Abstract
The purpose of this study was to compare the muscle activation of singers and instrumentalists while performing simple vocal exercises. Volunteer participants (N = 28) were undergraduate music majors and minors, with an equal number being vocalists and instrumentalists. Participants performed five vowel sounds (ah, eh, ee, oh, oo), while electromyography of the zygomaticus and masseter muscles was sampled at 1,000 Hz. A statistically significant multivariate analysis of variance effect was obtained and follow-up analyses of variance showed instrumentalists had more masseter muscle activation than vocalists when performing “eh” and “ee.” Instrumentalists also had more zygomaticus muscle activation than vocalists when performing the “eh” vowel, but vocalists had more zygomaticus muscle activation when performing the “ah” vowel.
Vocal technique can be defined as “the specific learned process through which a singer controls the coordination of the four physiologic systems (respiration, phonation, articulation, and resonation)” (Bickel, 2008, p. 1). Virtually all vocal pedagogues position relaxation, or lack of tension, as an integral part of vocal technique (Bauer, 2013; Bickel, 2008; Davids & LaTour, 2012; LaPine, 2008; Large, 1980; Miller, 1996; Vennard, 1967; White, 1976; Wilson, 1991). Vocalists are trained to sing with relaxed posture, throats, and jaw and facial muscles, primarily to avoid vocal damage and fatigue. Kitch and Oates (1994) described several visual indicators of excessive vocal tension, which included head tilting, jaw tightness, and overly active facial muscles. Miller (1996) asserted that the tongue, neck, and jaw are three major sources of vocal tension and that tension in a particular muscle group can affect other muscle groups. Miller also suggested that the neutral position of the jaw, with teeth slightly parted, occurs unless one consciously relaxes the mandible to allow the jaw to “hang” or consciously forces the jaw shut. Tension in the masseter muscles (i.e., the muscles responsible for closing the jaw) will cause the jaw to clench or tighten, which can lengthen the vocal tract and negatively alter the vocal tone quality. Additionally, ongoing tension in the masseter muscles can lead to issues with the temporomandibular joint (TMJ) and cause headaches.
According to McQuade (2016), the zygomaticus muscles are also crucial in producing a resonant sound. The zygomaticus major is a muscle of facial expression that draws the angle of the mouth superiorly (up) and posteriorly (back). The zygomaticus muscles are a common site for chronic tension, “which shortens the vocal tract and reduces resonance” (DeVore & Cookman, 2009, p. 179). Miller (1996) posited that the position of the zygomatic muscles can alter the shape of the vocal tract, which would affect vocal timbre and amplitude.
In stark contrast to vocalists, who focus on muscle relaxation as a key component of good vocal production, wind instrumentalists intentionally train muscles to activate on command and form an embouchure. Frei and Truong (2017) defined wind instrument embouchure as “the set pattern of perioral and jaw muscles used to initiate and control the amplitude and force of airflow in the mouthpiece of a woodwind or brass instrument” (pp. 377–378). Without activation from various muscles of the face, wind instrumentalists are unable to sound their instruments appropriately. In this way, activation of facial muscles is integral to good tone production in wind instrumentalists. However, overuse of these muscles may contribute to performance-related health issues, such as focal dystonia (Lederman, 1991; Schuele & Lederman, 2003), muscle fatigue, dental problems, lip irritations (Howard & Lovrovich, 1989), tightness of the lips and facial muscles, TMJ, and jaw pain (Taddey, 1992), as well as pain in the neck, hands, shoulder, and back (Chesky et al., 2002). Some woodwind instructors use vowels to teach vocal tract configurations with the goal of achieving better sound quality, especially in various registers of the instrument (González & Payri, 2017). The literature in this area has focused on the larynx and tongue position used to achieve each vowel, but appears to ignore the potential impact of facial muscles, like the zygomaticus, that may also be altered based on the vowel used. If facial muscle activation contributes to problems with playing wind instruments, it is possible that it may also contribute to problems with vocal production.
In recent years, the desire of music education researchers to understand the physiology of musicians has led to the increased use of electromyography (EMG). EMG uses electrodes to record action potentials, or electrical activity, of a muscle, thereby allowing quantification of muscle activation levels. Researchers utilizing EMG to study vocalists have focused primarily on the larynx in relationship to vocal health (Benninger et al., 1994; Colton & Casper, 1996; Ludlow, 1991; Sundberg, 1987). Additionally, researchers have used surface EMG to measure neck and shoulder muscle activation while singing and speaking (Manternach, 2016; Pettersen et al., 2005; Pettersen & Westgaard, 2002, 2004a, 2004b, 2005). Russo et al. (2009) used surface EMG to examine activity of the zygomaticus and corrugator cilli (frowning muscle) in musically trained individuals mimicking emotional ideas. Few researchers have used surface EMG to examine facial muscle activity during singing.
