approach. Journal of the Acoustical Society of America 85 (3), 1322-1328. Jones, J., & Munhall, K. (2000). Perceptual calibration of F0 produc tion: Evidence from feedback perturbation. Journal of the Acoustical Society of America 108 (3), Pt. 1, 1246-1251. Jones, M. (1989). A study of vocal pitch-ma tching skills among undergraduate education majors using classroom instruments. Update: The Application of Research in Music Education 7 (2), 39-41. Johnsrude, I., Penhune, V., & Zatorre, R. ( 2000). Functional specificity in the right human auditory cortex for perceiving pitch direction. Brain, 123, 155-163. Joyner, D. (1969). The monotone problem. Journal of Research in Music Education 17 (1), 115-124. Jrgens, U. (2002). Neural path ways underlying vocal control. Neuroscience and Biobehavioral Reviews 26, 235-258. Kawahara, H., & Williams, J. (1996). Eff ects of auditory feedback on voice pitch trajectories: Characteristic responses to pitch perturbations. In P. Davis & N. Fletcher (Eds.), Vocal fold physiology: Cont rolling complexity and chaos (pp. 263-278). San Diego: Singular Publishing Group, Inc. Keene, J. (1961). Muscle spindles in human laryngeal muscles. Journal of Anatomy 95, 25-29. Key, A., Dove, G., & Maguire, M. (2005). Link ing brainwaves to the brain: An ERP primer. Developmental Neuropsychology, 27 (2), 183-215.
176 Kirchner, J., & Wyke, B. (1965). Articular reflex mechanisms in the larynx. Annuals of Otology, Rhinology and Laryngology, 74, 748-768. Kishon-Rabin, L., Amir, O., Vexler, Y., & Za ltz, Y. (2001). Pitch discrimination: Are professional musicians bette r than non-musicians? Journal of Basic and Clinical Physiology & Pharmacology, 12, 125-144. Klein, M., Coles, M., & Donchin, E. (1984). Pe ople with absolute pitch process tones without producing a P300. Science 223, 1306-1309. Koelsch, S., & Friederici, A. (2003). Toward th e neural basis of pro cessing structure in music. Annals of the New York Academy of Sciences, 999, 15-28. Koelsch, S., Gunter, T., Friederici, A., & Sc hrger, E. (2000). Brain indices of music processing: Non-musicians are musical. Journal of Cognitive Neuroscience 12, 520-541. Koelsch, S., Schmidt, B., & Kansok, J. (2002). Effects of musical e xpertise on the early right anterior negativity: An ev ent-related brain potential study. Psychophysiology, 39, 657-663. Koelsch, S., Schrger, E., & Tervaniemi, M. (1999). Superior preattentive auditory processing in musicians. NeuroReport 10, 1309-1313. Lane, H., & Tranel, B. (1971). The Lombard sign and the role of hearing in speech. Journal of Speech and Hearing Research 14, 677-709. Lang, A., Eerola, O., Korpilahti, P., Holopain en, I., Salo, S., Uusipaikka, E., & Aaltonen, O. (1995). Clinical applications of the mismatch negativity. Ear & Hearing 16, 117-129. Langereis, M., Bosman, A., van Olphen, & Smo rrenburg, G. (1998). Effect of cochlear implantation on voice fundamental frequenc y in post-lingually deafened adults. Audiology 37, 219-230. Larson, C. (1998). Cross modality influences in speech motor control: The use of pitchshifting for the study of F0 control. Journal of Communication Disorders 31, 489-503. Leek, M. (2001). Adaptive procedur es in psychophysical research. Perception & Psychophysics 63 (8), 1279-1292. Leiberman, A., & Whalen, D. (2000). On the relation of speech to language. Trends in Cognitive Science, 4, 187-196.
177 Levitt, H. (1971). Transformed up-down methods in psychoacoustics. Journal of the Acoustical Society of America, 49, 467-477. Leydon, C., Bauer, J., & Larson, C. (2003). The role of auditory feedback in sustaining vocal vibrato. Journal of the Acoustical Society of America 114 (3), 1575-1581. Ligois-Chauvel, C., Giraud, K., Badier, J., Marquis, P., & Chauvel, P. (2001). Intracerebral evoked potentials in pitch perception reve al a functional asymmetry of the human auditory cortex. Annals of the New York Academy of Sciences, 930, 117-132. Ligois-Chauvel, C., Peretz, I., Baba, M., Laguitton, V., & Chauvel, P. (1998). Contribution of different cort ical areas in the temporal lobes to music processing. Brain 121, 1853-1867. Lopez, L., Jrgens, R., Diekmann, V., Becker, W., Reid, S., Grzinger, B., et al. (2003). Musicians versus non-musicians: A neurophysiological approach. Annals of the New York Academy of Sciences, 999, 124-130. Ludlow, C. (2005). Central nervous system c ontrol of the laryngeal muscles in humans. Respiratory Physiology and Neurobiology 147, 205-222. Martony, J. (1968). On the correction of voice pi tch level for severely hard of hearing subjects. American Annals of the Deaf 113, 195-202. McNamara, C. (2005). Ways to look at training and development processes: Informal/formal and self-directed/othe r-directed. Retrieved August 24, 2005, from http://www.managementhelp.org/trngdev/ways/ways.htm Menning, H., Roberts, L., & Pantev, C. (2000). Plastic changes in the auditory cortex induced by intensive frequency discrimination training. NeuroReport, 11(4), 817822. Merzenich, M., Wright, B., Jenkins, W., Xerri, C., Byl, N., & Miller, S., et al. (1996). Cortical plasticity underlying perceptual, motor, and cognitive skill development: Implications for neurorehabilitation. Language comprehension in languagelearning impaired children improved with acoustically modified speech. Cold Spring Harbor Symposium on Quantitative Biology, 61 (5245), 1-8. Miller, G. (2000). The mating mind. New York: Doubleday.
178 Mody, M. (2004). Neurobiological correlates of language and reading impairments. In C. Stone, E. Silliman, B. Ehren, & K. Apel (Eds.), Handbook of language and literacy (pp. 49-72). New York: The Guilford Press. Monaghan, P., Metcalfe, N., & Ruxton, G. (1998). Does practice shape the brain? Nature 394, 4343. Moore, B. (1989). An introduction to the psychology of hearing (pp. 159-167). San Diego, CA: Academic Press. Moore, B. (1997). An introduction to the psychology of hearing, 4th Edition London: Academic Press Inc. Moore, B., & Peters, R. (1992). Pitch disc rimination and phase sensitivity in young and elderly subjects and its relations hip to frequency selectivity. Journal of the Acoustical Society of America 91 (5), 2881-2893. Moore, R. (1994). Effects of age, sex, and melodic/harmonic patterns on vocal pitchmatching skills of talented 8-11 year olds. Journal of Research in Music Education 42 (1), 5-13. Morr, M., Shafer, V., Kreuzer, J., & Kurt zberg, D. (2002). Maturation of mismatch negativity in typically developing infants a nd preschool children. Ear & Hearing 23 (2), 118-136. Mnte, T., Nager, W., Beiss, R., Schroeder, C., & Altenmller, E. (2003). Specialization of the specialized: Electrophysiological inve stigations in professional musicians. Annals of the New York Academy of Sciences, 999, 131-139. Mrbe, D., Pabst, F., Hofman, G., & Sundberg J. (2002). Significance of auditory and kinesthetic feedback to singers pitch control. Journal of Voice, 16 (1), 44-51. Mrbe, D., Pabst, F., Hofman, G., & Sundberg J. (2004). Effects of a professional solo singer education on auditory and kinesthe tic feedback A longitudinal study of singers pitch control. Journal of Voice 18 (2), 236-241. Murry, T. (1990). Pitch-matching accuracy in singers and non-singers. Journal of Voice, 4, 317-321. Ntnen, R. (1995). The mismatch negativity: A powerful tool for cognitive neuroscience. Ear & Hearing 16, 6-18. Ntnen, R., Gaillard, A., & Mntysalo, S. ( 1978). Early selection-attention effect on evoked potential reinterpreted. Acta Psychologica 42 (4), 313-329.
179 Ntnen, R., Paavilainen, P ., Tiitinen, H., Jiang, D., & Alho, K. (1993). Attention and mismatch negativity. Psychophysiology 30, 436-450. Ntnen, R., Pakarinen, S., Rinne, T., & Ta kegata, R. (2004). The mismatch negativity (MMN): towards the optimal paradigm. Clinical Neurophysiology 115, 140-144. Ntnen, R., & Picton, T. (1987). The N1 wa ve of the human electric and magnetic response to sound: A review and anal ysis of the component structure. Psychophysiology, 24, 375-425. Nager, W., Teder-Slejrvi, W., Kunze, S., & Mnte, T. (2003). Preattentive evaluation of multiple perceptual streams in human audition. Cognitive Neuroscience and Neuropsychology 14 (6), 871-874. Nousak, J., Deacon, D., Ritter, W., & Vaughan, H. (1996). Storage of information in transient auditory memory. Cognitive Brain Research 4, 305-317. Novak, G., Ritter, W., & Vaughan, H. (1992). Mismatch detection and the latency of temporal judgments. Psychophysiology 29, 398-411. Novitski, N., Tervaniemi, M., Huotilaine n, M., & Ntnen, R. (2004). Frequency discrimination at different frequency le vels as indexed by electrophysiological and behavioral measures. Cognitive Brain Research, 20, 26-36. Opitz, B., Mecklinger, A., Friederici, A., & von Cramon, D. (1999). The functional neuroanatomy of novelty processing: integrating ERP and fMRI results. Cerebral Cortex, 9, 379-391. Overy, K. (2003). Dyslexia and music. Annals of the New York Academy of Sciences, 999, 497-505. Pantev, C., Engelien, A., Candia, V., & Elbe rt, T. (2001). Represen tational cortex in musicians plastic alterations in response to musical practice. Annals of the New York Academy of Sciences, 930, 300-314. Pantev, C., Oostenveld, R., Engelien, A., Ross, B., Roberts, L., & Hoke, M. (1998). Increased auditory cortical representation in musicians, Nature, 23, 811-814. Pantev, C., Ross, B., Fujioka, T., Trainor, L., Schulte, M., & Schulz, M. (2003). Music and learning-induced cortical plasticity. Annals of the New York Academy of Sciences, 999, 438-450.
