Abstract
The purpose of this study was to describe the voice change and voice part assignment of male middle school choir members. Volunteers (N = 92) were recruited from three public middle school choral programs (Grades 6-8). Participants were audio-recorded performing simple vocal tasks in order to assess vocal range and asked to share the music they were currently singing in class. Results revealed 23.91% of participants’ voices could be categorized as unchanged, 14.13% as Stage 1, 3.26% as Stage 2, 10.87% as Stage 3, 26.09% as Stage 4, and 21.74% as Stage 5. The majority of sixth-grade participants were classified as unchanged or in Stage 1 of the voice change and the majority of eighth-grade participants were classified in Stages 4 to 5 of the voice change. Of the participants labeled “tenors” in their choir, over 60% were classified as either unchanged voices or in Stage 1 of the voice change.
Teaching middle school aged students is both rewarding and challenging. Directors of middle school choirs serve at a pivotal time in young adolescents’ vocal development and the skills needed to help students navigate their vocal journey are myriad. Though much research exists to inform and aid middle school choir directors on the voice change, evidence exists that a gap between research and practice remains. Both male and female singers in middle school choirs encounter vocal change as a result of puberty that alters their vocal characteristics; however, the male vocal change results in more severe physiological changes of the vocal mechanism.
Research on the male voice change has increased dramatically, and our understanding of the physical characteristics associated with the male voice change has evolved due to the work of pioneers like McKenzie, Cooper, Swanson, and Cooksey (see Fisher, 2009, for a thorough review of the history of male voice change theories). A five-stage classification system was developed by Cooksey and is frequently used by researchers: Midvoice I, Midvoice II, Midvoice IIA, New Voice, and Emerging Adult Voice (Cooksey, 1999; Cooksey & Welch, 1998; see Figure 1). As the onset of puberty in males has been found to occur earlier in the United States now compared with previous generations (Herman-Giddens et al., 2001), so too has the onset of the male voice change. Fisher (2014) found that male singers were beginning and progressing through the voice change earlier than reported in previous research. Killian and Wayman (2010) found that the majority of boys in sixth grade and all of the boys in ninth grade had started the voice change, whereas Fisher (2010) found that almost half of fourth-grade participants had changing voices. A larger percentage of African American fourth-, fifth-, and sixth-grade students’ voices were changing than Hispanic or White students (Fisher, 2010). Choir members in the male voice change reportedly had higher vocal range than band members (Killian & Wayman, 2010). According to Fisher (2010), males who sing in choirs may not enter the male voice change or progress at the same rate of the voice change when compared with their nonsinging male peers.

Mean ranges for the Cooksey voice change stages.
The voice change can negatively affect male singers’ perception of their voice quality, but correct information about the nature and progression of the voice change can reduce negative perceptions of voice quality (Freer, 2015; Kennedy, 2004) and reduce behaviors harmful to the voice (Kennedy, 2004). Owen (2017) found that singing exercises increased adolescent females’ singing self-efficacy and Killian (1997) concluded that singing through the voice change affects males’ memories and perceptions of the voice change. Singing in choirs multiple years may positively affect the singing self-efficacy of male singers regardless of the voice change (Fisher, 2014).
Research has been conducted as to how choral educators can most effectively help students during the adolescent voice change. For example, choral educators indicated that more information on the male voice change would be helpful to them (Killian, 2003; Killian et al., 2020). Usher (2005) found that choral educators did not feel prepared to help students through the male voice change by their teacher preparation programs, and that participants specifically wanted more information about adolescent developmental characteristics, vocal physiology, and classroom management. Choral educators indicated that their teacher preparation programs were least helpful in providing information on the adolescent voice change, while they thought they learned the most from observations of other choral educators and attending professional development workshops (Clark, 2019). Choral educators’ self-efficacy for teaching music and teaching students in the voice change positively correlated with how often they provided instruction related to the voice change (Clark, 2019).
