Introduction
Why is it that some people seem unable to sing a melody correctly and with reliable pitch? This question has come to my attention as a music teacher in kindergarten teacher education; I have observed that approximately 10% of students are poor pitch singers (PPS), an observation that is also supported by research (Dalla Bella, Giguère & Peretz, Reference DALLA BELLA, GIGUÈRE and PERETZ2009).
Various aspects of poor pitch singing have been investigated during the last few decades. One focus has been people's tendency to define themselves as tone-deaf because of a perceived inability to sing, and to consider their tone-deafness to be a permanent condition which limits their potential musical achievements (Sloboda, Wise & Peretz, Reference SLOBODA, WISE and PERETZ2005; Knight, Reference KNIGHT2013). The term tone-deafness is often associated with the term Congenital Amusia (Peretz, Champod & Hyde, Reference PERETZ, CHAMPOD and HYDE2003), which is defined as a music-specific perceptual disorder appearing in individuals who have severe difficulties perceiving and processing melodic and temporal information. Congenital Amusia (CA) is thought to affect about 4% of the population (Hyde & Peretz, Reference HYDE and PERETZ2004), whereas other research has suggested that around 17% of the Western population define themselves as tone-deaf (Cuddy et al., Reference CUDDY, BALKWILL, PERETZ and HOLDEN2005). However, studies (Cuddy et al., Reference CUDDY, BALKWILL, PERETZ and HOLDEN2005; Wise & Sloboda, Reference WISE and SLOBODA2008) show that most people who define themselves as tone-deaf score within the normal range on the Montreal Battery of Evaluation of Amusia (MBEA), which consists of six subtests: contour, interval, scale, rhythm, meter, and memory. While faulty tonal perception may cause difficulties in singing, it is not the only factor; Pfordresher and Brown (Reference PFORDRESHER and BROWN2007) have examined the processes of production, memory and/or sensorimotor integration among poor pitch singers. They favor a sensorimotor model of mistranslation in converting heard notes to phonation targets. Building upon this research, Hutchins, Peretz and Gauthier (Reference HUTCHINS, PERETZ and GAUTHIER2012) reviewed the evidence for each of these factors, and found that, although the Pfordresher and Brown (Reference PFORDRESHER and BROWN2007) study demonstrates multiple possible causes of poor singing, most of the problems lie in motor control and timbral-translation errors, rather than with a purely perceptual deficit. The impact of timbre on imitation of novel songs has been studied by Lévêque, Giovanni and Schön (Reference LÉVÊQUE, GIOVANNI and SCHÖN2012), who found that PPS perform better when they have a human voice rather than a synthetic sound source as a model. The impact of lyrics on singing skills varies. CAs seem to perform better when using lyrics (Dalla Bella et al., Reference DALLA BELLA, GIGUÈRE and PERETZ2009), whereas occasional singers sing more accurately on non-lyric syllables (Berkowska & Dalla Bella, Reference BERKOWSKA and DALLA BELLA2013).
Research on poor pitch singing has shown that there are multiple causes and an array of combinations of these within the groups of singers studied. Heresniak (Reference HERESNIAK2004) discusses this issue in his article about adult ‘bluebirds’, in which he describes ten very different cases of PPS. These were having difficulties in singing due to audiological deficits, lack of practice, voice or body tensions, lack of breath, energy, and various perceptional challenges.
Developmental psychology offers another approach to this issue, based on the notion that the ability to relate to music may be considered a universal feature in all human beings. Welch (Reference WELCH1998) proposes four phases in the development of pitch accuracy in children's singing: from the initial ‘chant-like’ and word-focused singing within a restricted range, through contour singing, more precise interval matching, and finally to accurate singing in which a consistent key is maintained throughout a simple song. Children move through these phases at different speed. Moreover, there is evidence that environmental factors such as the song curriculum of childhood, poor vocal models, quality of available feedback, and negative peer pressure influence on singing development (Welch, Reference WELCH, Ostrem, Thurman and Welch2000). This has led to a hypothesis that the development of some adults who experience difficulty singing may have been halted in the early stages due to adverse environmental factors (Welch, Reference WELCH2001; Knight, Reference KNIGHT2013).
In children's singing development, the maintenance of a consistent tonality is the last feature achieved (Welch, Reference WELCH1998). This could be due to motoric development as well as to development of music processing skills.