180 Pascual-Leone, A. (2001). The brain th at plays music and is changed by it. Annals of the New York Academy of Sciences, 930, 315-329. Pascual-Leone, A., Nguyet, D., Cohen, L., Br asil-Neto, J., Cammarota, A., & Hallett, M. (1995). Modulation of muscle response s evoked by transcranial magnetic stimulation during the acquisition of new fine motor skills. Journal of Neurophysiology 74, 1037-1044. Patel, A., & Balaban, E. (2001). Human pitch perception is reflected in the timing of stimulus-related cortical activity. Nature Neuroscience 4 (8), 839-844. Pedersen, D., & Pedersen, N. (1970). The relationship between pitch recognition and vocal pitch production in sixth grade students. Journal of Research in Music Education, 18, 265-272. Pekkonen, E., Rinne, T., & Ntnen, R. (1995) Variability and replicability of the mismatch negativity. Electroencephalography and Clinical Neurophysiology, 96, 546-554. Peretz, I. (1990). Processing of local and global musical in formation by unilateral braindamaged patients. Brain, 113, 1185-1205. Peretz, I., & Coltheart, M. (2003) Modularity of music processing. Nature Neuroscience, 6, 688-691. Perrin, E., Berger-Vachon, C., Topouzkha nian, A., Truy, E., & Morgon, A. (1999). Evaluation of cochlear implanted childrens voices. International Journal of Pediatric Otorhinolaryngology 47, 181-186. Perry, D., Zatorre, R., Petrides, M., Alivis atos, B., Meyer, E., & Evans, A. (1999). Localization of cerebral ac tivity during simple singing. NeuroReport 10 (18), 3979-3984. Pettigrew, C., Murdoch, B., Ponton, C., Finniga n, S., Alku, P., Kei, J., et al., (2004). Automatic auditory processing of Englis h words as indexed by the mismatch negativity, using a multiple deviant paradigm. Ear & Hearing 25 (3), 284-301. Phillips, K., & Aitchison, R. (1997). The relationship of singing accuracy to pitch discrimination and tonal aptit ude among third-grade students. Contributions to Music Education 24 (1), 7-22. Picton, T. (1995). The neurophysiological ev aluation of auditory discrimination. Ear & Hearing 16, 1-5.
181 Porter, S. (1977). The effect of multiple discrimination training on pitch-matching behaviors of uncertain singers. Journal of Research in Music Education 25, 6882. Platel, H. (2002). Neuropsy chology of musical percep tion: new perspectives. Brain 125, 223-224. Ramig, L., Yoshiyuki, H., & Bonitati, C. ( 1991). The efficacy of voice therapy for patients with Parkinsons disease. NCVS Status and Progress Report -1, 61-86. Rauschecker, J. (2001). Cortical plasticity and music. Annals of the New York Academy of Sciences, 930, 330-336. Ries, N. (1987). An analysis of the characte ristics of infant-child singing expressions: Replication report. Canadian Journal of Research in Music Education, 29 (1), 520. Robertson, I., & Murre, J. (1999). Rehabilita tion of brain damage: Brain plasticity and principles of guided recovery. Psychological Bulletin 125 (5), 544-575. Roubeau, B., Chevrie-Muller, C., & Saint Gu ily, J. (1997). Electromyographic activity of strap and cricothyroid mu scles in pitch change. Acta Otolaryngologica 117, 459464. Sams, M., Paavilainen, P., Alho, K., & N tnen, R. (1985). Auditory frequency discrimination and event-related potentials. Electroencephalography and Clinical Neurophysiology/Evoked Potentials 62 (6), 437-448. Sapir, S., McClean, M., & Larson, C. (1983). Human laryngeal responses to auditory stimulation. Journal of the Acoustical Society of America, 73, 315-321. Sawashima, M., & Hirose, H. (1983). Laryng eal gestures in speech production. In P. McNeilage (Ed.), The production of speech. New York: Springer. Sawashima, M., Hirose, H., Honda, K., Yoshioka H., Hibi, S., Kawase, N., et al. (1983). Stereoendoscopic measurement of the laryng eal structure. In D. Bless and J. Albbs (Eds.), Vocal fold physiology. Contemporar y research and clinical issues (pp. 264-276). San Diego: College-Hall. Scharf, B. (1970). Critical bands. In J. Tobias (Ed.), Foundations of modern auditory theory, Vol. 1. (pp. 157-202). New York: Academic Press. Schellenberg, E. (2004). Music lessons enhance IQ. Psychological Science, 15 (8), 511514.
182 Schenk, B., Baumgartner, W., & Hamzavi, J. (2003). Changes in vowel quality after cochlear implantation. Journal for Oto-Rhino-Laryngology and Its Related Specialties 65 (3), 184-188. Scherg, M., Vajsar, J., & Pict on, T. (1989). A source analysis for the late human auditory evoked potentials. Journal of Cognitive Neuroscience, 1, 336-355. Schlaug, G. (2001). The brain of musicians: A model for functional and structural adaptation. Annals of the New York Academy of Sciences, 930, 281-299. Schmithorst, V., & Holland, S. (2003). The eff ect of music training on music processing: A fMRI study in humans. Neuroscience 348, 65-68. Schn, D., Magne, C., & Besson, M. (2004). The music of speech: Music training facilitates pitch processing in both music and language. Psychophysiology, 41, 341-349. Schrger, E., & Winkler, I. (1995). Presentati on rate and magnitude of stimulus deviance effects on human pre-atte ntive change detection. Neuroscience Letters 193, 185188. Seashore, C. (1919). The psychology of musical talent. New York: Silver Burdett Company. Seashore, C. (1939). Measurement of musical talent. Camden, NJ: RCA Victor Division. Seifert, E., Oswald, M., Bruns, U., Vische r, M., Kompis, M., & Haeusler, R. (2002). Changes of voice and articulation in children with cochlear implants. International Journal of Pediatric Otorhinolaryngology 66, 115-123. Sek, A., & Moore, B. (1995). Frequency di scrimination as a function of frequency, measured in several ways. Journal of the Acoustical Society of America, 97 (4), 2479-2486. Shahin, A., Bosnyak, D., Trainor, L., & Robert s, L. (2003). Enhancement of neuroplastic P2 and N1c auditory evoked potentials in musicians. The Journal of Neuroscience, 23 (12), 5545-5552. Shipp, T., & Izdebski, K. (1981). Current evidence for existence of laryngeal macrotremor and microtremor. Journal of Forensic Science, 26, 501-505.
183 Sinkkonen, J., Kaski, S., Huotilainen, M., Il moniemi, R., Ntnen, R., & Kaila, K. (1996). Optimal resource allocation for novelty detection in a human auditory memory. NeuroReport 7, 2479-2482. Sinkkonen, J., & Tervaniemi, M. (2000). Towards optimal recording and analysis of the mismatch negativity. Audiology & Neuro-Otology 5, 235-246. Speigel, M., & Watson, C. (1984). Performa nce on frequency discrimination tasks by musicians and non-musicians. Journal of the Acoustical Society of America, 76, 1690-1696. Steinmetz, H., & Seitz, R. (1991). Functi onal anatomy of language processing: Neuroimaging and the problem of individual variability. Neuropsychologia, 29, 1149-1161. Sundberg, J. (1987). The science of the singing voice DeKalb, IL: Northern Illinois University Press. Sundberg, J. (1994). Perceptu al aspects of singing. Journal of Voice, 8 (2), 106-122. Suziki, M., & Sasaki, C. (1977) Effect of various sensor y stimuli on reflex laryngeal adduction. Annuals of Otolaryngology 86, 30-36. Tervaniemi, M. (1993). Absolute pitch and event-related brain potentials. Music Perception, 10, 305-316. Tervaniemi, M. (2001). Musical sound processi ng in the human brain: Evidence from electric and magnetic recordings. Annals of the New York Academy of Sciences 930, 259-272. Tervaniemi, M., & Brattico, E. (2004). From sounds to music. Journal of Consciousness Studies, 11 (3-4), 9-27. Tervaniemi, M., Castaneda, A., Knoll, & Uther, M. (in press) Sound processing in amateur musicians and nonmusicians: ERP and behavioral indices. NeuroReport Tervaniemi, M., & Hugdahl, K. (2003). Late ralization of auditory-cortex functions. Brain Research Reviews, 43, 231-246. Tervaniemi, M., & Huotilainen, M. (2003). The promises of change-related brain potentials in cognitive neuroscience of music. Annals of the New York Academy of Sciences, 999, 29-39.
184 Tervaniemi, M., Ilvonen, T., Sinkkonen, J., Kujala, A., Alho, K., Huotilainen, M., et al. (2000). Harmonic partials facilitate pitch discrimination in humans: Electrophysiology and behavioral evidence. Neuroscience Letters, 279, 29-32. Tervaniemi, M., Just, V., Koelsch, S., Widm ann, A., & Schrger, E. (2005). Pitch discrimination accuracy in musicians vs nonmusicians: An event-related potential and behavioral study. Experimental Brain Research 161, 1-10. Tervaniemi, M., Medvedev, S., Alho, K., Pakhom ov, S., Roudas, M., van Zuijen, T. et al. (2000). Lateralized automatic auditory processing of phonetic versus musical information: a PET study. Human Brain Mapping, 10, 74-79. Tervaniemi, M., Schrger, E., Saher, M., & Ntnen, R. (2000). Effects of spectral complexity and sound duration in comple x-sound pitch processing in humans: A mismatch negativity study. Neuroscience Letters 290, 66-70. Teter, B., & Ashford, J. (2002). Neuropl asticity in Alzheimers disease. Journal of Neuro-Science Research 70 (3), 402-437. Titze, I. (1994). Principles of voice production. Englewood Cliffs, NJ: Prentice-Hall, Inc. Titze, I., Luchei, E., & Hirano, M. (1989). Ro le of the thyroid-arytenoid muscle in regulation of fundamental frequency. Journal of Voice 3, 213-224. Trainor, L., McDonald, K., & Al ain, C. (2002). Automatic a nd controlled processing of melodic contour and interval information measured by electrical brain activity. Journal of Cognitive Neuroscience 14, 430-442. Trainor, L., Shahin, A., & Roberts, L. (2003) Effects of music training on the auditory cortex in children. Annals of the New York Academy of Sciences, 999, 506-513. Turner, C., & Nelson, D. (1982). Frequency di scrimination in regions of normal and impaired sensitivity. Journal of Speech and Hearing Research, 25, 34-41. van Zuijen, T., Sussman, E., Winkler, I., N tnen, R., & Tervaniemi, M. (2004). Grouping of sequential sounds-An eventrelated potential study comparing musicians and nonmusicians. Journal of Cognitive Neuroscience 16 (2), 331-338. Verleger, R. (1988). Event-related potentials and memory: A critique of the context updating hypothesis and an altern ative interpretation of P3. Behavioral and Brain Sciences 11, 343-356.
185 Waldstein, R. (1990). Effect s of postlingual deafness speech on speech production: Implications for the role of auditory feedback. Journal of the Acoustical Society of America 88, 2099-2114. Ward, D., & Burns, E. (1978). Singing without auditory feedback. Journal of Research in Singing 1, 24-44. Watts, C., Barnes-Burroughs, K., Adrianopou los, M., & Carr, M. (2003). Potential factors related to untrain ed singing talent: A survey of singing pedagogues. Journal of Voice 17 (3), 298-307. Watts, C., Murphy, J., & Barnes-Burroughs, K. (2003). Pitch matc hing accuracy of trained singers, untrained subjects with talented singing voices, and untrained subjects with non-talented singing voices in conditions of varying feedback. Journal of Voice 17 (2), 185-194. Webster, D. (1999). Neuroscience of communication (2nd Ed.). San Diego: Singular Publishing Group, Inc. Weinberger, N. (1994). Music and cognitive achievement in children. MuSICA Research Notes 1 (2). Retrieved October 19, 2003, from http://www.musica.uci.edu/mrn/V1I2F94.html Welch, G. (1994). The assessment of singing. Psychology of Music, 22, 3-19. Wood, C., & Wolpaw, J. (1982). Scalp distribut ion of human auditory evoked potentials. II. Evidence multiple sources and involvement of auditory cortex. Electroencephalography and Clinical Neurophysiology, 54, 25-38. Wyke, B. (1967). Advances in the neurology of phonation: Phonatory reflex mechanisms in the larynx. British Journal of Communication 2, 2-14. Wyke, B. (1974). Laryngeal neuromus cular control systems in singing. Folia Phoniatrica 26, 295-306. Yarbrough, C., Green, G., Benson, W., & Bowe rs, J. (1991). Inaccurate singers: An exploratory study of variable s affecting pitch-matching. Bulletin of the Council for Research in Music Education, 107, 23-33. Zatorre, R. (1988). Pitch per ception of complex tones and hu man temporal lobe function. Journal of the Acoustical Society of America, 84, 566-572. Zatorre, R. (2001). Neural specia lizations for tonal processing. Annuals of the New York Academy of Sciences, 930, 193-210.