Due in part to noninvasiveness, surface EMG also has been used extensively to assess muscle activation in instrumentalists, including muscle activation patterns in the arm musculature of string players (Guettler, 1992; Thiem et al., 1994), the embouchures of clarinetists (Belcher, 2004), and the embouchures of trumpet players (Heuser & McNitt-Gray, 1991, 1993, 1998; Iltis & Givens, 2005; E. White & Basmajian, 1974). Gossett (1989) used EMG to measure the abdominal and suprahyoid muscle activation of individuals concurrently studying voice and oboe and concluded that there were changes in muscle activity of both the abdominal and suprahyoid. The author speculated these changes may be attributed to the tight lip embouchure and closed jaw position required for producing a quality sound on the oboe.
In addition to the study of muscle activation, EMG has been used for purposes of enhancing performance via biofeedback in both vocalists and instrumentalists. Kirkpatrick and McLester (2012) examined the use of EMG biofeedback in teaching lower laryngeal position in singers. Several research studies have involved the examination of EMG biofeedback and its use in reducing muscular tension, including the trapezius in string players (Cutietta, 1986), facial and throat muscles in clarinetists (Levee et al., 1976; Morasky et al., 1983), and arm extensor muscles in a string player (Reynolds et al., 1981). According to these researchers, EMG biofeedback reduces muscle tension and yields some improvement in overall performance.
The pedagogical practice of vocalization and singing parts in band has become fairly prominent in recent years, especially because vocalization aids students in developing a consistent sense of tonality and intonation (Robinson, 1996; Wolbers, 2002) and also encourages higher level thinking skills (Jarvis, 1980). Some band directors are also called on to teach general music or choir. For these reasons, many music teacher education programs require instrumental students to take a voice lesson (typically class voice) and/or participate in a choral ensemble. Research is scant on the unique issues that instrumentalists may face when singing that could be attributed to the facial muscle behaviors required to play their respective instrument.
Researchers have demonstrated that facial muscle activation can change in response to emotional facial expressions in others (Dimberg et al., 2000; Lundqvist, 1995), as well as affective imagery (Schwartz et al., 1976; Schwartz et al., 1980; Vanman et al., 2004). If such responses are elicited by simple facial expressions and images, it stands to reason that music students may experience similar phenomena based on the actions of a teacher or conductor. Recent research on mimicking behaviors of singers in response to nonverbal conductor behaviors may have implications for music teacher education. Fuelberth (2004) suggested that left-hand conducting gestures can have a significant effect on “inappropriate vocal tension” in singers, and Daugherty and Brunkan (2013) found that singers adjust their lip postures in response to lip rounding by a conductor. Manternach (2012a) reported similar findings related to lip rounding with the addition of singers increasing their eyebrow height in response to the conductor’s eyebrow raise. Manternach (2012b) also concluded certain conducting behaviors could affect singers’ head and shoulder movement.
This begs the questions: If students are responding to or mimicking intentional actions of a conductor or teacher in such a profound way, how are they responding to unintentional actions of a conductor or teacher? Do students perceive and replicate “vocal tension” during teacher modeling? Many musicians, both vocal and instrumental, find themselves in situations in which they must act as vocal models. Before these questions can be answered, research comparing facial muscle activation of instrumentalists and vocalists is needed. Training differences for wind instrumentalists and vocalists may influence vocal production while singing or modeling, and music teacher education programs may need to incorporate or amend vocal training in the curriculum based on these possible differences. The purpose of this study was to compare the facial muscle activation of singers and instrumentalists while performing simple vocal exercises. The zygomaticus and masseter muscles were investigated due to their superficial location, as well as their function.
Method
Participants
Volunteer participants (N = 28) consisted of male (n = 19) and female (n = 9) undergraduate music majors and minors from a midsized, 4-year university and a small community college in the south-central region of the United States. Permission to conduct this research was granted by the university institutional review board through exempted review. Participants completed a demographic survey prior to data collection. Information collected included age, sex, ethnicity, primary instrument or voice part, years of experience in both ensembles and private study, and experience with ensembles and private study outside their primary instrument or voice part. Participants were also asked to list any medical issues related to the musculature of the face (e.g., TMJ, dental work); four participants (vocalist, n = 1; instrumentalists, n = 3) reported being diagnosed with problems attributed to TMJ in the past.