The Model of Music Processing (MMP), developed by Peretz and Coltheart (Reference PERETZ and COLTHEART2003), is a neurocognitive model describing how music is processed in order to develop musical schemata. The model suggests that musical input is analyzed in systems of pitch and temporal organization that work in parallel but independently. Within the system of pitch organization, processing is hierarchical, starting with contour analysis, continuing with interval analysis, and ending up with tonal encoding. Information is then stored in a musical lexicon. However, this model does not take into account how memory (the lexicon) affects processing. A key assumption in the MMP is that the components have a biological basis and therefore reflect basic building blocks of musical skill that are independent of training.
A study by Krumhansl and Kessler (Reference KRUMHANSL and KESSLER1982) cited in Sloboda (Reference SLOBODA2005) identifies which scale degrees are perceived as most fundamental of tonality. In this study the tonal context was given to the test persons followed by a test stimulus. The test persons were asked to rate the degree to which this stimulus fitted, in a musical sense, with the key-defining context. The results show that the notes of the tonic triad have the highest ratings, followed by the notes of the diatonic scale, with the chromatic notes being given the lowest rating. This is confirmed by traditional music theory. Many possible explanations have been given for this result (Sloboda, Reference SLOBODA2005), the most likely being that these notes are the most dominant in Western tonal music. Several studies (Youngblood, Reference YOUNGBLOOD1958; Hughes, Reference HUGHES and Yeston1977; Knopoff & Hutchinson, Reference KNOPOFF and HUTCHINSON1983; Krumhansl, Reference KRUMHANSL1990) have carried out note counts on melodic lines in non-modulating tonal music, showing the same order of importance of the scale degrees. Thus there is a strong correlation between these studies and those of Krumhansl and Kessler (Reference KRUMHANSL and KESSLER1982).
Interestingly, the same tonal hierarchies are used in parts of Western ear training literature: for example Ravn Jensen (Reference RAVN1956), Raitio (Reference RAITIO1971), and Jørgensen (Reference JØRGENSEN1975). Here the sight-reading exercises start with establishing a tonal scaffold consisting of the tonic, the fifth and the third scale degree, followed by diatonic passages, and finally modulations.
While the above studies examine functional explanations for poor pitch singing, there are also intervention studies aimed at investigating how PPS can develop their singing skills. Numminen (Reference NUMMINEN2005) and Anderson et al. (Reference ANDERSON, HIMONIDES, WISE, WELCH and STEWART2012) suggest that improvement is possible.
The present research developed from the process of assessing three regular singing examinations at a Norwegian university college. It sheds light on how features of the singing repertoire, self-concept and practice influence singing development with PPS.
Background
In my experience, the issue of poor pitch singing is difficult to explore because those who have problems singing are often reluctant to deal with them. They have low musical self-confidence and feel ashamed if anyone can hear them sing. Thus, it has been difficult to establish their cooperation in a research project. Nevertheless, my aim has been to explore this issue in order to gain more knowledge and to offer some kind of remediation for PPS.
The background for the present research was the singing examination in the kindergarten teacher program at a Norwegian university college between 2012 and 2014. Singing is a basic activity both in everyday life in Norwegian kindergartens and in the Norwegian Framework Plan for the Content and Tasks of Kindergartens (Ministry of Education and Research, 2011), described in the learning areas Communication, language and text, and Art, culture and creativity.
Thus, music has traditionally been a core subject in Norwegian kindergarten teacher education. At present, music has become part of the interdisciplinary subject Art, culture and creativity together with drama, pedagogy and arts. Here the students build and strengthen their basic musical and didactical skills, and they learn how music can contribute to children's personal and social development in an overall perspective.
The singing examination of 2012–2014 was a partial examination in music counting 30% of the overall grade, whereas planning and carrying out a music didactical situation for children was the content of the partial examination counting the last 70%. During the actual time span there were four music teachers in the kindergarten teacher program at my institution. The singing examination was usually assessed by the regular music teacher of the class, but in cases of doubt we exchanged sound files for second opinions.