186 Zatorre, R. (2003). Music and the brain. Annals of the New York Academy of Sciences, 999, 4-14. Zatorre, R., Belin, P., & Penhune, V. (2002). St ructure and function of auditory cortex: Music and speech. Trends in Cognitive Science, 6, 37-46. Zatorre, R, Evans, A., & Meyer, E. ( 1994). Neural mechanisms underlying melodic perception and memory for pitch. Journal of Neuroscience, 14 (4), 1908-1919. Zatorre, R., Evans, A., Meyer, E., & Gje dde, A. (1992). Lateralization of phonetic and pitch discrimination in speech processing. Science 256, 846-849. Zatorre, R., Perry, D., Beckett, C., West bury, C., & Evans, A. (1998). Functional anatomy of musical processing in listeners with absolute pitch and relative pitch. Proceedings of the National Academy of Science USA, 95, 3172-3177. Zemlin, W. (1999). Speech and hearing science (4th Ed.). Boston: AllynBacon. Zeng, F. (2002). Temporal pitch in electric hearing. Hearing Research 174, 101-106. Zwissler, R. (1972). An investigation of pitc h discrimination skills of first grade children identified as accurate singers and those id entified as inaccurate singers (Doctoral dissertation, University of Ca lifornia at Los Angeles, 1971). Dissertation Abstracts International, 32, 4056A-4057A.
188 Appendix A: Participan t Screening Questionnaire PARTICIPANT SCREENING QUESTIONNAIRE Subject Code: ________________________ Birth Date: _________________ INSTRUCTIONS: Please read each question and answer accordingly. Either circle YES or NO or complete a short response. If you have any qu estions, please feel free to ask the investigator. GENERAL INFORMATION 1. Are you female? YES NO 2. Are you between the ages of 18 and 35? YES NO 3. Are you a native speaker of English (learned English as a child)? YES NO 4. Are you a fluent speaker of a tonal language (e.g., Vietnamese)? YES NO 5. Are you naturally right-handed? YES NO 6. Have you participated in other experiments where you listened to musical sounds and determined if they were the same or different? YES NO 7. Do you have absolute pitch ability (AP), also known as perfect pitch? YES NO MEDICAL 8. Do you have any history of vocal cord disease or phonosurgery, YES NO such as nodules, polyps, cyst? 9. Do you have any history of neurological illness or disease? YES NO 10. Do you now or have you ever had a habit of cigarette smoking? YES NO 11. Do you now or have you ever abused alcohol or drug use? YES NO 12. Are you currently having allergy symptoms or respiratory problems that affect your voice or hearing? YES NO 13. Do you have any known hearing impairment? YES NO SCREENINGS 1. Hearing Screening @ 25dB R 250 500 1000 2000 4000 Hz L 250 500 1000 2000 4000 Hz PASS or FAIL 2. Vocal Sweep A3 A 4 PASS or FAIL ACCEPT DO NOT ACCEPT
189 Appendix B: Participant Information Questionnaire PARTICIPANT INFORMATION QUESTIONNAIRE Subject Code: ________________________ Birth Date: _________________________ INSTRUCTIONS: Please read each question and answer accordingly. Either circle YES or NO or complete a short response. If you have any qu estions, please feel free to ask the investigator. EDUCATION 1. Have you had less than 12 months of musical instruction, such as band, YES NO piano, guitar, or voice lessons? 2. Have you had 5 years or more of singing (voice) lessons/instruction? YES NO If YES, please answer questions a. through h. then go to question # 4. If NO, go to question # 3. a. At what age did you begin your music training? _________________ b. At what age did you begin your vocal training? __________________ c. How many total years of music training have you had? ___________ d. How many total years of just vocal training have you had? _________ e. Is your vocal training in classical voice? YES NO f. Is your vocal training in choral voice? YES NO g. Combination? If yes, please describe. ____________________________________________ ______________________________________________________________________________ h. Other? Please describe. _______________________________________________________ ______________________________________________________________________________ 3. Have you had 5 years or more of instrumental musical lessons? YES NO If yes, what is your primary instrument? ________________________ At what age did you begin your music training? _________________ How many total years of music training have you had? ___________ 4. Has any immediate family member also had 5 years or more of voice or YES NO instrumental lessons? If yes, explain _______________________________________________ ______________________________________________________________________________ 5. What was your best math score on your SAT exam? Please circle the range. Less than 400 400 499 500 599 600 699 700 800 6. What was your best verbal score on your SAT exam? Please circle the range. Less than 400 400 499 500 599 600 699 700 800
190 Appendix C: Informed Consent Form Space below reserved for IRB Stamp Please leave blank Informed Consent Social and Behavioral Sciences University of South Florida Information for People Who Take Part in Research Studies The following information is being presented to help you decide whether or not you want to take part in a minimal risk research study. Please read this carefully. If you do not understand anything, ask the person in charge of the study. Title of Study: Auditory Neural Plasticity in Trained Vocal Musicians Principal Investigator: Deborah Adams Nikjeh Study Location(s): University of South Florida; Dept. of Communication Sciences & Disorders, PCD 3008 and PCD 3006 You are being asked to participate because this st udy will compare pitch perception and production among trained female vocal musicians an d musically untrained females. General Information about the Research Study The purpose of this research study is to assess, compare and correlate three identified physiological variables that contribute to the performance of the si nging voice. Those three va riables are: (1) auditory pitch discrimination, (2) vocal pitch matching accuracy, and (3) pre-attentive auditory pitch discrimination. Plan of Study If you agree to participate, you will first be asked to provide some basic information about your music training, education and language background, and general health. If you meet the criteria for this study, you will then have your vocal pitch range and your hear ing screened. If you pass these two screenings, you are ready to begin the study. Ther e are three tasks in this study. First, you will be seated in front of a computer monitor in a sound treated booth. You will hear two sounds through earphones. Then, two boxes will appear on the computer screen, labeled same and d ifferent. You decide if the two sounds are same or different and mouse click on the proper box. Fo r the next task, you will also wear earphones with a microphone in the sound treated booth. You will hear a piano tone for 2 seconds. You then sing the same tone (pitch) aloud on the sound ah for 3 seconds into the microphone and your voice is recorded. For the third task, you may rest comfortably in a reclining chair in a sound treated booth and watch a closed captioned movie of your choice without the sound. Instead, you will be wearing earphones and very soft and fast sounds will be played thro ugh the earphones. A cap with sma ll electrodes will be placed on your head to record your brains responses to the sounds. A tiny amount of cream is applied between each electrode and your scalp. There is no pain whatsoever. You can relax and watch the movie while your brain waves are recorded. The total time for participation is typically 2 to 3 hours and will be broken down into 2 visits, approximately 60 to 90 minutes each.
Appendix C: (Continued) Payment for Participation You will be entitled to one of the following: (1) Extr a credit in a pre-determined course in Communication Sciences and Disorders, (2) Extra credit in a pre-dete rmined course in the School of Music or (3) $10 per hour for your participation. If you withdraw from th e study before completion, payment or extra credit will be pro-rated based on actual time volunteered to the closest hour. Benefits of Being a Part of this Research Study You will not benefit directly from this study. Ho wever, your participation will help to increase our understanding of the function and adaptability of the human brain. Risks of Being a Part of this Research Study There are no known risks from participation in this study. Confidentiality of Your Records Your privacy and research records will be kept confiden tial to the extent of the law. Authorized research personnel, employees of the Department of Health and Human Services, and the USF Institutional Review Board may inspect the records fr om this research project. The results of this study may be published. However, the data obtained from you will be combined with data from others in the publication. The publis hed results will not include your name or any other information that would personally identify you in any way. The computer files with your data will be identified by an arbitrary code that will not be connected to your name. The consent forms and questionnaires will be kept separately in a locked file cabinet. Volunteering to Be Part of this Research Study Your decision to participate in this research study is completely voluntary. You are free to participate in this research study or to withdraw at any time. Th ere will be no penalty or negative consequences if you stop taking part in the study. Questions and Contacts If you have any questions about this research study, contact Dee Adams Nikjeh at Nikjeh@mail.usf.edu or Dr. Stefan A. Frisch at (813) 974-6563 or Frisch@cas.usf.edu If you have questions about your rights as a person who is taking part in a research study, you may contact the Division of Research Compliance of the University of South Florida at (813) 9745638. Consent to Take Part in This Research Study By signing this form I agree that: I have fully read or have had read and explained to me this informed consent form describing this research project. I have had the opportunity to question one of the persons in charge of this research and have received satisfactory answers. I understand that I am being asked to participate in research. I understand the risks and benefits, and I freely give my consent to participate in the research project outlined in this form, under the conditions indicated in it.