The progression of the male voice change is unique for each student, and choir directors should consider means of monitoring students’ voices and individualizing strategies used to assist students (Kennedy, 2004). Directors of successful choral ensembles involving male singers in the voice change kept a continuous record of developing vocal capabilities (Keating, 2004). More choral educators engage in group testing than in individual testing, and most teachers test their students two to three times a year (Usher, 2005). Choral educators who individually tested junior high students’ voices were much more likely to participate successfully in adjudicated contests; however, some choral educators may not be using testing procedures recommended by experts in the male voice change (Usher, 2005). Killian (2003) found some directors assigned “nonsinging” roles to males in the voice change which, according to Freer (2006), may result in males leaving choral singing due to perceptions that they are bad singers. Adler (2002) found that many choral educators modify parts and provide additional voice instruction for students in the male voice change. Of the strategies provided for assisting singers through the male voice change, choral educators indicated that they asked singers to take a part down the octave and rewrote parts most often (Killian, 2003; Killian et al., 2020). Limiting the voice parts adolescent male students sing to tenor and bass does not seem to be helpful (Kennedy, 2004).
Researchers have also recommended that choir directors should carefully consider adolescent students’ vocal range and vocal characteristics when assigning voice parts (Fisher, 2020; Friddle, 2005); however, Sweet (2018) found some females felt their choir directors assigned their voice part based on the needs of the ensemble rather than the individual singer. Additional factors that choir directors appear to consider when assigning voice parts include students’ choral experience, balance requirements and music reading ability (Dillon, 2013). Because students’ assigned voice parts may be misclassified (Dillon, 2013), researchers have recommended that choir directors avoid labeling singers to specific voice parts and allow them to gain experience singing various parts throughout their time in choir (Silvey, 2015; Sweet & Parker, 2019). Some community choral ensembles for children assign parts based on vocal range rather than biological sex (Keating, 2004). Fredenburgh (2003) found part-switching to have multiple benefits for students in treble choirs, such as relieving vocal strain from regularly singing in one register, facilitating the exploration of students’ full vocal range, developing students’ aural skills, and mitigating negative effects of identifying with a choral voice part. According to Kennedy (2004), the American Boychoir School (ABS) assigned voice parts during the voice change based on the needs of the student rather than the needs of the choir. This meant students may have been assigned different voice parts depending on the vocal demands of each piece rather than being labeled one voice part for all choral pieces.
Silvey (2015) recommended choral directors should distinguish between the function of a voice part and the vocal identity of the singers performing that voice part, and also encouraged directors to be transparent with their students in their rationale for voice part assignments. Students’ gender identity in combination with their voice identity should also be considered when assigning voice parts (Palkki, 2017). The terms soprano and alto had emotional connotations for female undergraduate and graduate voice students (Sweet & Parker, 2019), leading the researchers to recommend choir directors avoid labeling female singers as soprano or alto until laryngeal growth is complete. Numerous methods have been developed for assigning voice parts for adolescent males that are linked with the various theories of the male voice change (Cooksey, 1999; Cooper & Kuersteiner, 1970; McKenzie, 1956; Swanson, 1977).
Given the impact of repertoire on singers’ choral experience, singers may still be adversely affected by inappropriate or unsuitable choral literature, even if choir directors are carefully tracking their singers’ vocal ranges and assigning voice parts based on singers’ vocal characteristics. Poorly selected repertoire can negatively affect voice students’ music achievement as well as male singers’ perception of their success while navigating the voice change (Freer, 2015; Nix, 2002). Palkki (2015) recommended choir directors should consider the text of selected repertoire in confronting gender stereotypes. Nix (2002) classified repertoire selection into four categories of criteria: singers’ physical limitations, voice classification, music skills, and factors relating to music expression and emotion. For young singers, the most important criteria are the students’ vocal limitations and voice classification (Nix, 2002). In a survey of choral directors, Killian (2003) found most respondents chose to use two-part treble and three-part mixed repertoire for choirs with singers going through the male voice change. According to Funderburk-Galván (1987), a majority of junior high school choral directors (eighth and ninth grades) recommended using three-part mixed or SAB repertoire with changing male voices. The majority of directors in that sample used the tenor range of G3 to G4 and the baritone range of B2 to E4 as a guide when selecting literature for their junior high ensembles. Though choral directors carefully consider vocal range and difficulty of repertoire when selecting music, these considerations may have less of an impact on adolescent males’ enjoyment of choir than the style and content of the music (Kennedy, 2002).