Each student handed in ten self-recorded songs he/she had chosen for assessment. The only criterion for selection of songs was that the music teacher should be able to check for correct performance using notated versions or other recordings. The recording situation was chosen and controlled by the students. They could record their songs as many times as they wanted, and choose the best ones for assessment, and they could do the recordings at home if they preferred. The students were graded A-E (F) on the following rating scale:
A Healthy sounding voice. The student sang accurately in tune within a range suitable for children (mostly above middle C). The student sang with good phrasing, giving each song its own style.
B As A, but the student might have small inaccuracies in intonation or small errors in melody.
C As B. At least five songs had to be performed correctly.
D Three to four songs had to be performed correctly. Key stability could seem poor.
E Melody was recognizable. Key changes could occur during the song, and some melody fragments might be unrecognizable.
F (F Fail. There was no description of this grade, since it was hardly in use.)
This assessment scale resembles parts of the rating scale for singing accuracy in song performance (Wise & Sloboda, Reference WISE and SLOBODA2008) but was developed independently. As the music teacher for nine classes between 2012 and 2014, I assessed 239 students (2390 recordings). Out of these, 25 students were graded D–F for their recordings and therefore were defined as PPS. This group, which consisted of both female and male students, had varying patterns of difficulties in singing, related to such features as vocal range, ability to sing intervals accurately, or ability to stay within a constant key. What appeared surprising and interesting with regard to the portfolios of these students, was the varying pattern of mastery each showed in their different recordings. One student, for instance, performed three songs at a B-level, whereas the rest of her songs were at levels D and E. The only student who was graded F had one recording equivalent with the grade C, whereas the rest of her recordings were very weak. This raised the initial research question: Were there features within the musical material itself that could explain the varying degrees of mastery?
Methods
2011/2012
The present research started with the assessment of 74 students in 2012. I registered information about every song regarding key and vocal range, and whether the song was performed correctly, accurately, in a consistent tonality, and in an appropriate tessitura. This was information directly related to the grading scale. Next, for my own interest, I registered detailed information about the weak performances: what went wrong, where the problems started and how and to which degree the students dealt with them. I also noticed that some of the stronger students (mostly those who were graded C) sometimes could stumble at similar points as the weaker ones. Seven students were graded below C and in the following research, they are defined as PPS.
As I noticed the varying patterns of mastery within the portfolios of the PPS, I went back to the detailed assessment notes to analyze the songs they performed the best with regard to correct reproduction and consistent tonality, as these two factors very often connected to each other. Why were these songs performed better than the others? Did the melodic structure of the songs anchor tonality in any way? Did repeats play a role? What about tempo and total amount of information in the song? Did the total vocal range of the songs play a role? I noted the structural characteristics of the best performances of the PPS, and made a list of assumable easy-to-sing-features and an opposite list with difficult-features based upon the most common stumbling factors of the PPS. I also assumed it might play a role how well known the songs were to the students, so I sent a mail asking how long they had known their best songs. The two students who replied told that they had known their best songs since childhood.
The analysis led to reflections on how this knowledge could be useful for the next year's students. Certain features of the sung material seemed to have had an impact on mastery. The PPS clearly seemed unable to predict whether their chosen songs would turn out to be easy or difficult to sing. The PPS of the next year's classes would possibly perform better if they chose songs with easy-to-sing features. I decided to facilitate this by choosing 29 nursery rhymes and traditional songs with easy-features that should be well known to Norwegian occasional singers and make recordings of them available at the school's digital learning platform as tools for their own practice.
Helping PPS to perform better concerns institutional accountability in multiple ways. If the students perform better they may be able to pass their singing examinations and thus gain a stronger musical self-confidence which may lead to more singing and further improvement, whereas less vocal mastery can lead to no further singing whatsoever. If the students are better performers, they may serve the kindergartens better as competent vocal models for the children. The staff in kindergartens should be ‘conscious of the ethical, aesthetic and cultural values that they are communicating’ (Ministry of Education and Research, 2011). Songs are cultural values; therefore, kindergarten teachers should be able to convey songs correctly to fulfil their roles in society.
During the academic year, I had already recorded and distributed sound files of songs learnt in music classes as support for the students’ memory and as help to find an appropriate tessitura for singing with children. I found sound files to be low threshold tools that the students could use when practicing on their own without feeling exposed or ashamed.