192 Appendix C: (Continued) I have been given a signed copy of this informed consent form, which is mine to keep. _________________________ ____________________________________________ Signature of Participant Printed Name of Participant Date Investigator Statement I have carefully explained to the subject the nature of the above research study. I hereby certify that to the best of my knowledge the subject signing this consent form understands the nature, demands, risks, and benefits involved in participating in this study. _________________________ ____________________________________________ Signature of Investigator Printed Name of Investigator Date Or authorized research investigator designated by the Principal Investigator
193 Appendix D: Participant Profile Data GROUP SUBJECTS AGE MOS AGE TRNG INITIATED TTL YRS TRNG YRS and INSTRUM AGE VCL TRNG INIT YEARS VOCAL MOTHER FATHER SIBLING SATMATH SATVERBAL CONTROL C01 245 4 4 CONTROL C02 266 CONTROL C03 399 4 6 CONTROL C04 250 CONTROL C05 406 5 5 CONTROL C06 242 5 4 CONTROL C07 249 5 5 CONTROL C08 241 CONTROL C09 256 5 5 CONTROL C10 239 5 5 CONTROL C11 403 6 7 CONTROL C12 361 X 5 5 CONTROL C13 253 5 7 CONTROL C14 247 4 4 CONTROL C15 252 6 4 CONTROL C16 256 6 5 CONTROL C17 265 5 6 CONTROL C18 254 5 4 CONTROL C19 292 5 6 CONTROL C20 262 5 5 CONTROL C21 251 X 5 5
194 Appendix D: (Continued) GROUP SUBJECTS AGE MOS AGE TRNG INITIATED TTL YRS TRNG YRS and INSTRUM AGE VCL TRNG INIT YEARS VOCAL MOTHER FATHER SIBLING SATMATH SATVERBAL INSTRUMENTAL IN01 260 11 10 10 wind 0 7 6 INSTRUMENTAL IN02 237 10 8 8 wind 0 X 5 7 INSTRUMENTAL IN03 254 3 15 15 string 0 5 5 INSTRUMENTAL IN04 223 7 8 8 string 0 6 5 INSTRUMENTAL IN05 265 6 15 15 wind 0 X 6 6 INSTRUMENTAL IN06 226 11 8 8 string 0 4 6 INSTRUMENTAL IN07 267 11 11 11 wind 0 4 5 INSTRUMENTAL IN08 244 10 9 9 wind 0 X 4 5 INSTRUMENTAL IN09 265 7 8 8 string 0 7 5 INSTRUMENTAL IN10 262 5 13 13 string 0 X 5 5 INSTRUMENTAL IN11 259 4 8 8 string 0 6 5 INSTRUMENTAL IN12 252 15 7 7 wind 0 4 4 INSTRUMENTAL IN13 228 11 7 7 wind 0 5 5 INSTRUMENTAL IN14 257 11 9 9 string 0 5 5 INSTRUMENTAL IN15 356 11 7 7 wind 0 X X X 6 6 INSTRUMENTAL IN16 303 10 15 15 string 0 X X X 5 6 INSTRUMENTAL IN17 335 11 9 9 wind 0 X 7 5 INSTRUMENTAL IN18 234 11 10 10 string 0 X 5 6 INSTRUMENTAL IN19 220 4 8 8 string 0 X 6 6 INSTRUMENTAL IN20 286 5 13 13 string 0 5 5 INSTRUMENTAL IN21 233 10 8 8 wind 0 6 4
195 Appendix D: (Continued) GROUP SUBJECTS AGE MOS AGE TRNG INITIATED TTL YRS TRNG YEARS INSTRUM AGE VCL TRNG INIT YEARS VOCAL MOTHER FATHER SIBLING SATMATH SATVERBAL VOCAL V01 224 5 13 0 5 13 4 5 VOCAL V02 293 11 13 9 17 7 X X 3 3 VOCAL V03 281 6 17 9 9 9 X 5 5 VOCAL V04 227 5 13 9 13 5 X 5 5 VOCAL V05 234 8 6 0 8 6 5 4 VOCAL V06 221 11 8 0 11 8 X 6 5 VOCAL V07 263 8 14 0 8 14 X 6 5 VOCAL V08 232 11 9 5 11 9 X 5 6 VOCAL V09 270 8 11 6 8 6 X 6 6 VOCAL V10 229 9 10 0 9 10 X X 6 5 VOCAL V11 225 7 11 4 12 6 X 5 6 VOCAL V12 235 9 10 0 9 7 5 5 VOCAL V13 270 6 8 5 8 8 X 6 5 VOCAL V14 248 12 8 6 16 5 6 5 VOCAL V15 258 11 10 0 14 7 5 5 VOCAL V16 276 12 11 0 12 11 X X X 4 5 VOCAL V17 335 8 19 13 8 19 X X 5 7 VOCAL V18 257 8 13 0 12 9 6 4 VOCAL V19 396 5 13 0 5 13 VOCAL V20 265 10 8 0 10 8 5 5
196 Appendix D: (Continued) SUMMARY OF SUBJECT DATA GROUP AGEYEARS AGE TRAINING INITIATED TOTAL YEARS TRAINING YEARS INSTRUM AGE VOCAL TRNG INIT YEARS VOCAL MOTHER FATHER SIBLING SATMATH SATVERBAL CONTROL Mean 23.4 N/A N/A N/A N/A N/A 1.0 0.0 1.0 5.0 5.1 Max 33.8 N/A N/A N/A N/A N/A 6.0 7.0 Min 19.9 N/A N/A N/A N/A N/A 4.0 4.0 INSTRUMENTAL Mean 21.7 8.8 9.8 9.8 N/A 0.0 4.0 2.0 7.0 5.4 5.3 Max 29.7 15.0 15.0 15.0 N/A 0.0 7.0 7.0 Min 18.3 3.0 7.0 7.0 N/A 0.0 4.0 4.0 VOCAL Mean 21.8 8.5 11.3 3.3 10.3 9.0 6.0 5.0 6.0 5.2 5.1 Max 33.0 12.0 19.0 13.0 17.0 19.0 6.0 7.0 Min 18.4 5.0 6.0 0.0 5.0 5.0 3.0 3.0
197 Appendix E: Individual Data for Pitch Production Accuracy Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff% Stimulus Prod Hz Diff Hz Diff% C01 261.63/C4 309.54 47.91 18.3% 269.48/C4+qtr 291.59 22.11 8.2% 277.32/C4# 265.97 11.35 4.1% C02 261.63/C4 272.88 11.25 4.3% 269.48/C4+qtr 321.95 52.47 19.5% 277.32/C4# 266.45 10.87 3.9% C03 261.63/C4 234.94 26.69 10.2% 269.48/C4+qtr 238.26 31.22 11.6% 277.32/C4# 266.65 10.67 3.8% C04 261.63/C4 194.68 66.95 25.6% 269.48/C4+qtr 205.27 64.21 23.8% 277.32/C4# 206.12 71.2 25.7% C05 261.63/C4 255.91 5.72 2.2% 269.48/C4+qtr 259.71 9.77 3.6% 277.32/C4# 266.53 10.79 3.9% C06 261.63/C4 261.33 0.3 0.1% 269.48/C4+qtr 259.52 9.96 3.7% 277.32/C4# 275.28 2.04 0.7% C07 261.63/C4 252.14 9.49 3.6% 269.48/C4+qtr 268.38 1.1 0.4% 277.32/C4# 277.9 0.58 0.2% C08 261.63/C4 263.23 1.6 0.6% 269.48/C4+qtr 273.82 4.34 1.6% 277.32/C4# 280.12 2.8 1.0% C09 261.63/C4 227.02 34.61 13.2% 269.48/C4+qtr 224.91 44.57 16.5% 277.32/C4# 231.27 46.05 16.6% C10 261.63/C4 236.53 25.1 9.6% 269.48/C4+qtr 249.35 20.13 7.5% 277.32/C4# 273.65 3.67 1.3% C11 261.63/C4 254.81 6.82 2.6% 269.48/C4+qtr 255.3 14.18 5.3% 277.32/C4# 276.1 1.22 0.4% C12 261.63/C4 252.28 9.35 3.6% 269.48/C4+qtr 264.81 4.67 1.7% 277.32/C4# 274.4 2.92 1.1% C13 261.63/C4 245.17 16.46 6.3% 269.48/C4+qtr 245.3 24.18 9.0% 277.32/C4# 253.74 23.58 8.5% C14 261.63/C4 265.06 3.43 1.3% 269.48/C4+qtr 270.23 0.75 0.3% 277.32/C4# 276.1 1.22 0.4% C15 261.63/C4 207.57 54.06 20.7% 269.48/C4+qtr 237.9 31.58 11.7% 277.32/C4# 219.87 57.45 20.7% C16 261.63/C4 202.68 58.95 22.5% 269.48/C4+qtr 187.17 82.31 30.5% 277.32/C4# 213.18 64.14 23.1% C17 261.63/C4 265.2 3.57 1.4% 269.48/C4+qtr 272.79 3.31 1.2% 277.32/C4# 282.12 4.8 1.7% C18 261.63/C4 254.33 7.3 2.8% 269.48/C4+qtr 257.74 11.74 4.4% 277.32/C4# 272.92 4.4 1.6% C19 261.63/C4 259.98 1.65 0.6% 269.48/C4+qtr 270.41 0.93 0.3% 277.32/C4# 283.3 5.98 2.2% C20 261.63/C4 253.18 8.45 3.2% 269.48/C4+qtr 252.92 16.56 6.1% 277.32/C4# 270.39 6.93 2.5% C21 261.63/C4 258.14 3.49 1.3% 269.48/C4+qtr 269.85 0.37 0.1% 277.32/C4# 277.59 0.27 0.1%