Given the importance of literature selection and voice part assignment, we chose to examine voice part assignment during the male voice change. To date, no researchers have measured assigned voice parts as they relate to the male voice change. The purpose of this study was to describe the voice change and voice part assignment of male middle school choir members. Our research questions were:
What are the vocal characteristics of emerging adolescent male choir members in sixth, seventh, and eighth grades?
What are the vocal characteristics by voice part of emerging adolescent male choir members in sixth, seventh, and eighth grades?
Method
Participants (N = 92) were recruited from three different public middle school choral programs composed of sixth-grade (n = 8), seventh-grade (n = 36), and eighth-grade (n = 48) students located in a large, urban area in the Mid-South. The sample represented Black (n = 48), White (n = 30), Latino (n = 4), Asian (n = 5), and mixed race (n = 5) males who ranged in age from 11.49 to 16.32 years old (M = 13.47, SD = 0.76). All of the choral programs involved in this study were selected because of their history of success at concert and sight-reading festivals as evidenced by consistent superior ratings in concert festivals over a 3-year span. In addition, they had a large population of male singers from which to sample for this study. The choral directors (female: n = 1, male: n = 2) at these programs had been at their respective schools for a minimum of 3 years. Institutional review board (IRB) permission was secured by the first author’s university.
Procedures
On securing parental consent, participants were tested individually during their choir class in their choir director’s office to minimize possible distractions. We first collected the two or three choral octavos their choir was preparing for the spring choral festival contest. We noted each piece, arranger/composer, voice part assigned, and the vocal range each piece required of the singer. If a teacher had adjusted or rearranged any of the assigned vocal parts, that was also noted. See Table 1 for the list of choral pieces participants were preparing when data were collected. We also noted the choir type each participant sang in (mixed chorus, tenor/bass chorus, or treble chorus). Additional demographic information (race, grade level, and date of birth) was also collected from the participants.
Choral Pieces Performed by Participants.
We then recorded each individual participant performing simple vocal tasks to assess vocal range. No stimulus pitch was used. We rehearsed these vocal exercises as a group prior to individual assessment to be sure all participants clearly understood the procedures. Once the recording began, each participant stated his name and counted backward from 20 to establish his fundamental speaking pitch (FSP) frequency. He was then instructed to perform three separate ascending glissandi on an “ah” vowel in attempts of singing his highest note possible. The same instruction was given to perform three separate descending glissandi in attempts of singing his lowest note possible.
All participants were recorded using a Zoom H4N Handy Recorder (ZOOM Corporation), and audio was stored on an SD card. The recorder was held approximately six inches from each participant’s mouth to minimize background noise. After data were collected, the SD card was inserted into the primary researcher’s computer and each audio file was analyzed using the tuner in GarageBand 6.0.5 (Apple, Inc.) calibrated to A4 = 440. The FSP frequency was established by analyzing the participant’s voice as he counted backward from 20 since the speaking voice has been documented to be three to four semitones above the lowest terminal pitch (LTP) in males (Barresi & Bless, 1984; Moore, 1995). The highest note reached in the ascending glissandi was labeled as the highest terminal pitch (HTP) and recorded in hertz. The lowest note reached in the descending glissandi was labeled as the LTP and recorded in hertz. The primary researcher also documented vocal characteristics like presence of falsetto and phonational gaps since these characteristics are associated with certain stages of Cooksey’s voice change classification system.
To assess the reliability of the vocal range test, the first author randomly selected the audio files from 20% of the sample for independent analysis by the second author. Interjudge reliability was calculated by dividing agreements by agreements plus disagreements. An agreement existed if the HTP and LTP for each participant was within one half step between the two authors. Interjudge reliability was high between first author and second author (HTP = 95%; LTP = 95%). After vocal ranges were established, each participant was classified using Cooksey’s voice change stages. Participants classified by the first author as unchanged were coded as 0, whereas those in Stages 1 through 5 were coded with the parallel number.