The sample included vocalists (n = 14) and instrumentalists (n = 14) with ages ranging from 18 to 24 years (M = 21.07, SD = 1.65). Participants were grouped as vocalists or wind instrumentalists according to experience. Participants who identified as wind instrumentalists (flute, n = 1; clarinet, n = 7; horn, n = 2; trumpet, n = 1; trombone, n = 2; tuba, n = 1) possessed no more than 2 years of vocal or choral training, whereas participants who identified as vocalists (soprano, n = 3; mezzo soprano, n = 1; tenor, n = 4; baritone/bass, n = 6) possessed no more than 2 years of training on a wind instrument. Most of the participants identified as Caucasian (n = 26) with one participant identifying as African American, and one participant identifying as Asian American.
Equipment
Electromyography data were collected using the Telemyo 2400TG2 EMG system (Noraxon USA, Inc., Scottsdale, AZ) interfaced with a Dell Latitude D830 computer (Dell, Round Rock, TX). All EMG data were collected at a sampling rate of 1,000 Hz. A Panasonic PV GS300 camcorder (Panasonic Corporation of North America, Secaucus, NJ) was interfaced with Myoresearch XP version 1.06.64 software (Noraxon USA Inc., Scottsdale, AZ) for data collection and processing. Disposable 9 mm pregelled bipolar silver-silver chloride electrodes (Noraxon USA, Inc., Scottsdale, AZ) were also used for data collection. The use of surface electrodes was preferable in that they record data from a substantial portion of the muscle as compared to fine wire electrodes. Furthermore, surface electrode data demonstrate less variation within the same subject than do data from fine wire electrodes, which can migrate within the muscle, causing signal changes (Basmajian & De Luca, 1985). For these reasons, surface EMG is more representative of overall muscle activity, which is appropriate for the purposes of this study.
Procedure
Data were collected in the Human Performance Lab on the campus where the study took place over a 6-month period from late spring to late summer. Data collection sessions (one per participant) lasted an average of 35 minutes. Prior to the session, each participant completed the demographic survey and was also given an informed consent cover letter. Participants performed five vowel sounds (ah, eh, ee, oh, oo) while EMG samples of the zygomaticus and masseter were obtained. To randomize the order of vowel sounds performed, vowel sounds were drawn prior to data collection.
Participants were seated on a table and the left side of their face was swabbed with 70% isopropyl alcohol prior to electrode placement. Surface electrodes were applied to the skin overlying the masseter (along the jawline parallel to the ear), to the skin overlying the zygomaticus (diagonally to the corner of the mouth), and to the skin overlying the chin bone as a ground (available in the online Supplemental Figure 1).
Each participant was asked to stand centered in front of the camera as indicated by a tape marking at the appropriate location on the floor. Next, they were asked to vocalize a comfortable, medium-low pitch in their chest voice (i.e., lowest register of the voice) to be used for the entirety of the experiment. Participants were asked to sing without vibrato for this experiment; according to some researchers, the use of vibrato may affect muscle activation (Sapir & Larson, 1993). Each participant performed each vowel three consecutive times while sustaining the vowel for five seconds. After 60 seconds of rest, the participant performed the next vowel using the same protocol until all five vowels in the sequence had been performed. Participants were reminded of the appropriate vowel sound before each performance and cued visually for appropriate timing.
Data Analysis
The EMG data were rectified and smoothed using a root mean square algorithm. The middle 3 seconds of the 5-second trials were analyzed. Mean EMG data for the two muscles were normalized to the peak EMG data recorded during the trials to yield a percentage of the peak EMG as used in Heuser and McNitt-Gray (1991). Peak EMG was defined as the highest EMG amplitude recorded for each participant across all trials for each of the two muscles. This provides a maximum effort of function during the trials. Normalization of the trial data allows for comparison across participants with different muscle activation levels and utilizing the peak EMG allows for a safe normalization procedure for the two muscles of interest. Musician type (vocalist or instrumentalist) served as the independent variable. Activation of the zygomaticus and masseter muscles while performing the five vowel sounds served as the dependent variables. Preliminary assumption testing for normality, linearity, univariate and multivariate outliers, homogeneity of variance-covariance matrices, and multicollinearity were conducted. No significant violation was found.