198 Appendix E: (Continued) Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff% Stimulus Prod Hz Diff Hz Diff% C01 329.63/E4 341.34 11.71 3.6% 320.03/E4-qtr 355.6 3557 11.1% 310.97/D4# 351 40.03 12.9% C02 329.63/E4 378.49 48.86 14.8% 320.03/E4-qtr 342.21 2218 6.9% 310.97/D4# 345.37 34.4 11.1% C03 329.63/E4 305.6 24.03 7.3% 320.03/E4-qtr 304.26 15.77 4.9% 310.97/D4# 276.7 3427 11.0% C04 329.63/E4 242.28 8735 26.5% 320.03/E4-qtr 221.11 9892 30.9% 310.97/D4# 233.98 7699 24.8% C05 329.63/E4 338.53 8.9 2.7% 320.03/E4-qtr 315.01 5.02 1.6% 310.97/D4# 276.4 3457 11.1% C06 329.63/E4 322.78 6.85 2.1% 320.03/E4-qtr 314.24 5.79 1.8% 310.97/D4# 308.32 2.65 0.9% C07 329.63/E4 327.36 227 0.7% 320.03/E4-qtr 304.26 15.77 4.9% 310.97/D4# 291.47 19.5 6.3% C08 329.63/E4 323.63 6 1.8% 320.03/E4-qtr 315.55 4.48 1.4% 310.97/D4# 307.88 3.09 1.0% C09 329.63/E4 238.75 90.88 27.6% 320.03/E4-qtr 236.84 8319 26.0% 310.97/D4# 277.03 3394 10.9% C10 329.63/E4 326.47 316 1.0% 320.03/E4-qtr 325.68 5.65 1.8% 310.97/D4# 301.15 9.82 3.2% C11 329.63/E4 326.82 2.81 0.9% 320.03/E4-qtr 317.46 257 0.8% 310.97/D4# 310.67 0.3 0.1% C12 329.63/E4 314.07 1556 4.7% 320.03/E4-qtr 325.78 5.75 1.8% 310.97/D4# 301.6 9.37 3.0% C13 329.63/E4 318.82 10.81 3.3% 320.03/E4-qtr 286.78 3325 10.4% 310.97/D4# 255.22 55.75 17.9% C14 329.63/E4 323.77 5.86 1.8% 320.03/E4-qtr 319.77 026 0.1% 310.97/D4# 309.61 136 0.4% C15 329.63/E4 238.63 91 27.6% 320.03/E4-qtr 249.17 70.86 22.1% 310.97/D4# 283.81 2716 8.7% C16 329.63/E4 307.95 21.68 6.6% 320.03/E4-qtr 227.31 92.72 29.0% 310.97/D4# 219.93 91.04 29.3% C17 329.63/E4 331.67 2.04 0.6% 320.03/E4-qtr 301.54 18.49 5.8% 310.97/D4# 311.7 0.73 0.2% C18 329.63/E4 312.22 17.41 5.3% 320.03/E4-qtr 311.83 8.2 2.6% 310.97/D4# 294.37 16.6 5.3% C19 329.63/E4 286.15 43.48 13.2% 320.03/E4-qtr 316.82 321 1.0% 310.97/D4# 311.61 0.64 0.2% C20 329.63/E4 300.31 2932 8.9% 320.03/E4-qtr 272.76 4727 14.8% 310.97/D4# 264.92 46.05 14.8% C21 329.63/E4 324.43 5.2 1.6% 320.03/E4-qtr 318.77 126 0.4% 310.97/D4# 305.53 5.44 1.7%
199 Appendix E: (continued) Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % C01 392.00/G4 405.44 13.44 3.4% 380.58/G4-qtr 412.64 32.06 8.4% 369.81/F4# 376.96 7.15 1.9% C02 392.00/G4 428.19 36.19 9.2% 380.58/G4-qtr 405.95 25.37 6.7% 369.81/F4# 373.7 3.89 1.1% C03 392.00/G4 360.43 31.57 8.1% 380.58/G4-qtr 340.99 39.59 10.4% 369.81/F4# 329.75 40.06 10.8% C04 392.00/G4 255.13 136.87 34.9% 380.58/G4-qtr 258.88 121.7 32.0% 369.81/F4# 230.85 138.96 37.6% C05 392.00/G4 382.85 9.15 2.3% 380.58/G4 -qtr 373.66 6.92 1.8% 369.81/F4# 364.87 4.94 1.3% C06 392.00/G4 387.46 4.54 1.2% 380.58/G4 -qtr 374.8 5.78 1.5% 369.81/F4# 364.88 4.93 1.3% C07 392.00/G4 390.48 1.52 0.4% 380.58/G4 -qtr 381.36 0.78 0.2% 369.81/F4# 372.81 3 0.8% C08 392.00/G4 385.08 6.92 1.8% 380.58/G4 -qtr 372.15 8.43 2.2% 369.81/F4# 366.93 2.88 0.8% C09 392.00/G4 248.34 143.66 36.6% 380.58/G4 -qtr 253.93 126.65 33.3% 369.81/F4# 242.98 126.83 34.3% C10 392.00/G4 429.68 37.68 9.6% 380.58/G4 -qtr 438.65 58.07 15.3% 369.81/F4# 371.74 1.93 0.5% C11 392.00/G4 384.12 7.88 2.0% 380.58/G4 -qtr 377.39 3.19 0.8% 369.81/F4# 367.87 1.94 0.5% C12 392.00/G4 384.6 7.4 1.9% 380.58/G4 -qtr 368.76 11.82 3.1% 369.81/F4# 357.12 12.69 3.4% C13 392.00/G4 315.56 76.44 19.5% 380.58/G4 -qtr 301.49 79.09 20.8% 369.81/F4# 349.78 20.03 5.4% C14 392.00/G4 393.23 1.23 0.3% 380.58/G4 -qtr 461.27 80.69 21.2% 369.81/F4# 367.18 2.63 0.7% C15 392.00/G4 354.39 37.61 9.6% 380.58/G4 -qtr 345.36 35.22 9.3% 369.81/F4# 357.57 12.24 3.3% C16 392.00/G4 361.83 30.17 7.7% 380.58/G4 -qtr 374.78 5.8 1.5% 369.81/F4# 300.9 68.91 18.6% C17 392.00/G4 396.31 4.31 1.1% 380.58/G4 -qtr 377.14 3.44 0.9% 369.81/F4# 367.89 1.92 0.5% C18 392.00/G4 386.88 5.12 1.3% 380.58/G4 -qtr 363.55 17.03 4.5% 369.81/F4# 371.46 1.65 0.4% C19 392.00/G4 385.16 6.84 1.7% 380.58/G4 -qtr 344.33 36.25 9.5% 369.81/F4# 369.93 0.12 0.0% C20 392.00/G4 318.55 73.45 18.7% 380.58/G4 -qtr 321.26 59.32 15.6% 369.81/F4# 315.51 54.3 14.7% C21 392.00/G4 380.15 11.85 3.0% 380.58/G4 -qtr 365.48 15.1 4.0% 369.81/F4# 363.31 6.5 1.8%
200 Appendix E: (Continued) Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % IN01 261.63/C4 253.71 7.92 3.0% 269.48/C4+qtr 267.6 1.88 0.7% 277.32/C4# 278.28 0.96 0.3% IN02 261.63/C4 250.81 10.82 4.1% 269.48/C4+qtr 257.93 11.55 4.3% 277.32/C4# 275.86 1.46 0.5% IN03 261.63/C4 258.23 3.4 1.3% 269.48/C4+qtr 268.5 0.98 0.4% 277.32/C4# 273.39 3.93 1.4% IN04 261.63/C4 259.67 1.96 0.7% 269.48/C4+qtr 269.81 0.33 0.1% 277.32/C4# 276.6 0.72 0.3% IN05 261.63/C4 262.75 1.12 0.4% 269.48/C4+qtr 274.51 5.03 1.9% 277.32/C4# 282.87 5.55 2.0% IN06 261.63/C4 260.42 1.21 0.5% 269.48/C4+qtr 269.47 0.01 0.0% 277.32/C4# 275.19 2.13 0.8% IN07 261.63/C4 239.41 22.22 8.5% 269.48/C4+qtr 270.41 0.93 0.3% 277.32/C4# 278.7 1.38 0.5% IN08 261.63/C4 260.87 0.76 0.3% 269.48/C4+qtr 273.94 4.46 1.7% 277.32/C4# 277.94 0.62 0.2% IN09 261.63/C4 263.61 1.98 0.8% 269.48/C4+qtr 263.56 5.92 2.2% 277.32/C4# 274.28 3.04 1.1% IN10 261.63/C4 260.47 1.16 0.4% 269.48/C4+qtr 267.53 1.95 0.7% 277.32/C4# 277.15 0.17 0.1% IN11 261.63/C4 254.73 6.9 2.6% 269.48/C4+qtr 248.92 20.56 7.6% 277.32/C4# 250.49 26.83 9.7% IN12 261.63/C4 268.36 6.73 2.6% 269.48/C4+qtr 271.64 2.16 0.8% 277.32/C4# 281.51 4.19 1.5% IN13 261.63/C4 260.77 0.86 0.3% 269.48/C4+qtr 270.17 0.69 0.3% 277.32/C4# 276.78 0.54 0.2% IN14 261.63/C4 266.3 4.67 1.8% 269.48/C4+qtr 271.8 2.32 0.9% 277.32/C4# 281.25 3.93 1.4% IN15 261.63/C4 259.73 1.9 0.7% 269.48/C4+qtr 268.8 0.68 0.3% 277.32/C4# 275.82 1.5 0.5% IN16 261.63/C4 263.52 1.89 0.7% 269.48/C4+qtr 272.69 3.21 1.2% 277.32/C4# 276.57 0.75 0.3% IN17 261.63/C4 259.12 2.51 1.0% 269.48/C4+qtr 266.15 3.33 1.2% 277.32/C4# 276.6 0.72 0.3% IN18 261.63/C4 263.1 1.47 0.6% 269.48/C4+qtr 270.4 0.92 0.3% 277.32/C4# 279.2 1.88 0.7% IN19 261.63/C4 261.09 0.54 0.2% 269.48/C4+qtr 268.95 0.53 0.2% 277.32/C4# 279.97 2.65 1.0% IN20 261.63/C4 254.09 7.54 2.9% 269.48/C4+qtr 271.18 1.7 0.6% 277.32/C4# 268.06 9.26 3.3% IN21 261.63/C4 255.96 5.67 2.2% 269.48/C4+qtr 273.17 3.69 1.4% 277.32/C4# 273.27 4.05 1.5%
201 Appendix E: (Continued) Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % IN01 329.63/E4 344.4 14.77 4.5% 320.03/E4 -qtr 317.78 2.25 0.7% 310.97/D4# 332.46 21.49 6.9% IN02 329.63/E4 321.83 7.8 2.4% 320.03/E4 -qtr 305.52 14.51 4.5% 310.97/D4# 309.38 1.59 0.5% IN03 329.63/E4 324.68 4.95 1.5% 320.03/E4 -qtr 315 5.03 1.6% 310.97/D4# 305.89 5.08 1.6% IN04 329.63/E4 330.71 1.08 0.3% 320.03/E4 -qtr 322.22 2.19 0.7% 310.97/D4# 308.64 2.33 0.7% IN05 329.63/E4 330.08 0.45 0.1% 320.03/E4 -qtr 319.24 0.79 0.2% 310.97/D4# 311.69 0.72 0.2% IN06 329.63/E4 325.38 4.25 1.3% 320.03/E4 -qtr 315.41 4.62 1.4% 310.97/D4# 307.51 3.46 1.1% IN07 329.63/E4 334.2 4.57 1.4% 320.03/E4 -qtr 314.63 5.4 1.7% 310.97/D4# 304.58 6.39 2.1% IN08 329.63/E4 329.89 0.26 0.1% 320.03/E4 -qtr 323.95 3.92 1.2% 310.97/D4# 311.72 0.75 0.2% IN09 329.63/E4 328.42 1.21 0.4% 320.03/E4 -qtr 320.67 0.64 0.2% 310.97/D4# 308.95 2.02 0.6% IN10 329.63/E4 327.99 1.64 0.5% 320.03/E4 -qtr 318.52 1.51 0.5% 310.97/D4# 312.06 1.09 0.4% IN11 329.63/E4 269.7 59.93 18.2% 320.03/E4 -qtr 322.09 2.06 0.6% 310.97/D4# 307.15 3.82 1.2% IN12 329.63/E4 330.96 1.33 0.4% 320.03/E4 -qtr 321.8 1.77 0.6% 310.97/D4# 312.04 1.07 0.3% IN13 329.63/E4 328.89 0.74 0.2% 320.03/E4 -qtr 319.89 0.14 0.0% 310.97/D4# 312.19 1.22 0.4% IN14 329.63/E4 330.48 0.85 0.3% 320.03/E4 -qtr 318.34 1.69 0.5% 310.97/D4# 313.3 2.33 0.7% IN15 329.63/E4 326.84 2.79 0.8% 320.03/E4 -qtr 317.46 2.57 0.8% 310.97/D4# 313.07 2.1 0.7% IN16 329.63/E4 325.26 4.37 1.3% 320.03/E4 -qtr 317.17 2.86 0.9% 310.97/D4# 307.2 3.77 1.2% IN17 329.63/E4 329.18 0.45 0.1% 320.03/E4 -qtr 316.85 3.18 1.0% 310.97/D4# 307.9 3.07 1.0% IN18 329.63/E4 330.67 1.04 0.3% 320.03/E4 -qtr 316.55 3.48 1.1% 310.97/D4# 310.6 0.37 0.1% IN19 329.63/E4 329.48 0.15 0.0% 320.03/E4 -qtr 321.77 1.74 0.5% 310.97/D4# 307.51 3.46 1.1% IN20 329.63/E4 292.86 36.77 11.2% 320.03/E4 -qtr 323.53 3.5 1.1% 310.97/D4# 291.85 19.12 6.1% IN21 329.63/E4 347.67 18.04 5.5% 320.03/E4 -qtr 321.41 1.38 0.4% 310.97/D4# 305.04 5.93 1.9%