The choral repertoire for participants was documented using the name of the piece, composer, publisher, and vocal range for each part. Each participant told the researchers which voice part he was assigned for each piece and that was documented in the dataset. Those with hybrid (e.g., tenor with some notes from bass part) or modified voice parts (e.g., choir director rewrote a portion of the assigned voice part) were also documented. We later added the lowest pitch and highest pitch of each piece for each participant to assess the actual overall vocal range each student would perform on the assigned music.
Results
Results revealed 23.91% (n = 22) of participants’ voices could be categorized as unchanged, 14.13% (n = 13) as Stage 1; 3.26% (n = 3) as Stage 2; 10.87% (n = 10) as Stage 3; 26.09% (n = 24) as Stage 4; and 21.74% (n = 20) as Stage 5. Participants’ FSP ranged from 78.00 Hz (Eb2) to 247.00 Hz (B3), M = 156.20 Hz (Eb3), SD = 46.81. Participants’ HTP ranged from 123.47 Hz (B2) to 1244.51 Hz (Eb6), M = 465.84 Hz (Bb4), SD = 263.57. Participants’ LTP ranged from 65.41 Hz (C2) to 233.08 Hz (Bb3), M = 145.49 Hz (D3), SD = 46.65. Over 83% (n = 77) of the participants performed in a mixed-voice choral ensemble (SATB, SAB, or 3-part mixed) with only 8% (n = 7) assigned to a treble-voice choral ensemble and 9% (n = 8) assigned to a tenor/bass ensemble. 1
Descriptive statistics by grade were calculated for FSP, HTP, LTP, Music HP, Music LP, and voice change stages. The results in Table 2 indicate that the mean FSP of participants lowered as grade level increases with the greatest mean difference appearing between sixth- and seventh-grade participants. Similar trends can be seen in participants’ mean HTP and mean LTP. A comparison of the mean HTP and music mean HP reveals that participants’ vocal range was higher than the music vocal range, especially for the sixth- and seventh-grade participants. This stands in contrast to comparison of the mean LTP and music mean LP which was much more closely aligned for each grade level.
Descriptive Statistics for Sixth-, Seventh-, and Eighth-Grade Participants.
Note. Numbers in parentheses denote standard deviations. FSP = fundamental speaking frequency; HTP = highest terminal pitch; LTP = lowest terminal pitch; HP = highest pitch; LP = lowest pitch. A4 = 440 Hz.
The frequency report of voice change stage by grade level revealed the majority of sixth-grade participants were classified as unchanged and in Stage 1 of the voice change, whereas the majority of eighth-grade participants were classified in Stages 3 to 5 of the voice change. Results for seventh-grade participants were more divided with 53% (n = 19) being classified as unchanged or in Stage 1 and 44% (n = 17) classified in Stages 3 to 5 of the voice change. In reviewing the frequency report of voice part assignment by grade level, it is clear that the majority of sixth-grade participants were assigned a treble voice part, the majority of seventh-grade participants assigned the tenor part, and the majority of eighth-grade participants assigned bass/baritone parts.
Descriptive statistics by assigned voice part were calculated for FSP, HTP, LTP, Music HP, and Music LP. The results in Table 3 indicate that the mean FSP, HTP, and LTP is higher for each successive voice part. A comparison of the mean HTP and music mean HP reveals close alignment for participants assigned the bass/baritone part, but a large mean difference (228.99 Hz) for participants assigned the tenor part. As an example, the mean HTP for tenor participants was around Eb5, whereas their music mean HP was around G4. A comparison of the mean LTP and music mean LP reveals close alignment regardless of assigned voice part.
Descriptive Statistics by Assigned Voice Part.
Note. Numbers in parentheses denote standard deviations. FSP = fundamental speaking frequency; HTP = highest terminal pitch; LTP = lowest terminal pitch; HP = highest pitch; LP = lowest pitch. A4 = 440 Hz.
The frequency report of voice change stage by assigned voice part reveals 78% (n = 35) of bass/baritone participants were classified in Stages 4 and 5 of the voice change. All of the participants assigned to a treble part were classified as unchanged voices or in Stage 1 of the voice change. Of the participants assigned the tenor part in their choirs, 64% (n = 25) were classified as either unchanged voices or in Stage 1 of the voice change.