Results
To compare the muscle activation of the zygomaticus and masseter for each performed vowel sound (ah, eh, ee, oh, oo), a one-way multivariate analysis of variance (MANOVA) was conducted. A statistically significant MANOVA effect was obtained, Pillai’s trace = .73, F(1, 27) = 4.54, p = .003. As a follow-up to the MANOVA, one-way analyses of variance for each of the 10 dependent variables were conducted (p < .05). A main effect was found for the muscle activation of the zygomaticus for the “ah” [F(1, 27) = 6.78, p = .02, partial η2 = .21] and “eh” [F(1, 27) = 5.76, p = .02, partial η2 = .18] vowels. A main effect was also found for the muscle activation of the masseter for the “eh” [F(1, 27) = 13.25, p = .001, partial η2 = .34] and “ee” [F(1, 27) = 5.46, p = .03, partial η2 = .17] vowels. 1
Vocalists (M = 34.39, SD = 14.14) had significantly greater activation of the zygomaticus muscle than instrumentalists (M = 21.71, SD = 11.50) when performing the “ah” vowel (see Table 1). However, instrumentalists (M = 37.46, SD = 12.99) had significantly greater activation of the zygomaticus muscle than vocalists (M = 27.03, SD = 9.77) when performing the “eh” vowel. Instrumentalists also had significantly greater activation of the masseter muscle when performing “eh” (M = 44.17, SD = 12.48) and “ee” (M = 44.28, SD = 10.96) vowels compared to vocalists performing “eh” (M = 28.07, SD = 10.86) and “ee” (M = 33.04, SD = 14.28).
Descriptive Statistics for Zygomaticus and Masseter Muscle Activation by Musician Type Across Vowel Sounds.
Note. Units = % peak EMG (electromyography).
p < .05.
Discussion
The purpose of this study was to compare the facial muscle activation of singers and instrumentalists while performing simple vocal exercises. We found the facial muscle activation of these instrumentalists and vocalists to be significantly different when performing “ah,” “eh,” and “ee” vowel sounds. The result of vocalist participants having significantly greater zygomatic muscle activation than instrumentalists when performing the “ah” vowel may be attributed to singers modifying vowels to increase amplitude while maintaining a resonant sound. The greater activation of the zygomaticus, which is associated with “lift” in vocal timbre (Miller, 1996) in the vocalist participants may indicate that vocalists make more use of space inside the oral cavity to produce an “ah” vowel, while instrumentalists may make more use of the lips and jaw. Miller recommended singers imagine inhaling a fragrance through the nose to describe the appropriate position of the zygomaticus when singing. Instrumentalists had significantly greater zygomatic muscle activation than vocalists when performing the “eh” vowel, which also may be due to contrasting vowel shapes.
Instrumentalists had significantly greater masseter muscle activation when performing the “eh” and “ee” vowels. This could be attributed to a lack of instruction in proper vowel shape (even though many of the instrumentalists in this study had at least participated in a choral ensemble prior to the study) or a wind embouchure influence. Typical training for vocalists involves the production of a resonant tone on each vowel without engaging the masseter. Voice pedagogues typically encourage singers to sing with an “inner smile,” which allows the jaw to be open without engaging the chewing muscles (McQuade, 2016). As Gossett (1989) concluded, oboists when concurrently studying voice had higher abdominal and neck muscle activation, which the author attributed to oboe embouchure. Researchers determined that instrumentalists have performance-related health issues like focal dystonia (Lederman, 1991; Schuele & Lederman, 2003), muscle fatigue, tightness of the lips and facial muscles, TMJ, and jaw pain (Taddey, 1992), which all affect facial muscle activation. All these conditions could be associated with over-use and embouchure, and it is reasonable to assume these muscular-related issues could influence facial muscle use while singing.
If trained vocalists shape vowels in alignment with popular vocal pedagogy practices, it stands to reason that untrained singers, in this case instrumentalists, may be producing vowel shapes based on instrument pedagogy principles rather than on what is considered to be healthy or technically correct in common vocal pedagogy, which may contribute to excessive or inappropriate muscle activation. A lack of vocal training, in combination with training in playing wind instruments, may lead to excessive or inappropriate muscle activation during singing, which is in opposition to most vocal pedagogy philosophies (Bickel, 2008; LaPine, 2008; Large, 1980; Miller, 1996; Vennard, 1967; Voice, 1977; R. C. White, 1976; Wilson, 1991). Miller (1996) might describe this higher level of muscle activation produced by the instrumental participants as vocal tension, though this terminology is not used in the field of biomechanics.