202 Appendix E: (Continued) Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % IN01 392.00/G4 391.15 0.85 0.2% 380.58/G4 -qtr 370.48 10.1 2.7% 369.81/F4# 370.62 0.81 0.2% IN02 392.00/G4 385.95 6.05 1.5% 380.58/G4 -qtr 371.7 8.88 2.3% 369.81/F4# 357.44 12.37 3.3% IN03 392.00/G4 382.64 9.36 2.4% 380.58/G4 -qtr 372.66 7.92 2.1% 369.81/F4# 355.87 13.94 3.8% IN04 392.00/G4 391.96 0.04 0.0% 380.58/G4 -qtr 370.65 9.93 2.6% 369.81/F4# 370.36 0.55 0.1% IN05 392.00/G4 392.06 0.06 0.0% 380.58/G4 -qtr 373.44 7.14 1.9% 369.81/F4# 365.86 3.95 1.1% IN06 392.00/G4 387.66 4.34 1.1% 380.58/G4 -qtr 369.69 10.89 2.9% 369.81/F4# 368.7 1.11 0.3% IN07 392.00/G4 364.92 27.08 6.9% 380.58/G4 -qtr 372.23 8.35 2.2% 369.81/F4# 348.76 21.05 5.7% IN08 392.00/G4 393.64 1.64 0.4% 380.58/G4 -qtr 380.34 0.24 0.1% 369.81/F4# 370.46 0.65 0.2% IN09 392.00/G4 389.31 2.69 0.7% 380.58/G4 -qtr 369.7 10.88 2.9% 369.81/F4# 372.62 2.81 0.8% IN10 392.00/G4 391.42 0.58 0.1% 380.58/G4 -qtr 379.09 1.49 0.4% 369.81/F4# 366.54 3.27 0.9% IN11 392.00/G4 388.41 3.59 0.9% 380.58/G4 -qtr 376.51 4.07 1.1% 369.81/F4# 359.37 10.44 2.8% IN12 392.00/G4 382.97 9.03 2.3% 380.58/G4 -qtr 382.98 2.4 0.6% 369.81/F4# 374.31 4.5 1.2% IN13 392.00/G4 390.51 1.49 0.4% 380.58/G4 -qtr 377.88 2.7 0.7% 369.81/F4# 366.68 3.13 0.8% IN14 392.00/G4 381.55 10.45 2.7% 380.58/G4 -qtr 378.67 1.91 0.5% 369.81/F4# 368.26 1.55 0.4% IN15 392.00/G4 394.7 2.7 0.7% 380.58/G4 -qtr 374.14 6.44 1.7% 369.81/F4# 370.42 0.61 0.2% IN16 392.00/G4 390.64 1.36 0.3% 380.58/G4 -qtr 375.84 4.74 1.2% 369.81/F4# 368.2 1.61 0.4% IN17 392.00/G4 390.18 1.82 0.5% 380.58/G4 -qtr 377.55 3.03 0.8% 369.81/F4# 367.52 2.29 0.6% IN18 392.00/G4 392.78 0.78 0.2% 380.58/G4 -qtr 382.59 2.01 0.5% 369.81/F4# 370.26 0.45 0.1% IN19 392.00/G4 387.79 4.21 1.1% 380.58/G4 -qtr 374.25 6.33 1.7% 369.81/F4# 362.5 7.31 2.0% IN20 392.00/G4 397.65 5.65 1.4% 380.58/G4 -qtr 355.64 24.94 6.6% 369.81/F4# 375.1 5.29 1.4% IN21 392.00/G4 379.05 12.95 3.3% 380.58/G4 -qtr 375.43 5.15 1.4% 369.81/F4# 375.75 5.94 1.6%
203 Appendix E: (Continued) Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % V01 261.63/C4 263.28 1.65 0.6% 269.48/C4+qtr 264.31 5.17 1.9% 277.32/C4# 274.09 3.23 1.2% V02 261.63/C4 261.05 0.58 0.2% 269.48/C4+qtr 268.13 1.35 0.5% 277.32/C4# 275.6 1.72 0.6% V03 261.63/C4 261.08 0.55 0.2% 269.48/C4+qtr 262.96 6.52 2.4% 277.32/C4# 275.52 1.8 0.6% V04 261.63/C4 257.32 4.31 1.6% 269.48/C4+qtr 263.94 5.54 2.1% 277.32/C4# 272.88 4.44 1.6% V05 261.63/C4 263.53 1.9 0.7% 269.48/C4+qtr 274.25 4.77 1.8% 277.32/C4# 279.84 2.52 0.9% V06 261.63/C4 258.89 2.74 1.0% 269.48/C4+qtr 269.47 0.01 0.0% 277.32/C4# 277.31 0.01 0.0% V07 261.63/C4 259.26 2.37 0.9% 269.48/C4+qtr 268.75 0.73 0.3% 277.32/C4# 274.46 2.86 1.0% V08 261.63/C4 258.8 2.83 1.1% 269.48/C4+qtr 275.57 6.09 2.3% 277.32/C4# 276.76 0.56 0.2% V09 261.63/C4 281.97 20.34 7.8% 269.48/C4+qtr 273.35 3.87 1.4% 277.32/C4# 276.14 1.18 0.4% V10 261.63/C4 258.65 2.98 1.1% 269.48/C4+qtr 265.73 3.75 1.4% 277.32/C4# 275.83 1.49 0.5% V11 261.63/C4 261.01 0.62 0.2% 269.48/C4+qtr 270.14 0.66 0.2% 277.32/C4# 277.28 0.04 0.0% V12 261.63/C4 258.1 3.53 1.3% 269.48/C4+qtr 268.03 1.45 0.5% 277.32/C4# 272.95 4.37 1.6% V13 261.63/C4 257.88 3.75 1.4% 269.48/C4+qtr 272.23 2.75 1.0% 277.32/C4# 276.01 1.31 0.5% V14 261.63/C4 255.35 6.28 2.4% 269.48/C4+qtr 271.27 1.79 0.7% 277.32/C4# 271.79 5.53 2.0% V15 261.63/C4 255.92 5.71 2.2% 269.48/C4+qtr 274.18 4.7 1.7% 277.32/C4# 279.43 2.11 0.8% V16 261.63/C4 260.57 1.06 0.4% 269.48/C4+qtr 268.86 0.62 0.2% 277.32/C4# 275 2.32 0.8% V17 261.63/C4 263.99 2.36 0.9% 269.48/C4+qtr 267.06 2.42 0.9% 277.32/C4# 278.32 1 0.4% V18 261.63/C4 259.46 2.17 0.8% 269.48/C4+qtr 266.73 2.75 1.0% 277.32/C4# 273.75 3.57 1.3% V19 261.63/C4 268.46 6.83 2.6% 269.48/C4+qtr 276.24 6.76 2.5% 277.32/C4# 271.45 5.87 2.1% V20 261.63/C4 260.64 0.99 0.4% 269.48/C4+qtr 273.34 3.86 1.4% 277.32/C4# 277.34 0.02 0.0%
204 Appendix E: (Continued) Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % V01 329.63/E4 327.65 1.98 0.6% 320.03/E4 -qtr 322.21 2.18 0.7% 310.97/D4# 309.39 1.58 0.5% V02 329.63/E4 327.22 2.41 0.7% 320.03/E4 -qtr 318.11 1.92 0.6% 310.97/D4# 312.1 1.13 0.4% V03 329.63/E4 330.8 1.17 0.4% 320.03/E4 -qtr 316.52 3.51 1.1% 310.97/D4# 308.83 2.14 0.7% V04 329.63/E4 326.94 2.69 0.8% 320.03/E4 -qtr 316.59 3.44 1.1% 310.97/D4# 307.47 3.5 1.1% V05 329.63/E4 328.5 1.13 0.3% 320.03/E4 -qtr 314.64 5.39 1.7% 310.97/D4# 310.38 0.59 0.2% V06 329.63/E4 332.04 2.41 0.7% 320.03/E4 -qtr 318.06 1.97 0.6% 310.97/D4# 310.96 0.01 0.0% V07 329.63/E4 324.25 5.38 1.6% 320.03/E4 -qtr 317.42 2.61 0.8% 310.97/D4# 309.03 1.94 0.6% V08 329.63/E4 333.21 3.58 1.1% 320.03/E4 -qtr 318.19 1.84 0.6% 310.97/D4# 315.01 4.04 1.3% V09 329.63/E4 333.32 3.69 1.1% 320.03/E4 -qtr 318.5 1.53 0.5% 310.97/D4# 314.68 3.71 1.2% V10 329.63/E4 325.22 4.41 1.3% 320.03/E4 -qtr 316.47 3.56 1.1% 310.97/D4# 309.54 1.43 0.5% V11 329.63/E4 329.06 0.57 0.2% 320.03/E4 -qtr 319.44 0.59 0.2% 310.97/D4# 312.43 1.46 0.5% V12 329.63/E4 323.63 6 1.8% 320.03/E4 -qtr 310.44 9.59 3.0% 310.97/D4# 310.26 0.71 0.2% V13 329.63/E4 326.75 2.88 0.9% 320.03/E4 -qtr 321.04 1.01 0.3% 310.97/D4# 314.22 3.25 1.0% V14 329.63/E4 323.87 5.76 1.7% 320.03/E4 -qtr 319.49 0.54 0.2% 310.97/D4# 307.24 3.73 1.2% V15 329.63/E4 328.87 0.76 0.2% 320.03/E4 -qtr 323.95 3.92 1.2% 310.97/D4# 311.54 0.57 0.2% V16 329.63/E4 324.87 4.76 1.4% 320.03/E4 -qtr 315.99 4.04 1.3% 310.97/D4# 306.6 4.37 1.4% V17 329.63/E4 328.87 0.76 0.2% 320.03/E4 -qtr 318.84 1.19 0.4% 310.97/D4# 313.85 2.88 0.9% V18 329.63/E4 327.6 2.03 0.6% 320.03/E4 -qtr 316.09 3.94 1.2% 310.97/D4# 307.68 3.29 1.1% V19 329.63/E4 324.95 4.68 1.4% 320.03/E4 -qtr 317.54 2.49 0.8% 310.97/D4# 313.54 2.57 0.8% V20 329.63/E4 327.63 2 0.6% 320.03/E4 -qtr 316.85 3.18 1.0% 310.97/D4# 309.66 1.31 0.4%