Discussion
The purpose of this study was to describe the voice change and voice part assignment of male middle school choir members. In this study, results revealed the majority of sixth-grade and seventh-grade males were classified as unchanged or in Stage 1 of the voice change. However, the majority of eighth-grade males were classified in Stages 3 to 5 of the voice change, which aligns closely with the findings in Fisher (2014). Killian and Wayman (2010), however, reported a much lower percentage of sixth- and seventh-grade participants to be classified as unchanged or in Stage 1 of the voice change. As expected, the overall mean FSP, HTP, and LTP lowers with each successive grade as more males’ voices are lowering due to puberty. The mean HTP and LTP for each grade in our current study were very similar to those reported in Fisher (2014).
Although the majority of sixth-grade participants were assigned treble voice parts, seventh- and eighth-grade participants were almost exclusively assigned tenor and bass/baritone parts even though unchanged voices were found in the sample of seventh- and eighth-grade males. For those participants who were assigned the tenor voice part, over 60% were classified as unchanged voices or in the early stages of the voice change. This finding aligns with Sweet (2018) in that participants felt directors assigned voice parts based on the needs of the choir rather than individual singers’ vocal needs. There appears to be a rush for choir directors to assign male singers the more stereotypical “male voice parts” once singers reach the seventh grade, regardless of the singers’ vocal range. The practice of assigning males’ voice parts due to the needs of the choral ensemble contradicts the practices of the ABS where voice part assignments were based on the needs of the student (Kennedy, 2004).
According to Sweet (2018), assigning voice parts based on the needs of the ensemble for appropriate balance may limit singers’ individual vocal development and singing potential. As participants in her study recalled, “they trusted their choral teachers and assumed their teachers had good intentions. However, perceptions of choral teachers’ decisions . . . now seemed misguided to many” (p. 145). Researchers have yet to measure if males singing voice parts outside of their natural vocal range results in potentially negative vocal habits or physical discomfort; however, this practice does not seem to align with the goal of optimal singing based on singers’ vocal characteristics. As Kennedy (2004) discovered in her research with the American Boychoir, singers were sometimes assigned different voice parts depending on the vocal demands of each piece rather than being labeled one voice part for all choral pieces. One participant recalled being allowed to attempt to sing in multiple registers during his voice change while maturing in the choirs as ABS. ABS also allowed for individualized voice instruction and more rehearsal time, which may not reflect the most typical school choral experience.
We also found that the majority of participants were members of a mixed-voice choral ensemble. The voicing for the majority of repertoire assigned to those in mixed-voice ensembles were three-part mixed or SAB, which mirrors the results of Funderburk-Galván (1987). Though our sample may not reflect the entire middle school choir population, it appears many middle school programs provide mixed-voice ensemble opportunities, especially for Grades 7 and 8. Although mixed-voice ensembles provide great opportunities for assigning males a voice part most appropriate to their voice change stage, it may be that many choir directors are primarily assigning male singers to tenor or bass parts despite singers’ vocal ranges rather than taking advantage of the possibilities to assign treble-voice parts when appropriate.
Future researchers may wish to further investigate the decision-making process of middle school choral directors when selecting music each year. Ensemble type should also be considered in evaluating choral directors’ music selection process to determine whether attributes like voice change classification are not considered for some ensemble types. Research is also needed to examine the effects of assigning tenor or bass parts to unchanged voices on their vocal physiology and vocal health.
Implications for Middle School Choir Directors
Based on the findings of this study, we recommend choral directors regularly assess the vocal ranges and vocal characteristics of their male singers to ensure they are assigned appropriate voice parts regardless of traditional gender-stereotyped voicing (Keating, 2004; Kennedy, 2004; Usher, 2005). With smartphone tuner apps like Cleartune (Bitcount ltd.), directors can quickly measure singers’ HTP and LTP using the same procedures employed in this study. Tracking each male singer’s HTP and LTP while noting vocal characteristics like the emergence of falsetto can help directors identify the voice change stage of each singer, which may lead to a modification of vocal part assignment. This same process can be implemented with a piano, but we have found removing a stimulus pitch and using glissandi up and down on the “ah” vowel allows for faster identification of vocal range without sacrificing accuracy (Fisher, 2010, 2014). A more accurate assessment of students’ vocal range may result in choral directors either assigning the student an existing voice part or creating a hybrid part (e.g., portions of the alto and tenor parts) that fits the students’ range.