Implications for Music Teacher Education
Most music teacher educators prepare students for the possibility of teaching music in areas that lie outside their primary applied/performance area. For example, vocalists serving as high school choir teachers may need to conduct a pit orchestra during a musical, or instrumentalists teaching a middle school choir may need to vocally model in a pedagogically sound way. Based on recent research on mimicking behaviors, the results of this study may have implications for instrumentalists serving as choir directors or elementary music teachers. Nonverbal conductor behaviors have been shown to influence singers’ vowel shape (Daugherty & Brunkan, 2013; Manternach, 2012a), eyebrow lift (Manternach, 2012a), and neck and shoulder movement (Manternach, 2012b). If singers have the tendency to mimic behaviors of their conductor, then vowel execution may also be mimicked, for better or for worse. Research is needed to determine if facial muscle activation in instrumentalists can be manipulated by vocal training and also to determine if teacher modeling influences student facial muscle activation. Furthermore, a comparison of facial muscle activation between musicians who have both vocal and instrumental training and musicians who have trained solely as vocalists or instrumentalists may be useful in determining if similarities or differences exist between the facial muscle activation of these groups.
Choral music teachers are expected to be proficient vocal models and many instrumental music teachers already incorporate singing as a teaching tool (Robinson, 1996; Wolbers, 2002). Furthermore, most music education degrees lead toward certification in grades K–12, so many music education students may end up teaching elementary music. Elementary music classes often present the first introduction to singing for children, and anybody in such a teaching position bears the responsibility of teaching children to sing with healthy vocal technique. If untrained vocalists instinctively shape vowels in a manner that opposes popular vocal pedagogy, and if children form vocal habits based on vocal modeling, it seems that teachers licensed to teach K–12 music need to be proficient in healthy vocal production to better inform their own teaching. That being said, not all music education programs require that instrumental majors complete coursework dedicated to vocal technique, despite the fact that vocal modeling is a significant aspect to teaching music regardless of grade level. Faculty overseeing music education degree programs may need to require some type of vocal technique training for instrumental music education majors to prepare future music teachers for vocal modeling at the elementary level. Learning from and responding to healthy vocal models during students’ formative years may help protect their voices throughout their middle school and high school music experiences, and potentially for life.
Additionally, many vocalists in university music programs or choral programs have an instrumental background. Many music education programs require students in the instrumental licensure track to participate in a choral ensemble and/or take applied voice. Voice teachers and choral directors who work with students who have played an instrument for many years may need to be sensitive to what appears to be proclivities to have more muscle activation, especially more masseter muscle activation, which might translate into vocal tension. Based on our results, choir directors may need to provide more detailed instruction on proper vowel shape and advice on how to avoid muscular tension to choir members with a strong background in instrumental studies (particularly wind instrumentalists). Vocal modeling with relaxed facial muscles while singing the primary vowel sounds may prove extremely beneficial to instrumentalists in a choral setting. Literature is consistent that muscle activation can be reduced with quality instruction and monitored feedback.
In conclusion, our findings suggest that the facial muscle activation of vocalists and instrumentalists differ when singing certain vowels. University professors who work with instrumentalists on singing tasks should be prepared to devote extra class time focusing on healthy singing technique with direct instruction on proper vowel positions that promote relaxed facial muscle use. Likewise, music teacher educators should carefully observe facial muscle use in music teacher candidates, especially instrumentalists, when singing in front of or with K–12 students to reinforce quality, healthy vocal modeling and techniques.
Supplemental Material
Supplemental_Figure_1_for_Facial_Muscle_Activation_JMTE – Supplemental material for A Comparison of Facial Muscle Activation for Vocalists and Instrumentalists
Supplemental material, Supplemental_Figure_1_for_Facial_Muscle_Activation_JMTE for A Comparison of Facial Muscle Activation for Vocalists and Instrumentalists by Ryan A. Fisher, Aubrey R. Hoult and W. Steven Tucker in Journal of Music Teacher Education
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Student Undergraduate Research Fellowship grant from the Arkansas Department of Higher Education.
Supplemental Material
Supplemental material for this article is available online.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