205 Appendix E: (Continued) Subj Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % Stimulus Prod Hz Diff Hz Diff % V01 392.00/G4 388.05 3.95 1.0% 380.58/G4-qtr 374.8 5.78 1.5% 369.81/F4# 368.85 0.96 0.3% V02 392.00/G4 390.81 1.19 0.3% 380.58/G4-qtr 375.32 5.26 1.4% 369.81/F4# 370.96 1.15 0.3% V03 392.00/G4 392.65 0.65 0.2% 380.58/G4-qtr 373.59 6.99 1.8% 369.81/F4# 370.32 0.51 0.1% V04 392.00/G4 388.15 3.85 1.0% 380.58/G4-qtr 371.13 9.45 2.5% 369.81/F4# 367.89 1.92 0.5% V05 392.00/G4 387.09 4.91 1.3% 380.58/G4-qtr 369.41 11.17 2.9% 369.81/F4# 370.55 0.74 0.2% V06 392.00/G4 382.77 9.23 2.4% 380.58/G4-qtr 371.44 9.14 2.4% 369.81/F4# 363.75 6.06 1.6% V07 392.00/G4 392.06 0.06 0.0% 380.58/G4-qtr 375.67 4.91 1.3% 369.81/F4# 367.45 2.36 0.6% V08 392.00/G4 394.15 2.15 0.5% 380.58/G4-qtr 378.8 1.78 0.5% 369.81/F4# 364.95 4.86 1.3% V09 392.00/G4 393.57 1.57 0.4% 380.58/G4-qtr 380.62 0.04 0.0% 369.81/F4# 368.44 1.37 0.4% V10 392.00/G4 390.19 1.81 0.5% 380.58/G4-qtr 371.4 9.18 2.4% 369.81/F4# 362.54 7.27 2.0% V11 392.00/G4 388.48 3.52 0.9% 380.58/G4-qtr 377.54 3.04 0.8% 369.81/F4# 363.13 6.68 1.8% V12 392.00/G4 386.63 5.37 1.4% 380.58/G4-qtr 369.74 10.84 2.8% 369.81/F4# 363.96 5.85 1.6% V13 392.00/G4 389.53 2.47 0.6% 380.58/G4-qtr 373.1 7.48 2.0% 369.81/F4# 365.64 4.17 1.1% V14 392.00/G4 385.89 6.11 1.6% 380.58/G4-qtr 384.23 3.65 1.0% 369.81/F4# 366.08 3.73 1.0% V15 392.00/G4 389.49 2.51 0.6% 380.58/G4-qtr 372.33 8.25 2.2% 369.81/F4# 371.52 1.71 0.5% V16 392.00/G4 387.7 4.3 1.1% 380.58/G4-qtr 373.24 7.34 1.9% 369.81/F4# 366.32 3.49 0.9% V17 392.00/G4 389.25 2.75 0.7% 380.58/G4-qtr 374.8 5.78 1.5% 369.81/F4# 365.87 3.94 1.1% V18 392.00/G4 386.41 5.59 1.4% 380.58/G4-qtr 371.61 8.97 2.4% 369.81/F4# 363.22 6.59 1.8% V19 392.00/G4 390.75 1.25 0.3% 380.58/G4-qtr 383.81 3.23 0.8% 369.81/F4# 368.2 1.61 0.4% V20 392.00/G4 389.5 2.5 0.6% 380.58/G4-qtr 375.56 5.02 1.3% 369.81/F4# 363.74 6.07 1.6%
206 Appendix F: Individu al Data for Difference Limen for Frequency GROUP SUBJ STIM DLF Hz DLF % STIM DLF Hz DLF% STIM DLF Hz DLF% CONTROL C01 261.63/C4 10.43 3.99% 329.63/E4 6.65 2.02% 392.00/G4 5.34 1.36% CONTROL C02 261.63/C4 4.59 1.75% 329.63/E4 5.25 1.59% 392.00/G4 5.5 1.40% CONTROL C03 261.63/C4 7.66 2.93% 329.63/E4 9.54 2.89% 392.00/G4 13.32 3.40% CONTROL C04 261.63/C4 5.83 2.23% 329.63/E4 5.51 1.67% 392.00/G4 9.64 2.46% CONTROL C05 261.63/C4 11.33 4.33% 329.63/E4 32.96 10.00% 392.00/G4 10.99 2.80% CONTROL C06 261.63/C4 31.28 11.96% 329.63/E4 35.56 10.79% 392.00/G4 44.97 11.47% CONTROL C07 261.63/C4 4.92 1.88% 329.63/E4 9.01 2.73% 392.00/G4 5.82 1.48% CONTROL C08 261.63/C4 3.93 1.50% 329.63/E4 4.42 1.34% 392.00/G4 5.02 1.28% CONTROL C09 261.63/C4 7.11 2.72% 329.63/E4 5.24 1.59% 392.00/G4 7.19 1.83% CONTROL C10 261.63/C4 2.11 0.81% 329.63/E4 4.62 1.40% 392.00/G4 5.67 1.45% CONTROL C11 261.63/C4 3.26 1.25% 329.63/E4 4.01 1.22% 392.00/G4 4.45 1.14% CONTROL C12 261.63/C4 8.67 3.31% 329.63/E4 18.76 5.69% 392.00/G4 5.99 1.53% CONTROL C13 261.63/C4 35.85 13.70% 329.63/E4 57.35 17.40% 392.00/G4 48.08 12.27% CONTROL C14 261.63/C4 3.1 1.18% 329.63/E4 3.74 1.13% 392.00/G4 5.5 1.40% CONTROL C15 261.63/C4 6.25 2.39% 329.63/E4 5.24 1.59% 392.00/G4 7.44 1.90% CONTROL C16 261.63/C4 17.32 6.62% 329.63/E4 5.17 1.57% 392.00/G4 5.58 1.42% CONTROL C17 261.63/C4 3.7 1.41% 329.63/E4 4.63 1.40% 392.00/G4 5.1 1.30% CONTROL C18 261.63/C4 3.96 1.51% 329.63/E4 5.56 1.69% 392.00/G4 3.89 0.99% CONTROL C19 261.63/C4 5.8 2.22% 329.63/E4 6.78 2.06% 392.00/G4 5.18 1.32% CONTROL C20 261.63/C4 8.76 3.35% 329.63/E4 7.13 2.16% 392.00/G4 5.5 1.40% CONTROL C21 261.63/C4 4.14 1.58% 329.63/E4 4.7 1.43% 392.00/G4 4.78 1.22%
207 Appendix F: (Continued) GROUP SUBJ STIM DLF Hz DLF % STIM DLF Hz DLF% STIM DLF Hz DLF% INSTRUM IN01 261.63/C4 4.47 1.71% 329.63/E4 4.96 1.50% 392.00/G4 6.07 1.55% INSTRUM IN02 261.63/C4 3.54 1.35% 329.63/E4 7.79 2.36% 392.00/G4 6.47 1.65% INSTRUM IN03 261.63/C4 6.41 2.45% 329.63/E4 13.07 3.97% 392.00/G4 5.83 1.49% INSTRUM IN04 261.63/C4 2.38 0.91% 329.63/E4 4.15 1.26% 392.00/G4 5.58 1.42% INSTRUM IN05 261.63/C4 2.77 1.06% 329.63/E4 4.42 1.34% 392.00/G4 4.45 1.14% INSTRUM IN06 261.63/C4 4.42 1.69% 329.63/E4 4.08 1.24% 392.00/G4 5.1 1.30% INSTRUM IN07 261.63/C4 4.58 1.75% 329.63/E4 4.83 1.47% 392.00/G4 6.71 1.71% INSTRUM IN08 261.63/C4 3.26 1.25% 329.63/E4 5.51 1.67% 392.00/G4 5.1 1.30% INSTRUM IN09 261.63/C4 4.2 1.61% 329.63/E4 3.33 1.01% 392.00/G4 5.5 1.40% INSTRUM IN10 261.63/C4 1.89 0.72% 329.63/E4 3.34 1.01% 392.00/G4 3.88 0.99% INSTRUM IN11 261.63/C4 3.92 1.50% 329.63/E4 3.68 1.12% 392.00/G4 4.29 1.09% INSTRUM IN12 261.63/C4 3.48 1.33% 329.63/E4 3.4 1.03% 392.00/G4 3.89 0.99% INSTRUM IN13 261.63/C4 2.71 1.04% 329.63/E4 3.68 1.12% 392.00/G4 3.89 0.99% INSTRUM IN14 261.63/C4 3.48 1.33% 329.63/E4 3.75 1.14% 392.00/G4 4.62 1.18% INSTRUM IN15 261.63/C4 4.36 1.67% 329.63/E4 4.7 1.43% 392.00/G4 5.18 1.32% INSTRUM IN16 261.63/C4 1.95 0.75% 329.63/E4 2.93 0.89% 392.00/G4 3.72 0.95% INSTRUM IN17 261.63/C4 3.48 1.33% 329.63/E4 3.95 1.20% 392.00/G4 5.02 1.28% INSTRUM IN18 261.63/C4 4.75 1.82% 329.63/E4 4.42 1.34% 392.00/G4 6.71 1.71% INSTRUM IN19 261.63/C4 2.33 0.89% 329.63/E4 4.29 1.30% 392.00/G4 3.97 1.01% INSTRUM IN20 261.63/C4 6.55 2.50% 329.63/E4 6.85 2.08% 392.00/G4 5.1 1.30% INSTRUM IN21 261.63/C4 3.76 1.44% 329.63/E4 5.3 1.61% 392.00/G4 4.29 1.09%
208 Appendix F: (Continued) GROUP SUBJ STIM DLF Hz DLF % STIM DLF Hz DLF% STIM DLF Hz DLF% VOCAL V01 261.63/C4 2.82 1.08% 329.63/E4 3.4 1.03% 392.00/G4 4.86 1.24% VOCAL V02 261.63/C4 4.36 1.67% 329.63/E4 3.74 1.13% 392.00/G4 4.61 1.18% VOCAL V03 261.63/C4 3.87 1.48% 329.63/E4 3.54 1.07% 392.00/G4 4.94 1.26% VOCAL V04 261.63/C4 3.31 1.27% 329.63/E4 3.68 1.12% 392.00/G4 3.81 0.97% VOCAL V05 261.63/C4 4.53 1.73% 329.63/E4 11.95 3.63% 392.00/G4 11.29 2.88% VOCAL V06 261.63/C4 3.47 1.33% 329.63/E4 4.43 1.34% 392.00/G4 4.86 1.24% VOCAL V07 261.63/C4 2.66 1.02% 329.63/E4 4.22 1.28% 392.00/G4 5.42 1.38% VOCAL V08 261.63/C4 3.32 1.27% 329.63/E4 4.29 1.30% 392.00/G4 5.1 1.30% VOCAL V09 261.63/C4 2.38 0.91% 329.63/E4 4.13 1.25% 392.00/G4 4.86 1.24% VOCAL V10 261.63/C4 4.64 1.77% 329.63/E4 4.89 1.48% 392.00/G4 4.53 1.16% VOCAL V11 261.63/C4 3.42 1.31% 329.63/E4 4.08 1.24% 392.00/G4 4.94 1.26% VOCAL V12 261.63/C4 3.98 1.52% 329.63/E4 5.22 1.58% 392.00/G4 4.37 1.11% VOCAL V13 261.63/C4 3.04 1.16% 329.63/E4 4.22 1.28% 392.00/G4 4.78 1.22% VOCAL V14 261.63/C4 3.32 1.27% 329.63/E4 3.61 1.10% 392.00/G4 4.86 1.24% VOCAL V15 261.63/C4 2.71 1.04% 329.63/E4 4.29 1.30% 392.00/G4 5.02 1.28% VOCAL V16 261.63/C4 4.14 1.58% 329.63/E4 3.54 1.07% 392.00/G4 4.37 1.11% VOCAL V17 261.63/C4 3.15 1.20% 329.63/E4 3.88 1.18% 392.00/G4 5.1 1.30% VOCAL V18 261.63/C4 2.66 1.02% 329.63/E4 3.47 1.05% 392.00/G4 5.34 1.36% VOCAL V19 261.63/C4 6.92 2.64% 329.63/E4 5.37 1.63% 392.00/G4 9.68 2.47% VOCAL V20 261.63/C4 2.77 1.06% 329.63/E4 4.56 1.38% 392.00/G4 5.1 1.30%