We also encourage choir directors to freely discuss with their singers the various challenges and characteristics associated with the voice change and to regularly dialogue with their singers about voice part assignments and specific issues students may be encountering with their assigned voice parts (Palkki, 2017; Silvey, 2015; Sweet & Parker, 2019). This communication with singers may result in directors modifying individual singers’ voice parts or assigning a different voice part, but singers will feel empowered and included in the solution process. Accurate information about the nature of the voice change decreases students’ negative perceptions of the voice change as a “scary” or embarrassing process (Freer, 2015; Kennedy, 2004). Such information paired with useful vocal exercises can reduce harmful vocal behaviors on the part of singers going through the voice change process (Owen, 2017).
Given the importance of and constraints imposed by repertoire, choir directors should carefully consider the music selection process. Many choir directors use two-part treble and three-part mixed repertoire with students progressing through the male voice change (Killian, 2003; Killian et al., 2020); however, no single piece of music can perfectly accommodate and develop every student’s voice. Choir directors may wish to consider modifying pieces through rewriting parts or asking students to part-switch (Fredenburgh, 2003; Killian, 2003; Killian et al., 2020). Middle school choir directors should avoid placing males in the early stages of the voice change in a traditional “men’s chorus,” which typically performs tenor/bass literature. Though most treble voice singers can sing many tenor parts found in tenor/bass literature, they are not provided opportunities to use the beautiful and unique treble range associated with the unchanged male voice and instead are relegated to singing in the lowest extremes of their chest register which could contribute to unhealthy habits of lowering the larynx to produce these notes. Mixed-voice middle school ensembles provide directors more flexibility of assigning male singers to a more appropriate voice part based on singers’ vocal range.
While males are in the early stages of the voice change, we advise directors to avoid labeling singers with a voice part but rather expose their singers to various treble voice parts as long as their vocal range aligns with the range of the assigned vocal part (Palkki, 2017; Silvey, 2015; Sweet, 2018). Many directors have found success in relabeling voice parts with colors (i.e., Treble 1’s are Blue). This practice is appropriate for all singers regardless of sex to avoid permanent identification as a particular voice part. It is especially important for unchanged males singing treble voice parts in a mixed ensemble to not be labeled “soprano” or “alto” since those terms are often associated with gender. Choral directors might also clarify for students that voice part labels refer to range rather than gender. As male singers move through the final stages of the voice change, assigning a tenor or baritone voice part may be more appropriate; however, many baritones in high school may evolve to become tenors in college. Regardless of which part male singers are assigned, vocal technique must be addressed alongside rehearsal of the repertoire. Choral directors should use the warm-up sequence of each rehearsal to teach and reinforce concepts like appropriate breath regulation needed for singing, vocal resonance, and vocal agility while also allowing singers to experiment with emerging falsetto and vocal range extension. For more guidance on strategies for developing the male voice through the voice change, we refer readers to Cooksey (1999) and Fisher (2020).
Finally, we concur with Freer (2006) that choir directors should acknowledge the power of their words. Choir directors who assign students “nonsinging” activities or encourage students to mouth the words convey a message that the student is not a good singer and does not contribute to the overall success of the choral ensemble. According to Killian (2003), choral directors are still assigning students nonsinging roles in the choir and many directors are simply allowing male singers to sing everything down the octave (Killian et al., 2020) or self-assign their voice part. This may be due to some choral directors’ lack of knowledge related to the male voice change or a ceding of leadership and expertise of the director to the young student. Though no director will have all of the right answers, as experts in our field, we must, above all, value each singer and do our best to develop and enhance the voices of all of our students.
Footnotes
Author’s Note
Nancy L. Summitt is now Assistant Professor of Music Education in the School of Music at Central Michigan University.
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) received no financial support for the research, authorship, and/or publication of this article.