209 Appendix G: P1 Individual Peak Amplitude at Fz *V05 did not meet inclusion criteria Summary Data GROUP Mean Dev1/Amp Std Error Mean Dev2/Amp Std Error Mean Dev3/Amp Std Error Control 3.79 1.46 3.68 1.14 3.96 1.51 Musician 3.08 1.21 3.21 1.01 3.00 1.16 Vocal 3.24 1.12 3.22 1.04 3.13 1.16 Instrumental 2.93 1.30 3.20 1.14 2.87 1.17 Grand Mean 3.32 3.37 3.32 Group Dev 1 Amp Dev 2 Amp Dev 3 Amp Group Dev 1 Amp Dev 2 Amp Dev 3 Amp Group Dev 1 Amp Dev 2 Amp Dev 3 Amp C01 4.03 6.12 5.16 IN01 3.92 3.09 2.71 V01 5.71 4.33 5.69 C02 3.72 4.36 7.19 IN02 1.89 2.8 2.19 V02 3.83 3.01 2.49 C03 3.74 4.38 2.86 IN03 3.84 3.56 2.27 V03 4.28 3.94 3.62 C04 5.71 5.26 4.93 IN04 2.87 3.56 4.3 V04 4.23 3 4.11 C05 2.52 3.23 2.35 IN05 1.48 3.55 2.5 V05* C06 3.64 3.04 3.5 IN06 3.14 2.53 2.21 V06 4.2 4.37 4.51 C07 4.9 4.16 5.46 IN07 1.96 2.22 0.75 V07 1.45 2.2 1.8 C08 6.71 5.52 5.54 IN08 2.28 3.67 2.27 V08 1.86 4.7 0.9 C09 2.62 2.36 1.6 IN09 1.87 2.17 3.12 V09 2.66 4.04 3.14 C10 0.19 1.55 2.28 IN10 5.43 3.5 3.5 V10 3.16 4.02 4.3 C11 2.64 2.52 3.29 IN11 0.43 0.56 0.48 V11 2.11 1.64 1.41 C12 2.56 2.83 2.33 IN12 1.96 2.24 3.2 V12 2.64 2.81 3.24 C13 4.59 3.51 5.2 IN13 4.2 3.67 5.17 V13 4.31 4.99 3.57 C15 2.2 2.61 2.35 IN14 3.64 4.04 2.47 V14 2.31 2.73 3.64 C16 4.79 3.53 5.16 IN16 4.64 5.22 2.87 V15 3.48 1.62 1.65 C17 4.97 4.17 5.41 IN17 2.89 2.24 2.34 V16 2.93 2.25 2.58 C18 5.06 3.67 4.66 IN18 2.41 1.98 2.81 V17 4.18 3.36 3.57 C19 2.95 3.37 3.04 IN19 5.15 5.36 5.04 V18 1.79 2.58 2.72 C20 3.97 2.95 4.09 IN20 1.91 4.03 3.5 V19 4.15 4.11 3.96 C21 4.26 4.42 2.84 IN21 2.72 4.08 3.74 V20 3.41 2.76 3.08
210 Appendix H: P1 Individual Peak Latency at Fz *V05 did not meet inclusion criteria Summary Data GROUP Mean Dev 1/Amp Std Error Mean Dev 2/Amp Std Error Mean Dev 3/Amp Std Error Control -1.52 1.25 -1.99 1.69 -2.46 1.24 Musician -1.94 1.53 -1.67 1.11 -1.87 1.69 Vocal -2.26 1.51 -1.42 1.03 -1.92 1.43 Instrumental -1.61 1.51 -1.92 1.16 -1.82 1.96 Grand Mean -1.80 -1.78 -2.07 Group Dev 1 Amp Dev 2 Amp Dev 3 Amp Group Dev 1 Amp Dev 2 Amp Dev 3 Amp Group Dev 1 Amp Dev 2 Amp Dev 3 Amp C01 -1.12 -1.15 -0.91 IN01 -3.83 -4.41 -3.24 V01 -4.64 -2.7 -2.94 C02 -2.37 -1.09 -4.53 IN02 -1.44 -0.14 0.11 V02 -3.78 -2.57 -2.91 C03 -0.99 -0.83 -1.98 IN03 -1.06 -1.16 0.95 V03 -0.79 -0.5 -2.5 C04 -2.43 -2.93 -3.54 IN04 -2.03 -0.92 -1.19 V04 -2.39 -0.56 -4.82 C05 -1.04 -2.6 -3.07 IN05 -2.34 -3.49 -3.58 V05* C06 -0.98 -2.8 -2.93 IN06 -2.19 -1.42 -3.47 V06 -0.36 -1.47 -2.76 C07 -3.61 -3.5 -2.18 IN07 0.34 -0.04 -3.17 V07 -1.74 -0.82 -2.17 C08 -1.19 -3.79 -3.36 IN08 0.33 -2.01 -3.27 V08 -2.34 -2.15 -2.3 C09 0.75 1.55 -0.34 IN09 -0.67 -2.59 -2.72 V09 -3.26 -2.34 0.26 C10 -3.25 -0.58 -2.05 IN10 -1 -3.05 -6.25 V10 -2.62 -2.39 -1.65 C11 -1.36 -2.6 -2.55 IN11 0.02 -0.31 -0.1 V11 -3.7 -1.42 -2.75 C12 -0.09 -2.03 -3.13 IN12 -2.42 -2.23 -2.16 V12 -3.08 -2.13 -2.6 C13 -1.46 -4.21 -3.13 IN13 -2.81 -1.96 -0.09 V13 -3.42 0.43 1.38 C15 -1.24 -4.79 -5.1 IN14 -1.85 -2.69 -2.78 V14 -2.05 -2.97 -2.39 C16 -1.34 -2.56 -2.29 IN16 0.55 -1.88 -2.7 V15 0.51 0.05 -1.5 C17 -3.55 -3.51 -2.26 IN17 -0.95 -3.07 0.17 V16 -4.8 -1.99 -3.57 C18 0.64 -1.61 -2.19 IN18 -1.17 -1.78 2.13 V17 -1.61 -1.12 -2.09 C19 -3.4 1.1 -0.68 IN19 -3.25 -1.72 -1.5 V18 -1.88 0.37 0.23 C20 -1.02 0.07 -0.51 IN20 -0.98 -0.84 -2.51 V19 -1.26 -1.6 -1.6 C21 -1.38 -1.97 -2.48 IN21 -5.47 -2.77 -0.99 V20 -2.64 -1.82 -1.06
211 Appendix I: Mismatch Negativity Indivi dual Peak Amplitude at Fz *V05 did not meet inclusion criteria Summary Data GROUP Mean Dev 1/Lat Std Error Mean Dev 2/Lat Std Error Mean Dev 3/Lat Std Error Control 75.30 8.35 79.00 7.24 77.50 7.22 Musician 78.73 8.38 80.07 10.11 80.20 10.44 Vocal 79.30 9.84 79.75 9.27 80.65 8.46 Instrumental 78.15 6.82 80.40 11.12 79.75 12.32 Grand Mean 77.58 79.72 79.30 Group Dev 1 Lat Dev 2 Lat Dev 3 Lat Group Dev 1 Lat Dev 2 Lat Dev 3 Lat Group Dev 1 Lat Dev 2 Lat Dev 3 Lat C01 69 70 88 IN01 67 68 70 V01 76 74 74 C02 80 82 84 IN02 80 92 70 V02 80 103 84 C03 74 76 72 IN03 82 80 90 V03 74 82 84 C04 76 76 74 IN04 76 68 67 V04 82 82 84 C05 76 70 67 IN05 72 70 68 V05* C06 78 82 82 IN06 80 82 82 V06 84 86 82 C07 82 96 78 IN07 78 80 78 V07 100 94 98 C08 74 72 72 IN08 68 70 70 V08 100 84 80 C09 72 72 80 IN09 86 88 90 V09 72 70 68 C10 74 84 78 IN10 72 76 72 V10 80 78 80 C11 70 74 63 IN11 80 86 70 V11 82 82 82 C12 82 86 86 IN12 90 84 94 V12 65 70 65 C13 78 80 82 IN13 78 78 84 V13 74 74 74 C15 47 74 72 IN14 72 67 63 V14 72 72 74 C16 88 82 92 IN16 76 76 76 V15 68 68 72 C17 82 92 78 IN17 68 70 70 V16 94 94 90 C18 72 72 72 IN18 82 88 92 V17 80 80 78 C19 70 74 72 IN19 82 80 92 V18 80 80 84 C20 80 84 82 IN20 90 92 86 V19 82 78 82 C21 82 82 76 IN21 84 113 111 V20 76 76 80
212 Appendix J: Mismatch Negativity Individua l Peak Latency at Fz *V05 Did not meet inclusion criteria Summary Data GROUP Mean Dev 1/Lat Std Error Mean Dev 2/Lat Std Error Mean Dev 3/Lat Std Error Control 231.15 31.09 204.45 25.13 189.45 19.03 Musician 217.87 27.45 192.00 22.57 184.42 26.37 Vocal 216.35 27.55 192.80 26.15 184.90 32.24 Instrumental 219.40 27.98 191.20 18.98 183.95 19.70 Grand Mean 222.30 196.15 186.10 Group Dev 1 Lat Dev 2 Lat Dev 3 Lat Group Dev 1 Lat Dev 2 Lat Dev 3 Lat Group Dev 1 Lat Dev 2 Lat Dev 3 Lat C01 200 182 188 IN01 204 197 189 V01 230 197 175 C02 270 219 179 IN02 261 206 208 V02 189 156 167 C03 276 231 212 IN03 261 197 177 V03 197 216 173 C04 222 206 171 IN04 226 177 164 V04 200 197 189 C05 190 168 185 IN05 220 169 187 V05* C06 231 185 198 IN06 233 206 175 V06 245 183 202 C07 236 220 201 IN07 173 185 177 V07 237 228 189 C08 193 199 159 IN08 228 212 202 V08 267 216 179 C09 267 232 203 IN09 214 173 164 V09 235 185 166 C10 218 168 187 IN10 210 162 179 V10 206 208 197 C11 283 206 203 IN11 257 167 179 V11 158 150 146 C12 263 250 187 IN12 255 228 210 V12 195 191 185 C13 219 201 161 IN13 239 177 202 V13 230 235 241 C15 182 197 183 IN14 179 197 152 V14 228 164 142 C16 259 241 222 IN16 187 222 199 V15 195 202 195 C17 237 220 200 IN17 179 206 224 V16 202 160 166 C18 193 177 177 IN18 210 191 152 V17 197 210 152 C19 237 166 156 IN19 224 191 175 V18 212 160 171 C20 208 195 195 IN20 237 166 197 V19 218 166 167 C21 239 226 222 IN21 191 195 167 V20 212 208 218
About the Author Dee Adams Nikjeh received a Bachel ors Degree in Speech Pathology and Audiology in 1978 from West Virginia Universi ty where she also received her Master of Science Degree in Speech Pathology in 1979. Sh e has more than 25 years of clinical experience focusing in neurogenic communi cation disorders and has given numerous professional presentations. For the past 11 ye ars, Mrs. Nikjeh has been employed by an otolaryngology practice where she specializ es in endoscopic procedures for the evaluation and management of voice and swallowing disorders. She presently serves on the Amer ican Speech Language and Hearing Associations (ASHA) Health Care a nd Economics Committee and has extensive knowledge of national reimbursement and codi ng issues as they pertain to speechlanguage pathologists and audi ologists. Mrs. Nikjeh has served on ASHAs Legislative Council, the Executive Board of the Council of St ate Association Presidents and is a pastpresident of the Florida Asso ciation of Speech Language Pat hologists and Audiologists.