Since the target group for the 29 songs provided was future kindergarten teachers, the songs were recorded in tessituras suitable for singing with small children – that is, the songs were recorded in keys that allowed the main parts to stay above middle C. This represented a practical dilemma, as many untrained singers (especially females) seem to be uncomfortable using the vocal register above their speaking register, thus have a lower comfort zone often with the top point somewhere around G4. Brown (Reference BROWN1996) refers to the speaking register as Register 2 and the register above as Register 3, and indicates that Register 2 ends at A4–C5 depending on type of voice. However, my experience is that untrained singers with low self-confidence are lower pitched. This dilemma was handled by finding a small collection of songs with a restricted vocal range (from a perfect fourth to a major sixth). The recordings were kept within a limited number of keys to facilitate stabilization of tonality.
The starting note and the starting syllable of the song were sung separately before the song started to help the student tune in to the right pitch. In the time gap between the starting note and the song, the student could stop the recording, imitate the pitch and go back for a repeat if she felt that she did not hit the right pitch. The songs were recorded both in female and male versions by accurate but untrained singers. The male version was offered because - based on my own experience from teaching and choir conducting - untrained male singers often have difficulties when imitating a female voice; they find the right pitch more easily when imitating a man. All the songs were recorded without accompaniment in tranquil tempi to allow for better focus and precision as well as timbral imitation.
As additional tools for the students’ own practice, a table of the different features of the songs and a user's manual was made.
2012/2013
The new classes were informed directly in the classroom about the new singing aids available at the digital learning platform. Halfway through the year, the students answered a questionnaire about whether they had used the tools and whether they had found them helpful. I wanted to know if the new tools had any impact on their work and self-concept as singers. The answers could not show whether the students’ singing skills had actually improved if not combined with before-and-after tests of the students’ singing. However, I assumed that such a procedure would intimidate most of the students, and thus I found it unethical.
Many students reported that they had been using the sound files. Some of the recordings of 2013 suggested that the students had been using model sound files because their recordings appeared to be in the same key and tempo as the model recordings. However, the examination portfolios of 2013 were not very different in character and level from the ones of 2012. This year I assessed 79 students; ten were PPS.
At this point, I wanted to expand my knowledge through clarifying interviews and further exploring of the findings from the analyses by singing mentoring of the PPS through electronic exchange of new sound files. Hence, there was a need to formalize this as a study. The PPS of 2013 received a letter of information about the background and intentions of my study (improvement of adult singing skills), and were asked to join by giving their written consent to as many as possible of the following points of participation:
1. Allow me to keep their examination recordings.
2. Participate in an informal talk about singing background from family and school, feelings towards singing, their preparation for the singing exam; and finally about the assessment scale and how they performed on the exam with an emphasis on what they did well.
3. Participate in singing mentoring through electronic exchange of new sound files.
4. Allow me to use their sound files anonymously in oral presentations or electronic articles.
All participation was voluntary, and the PPS were free to change their minds about degree of participation or to leave the study at any time. The study was reported to and approved by the Norwegian Centre for Research Data (NSD). This is a resource center that assists researchers with regard to data gathering, data analysis, and issues of methodology, privacy, and research ethics.
In 2013, six of ten PPS came for an informal talk. Five of them agreed to participate in singing mentoring, but only two actually did. However, three of the PPS decided to repeat their examinations; one of them did so because of her failing grade. After six months, they handed in new sound files, which were assessed by one of the other music teachers who did not know the students’ original grades. All the students improved by one grade.
2013/2014
The new students were informed about the practicing tools, and the assessment, analysis and interview processes were repeated whereas I omitted the questionnaire. In 2014, I assessed 86 students, from where eight were PPS. Two male students came for the informal talk, they both agreed to singing mentoring but after summer holiday both withdrew.
The findings from the assessment and analysis processes of 2013 and 2014 were mostly similar to those of 2012. In this article, I primarily focus on data from the analysis of the examination sound files of the 25 PPS. Some of my experiences and reflections from the other parts of the project will be included in the discussion.
Findings
Physiological factors
The sound files revealed that some of the students clearly had physical difficulties in voice production, such as insufficient breath or range restriction equivalent to their speaking register only. Songs that required changes of vocal register were very difficult for some of these students, and could result in loss of tonality, either for physical or for attention-related reasons, or both. This group of students performed better in songs that did not require changes of vocal register, or songs having few such changes
Some students seemed to have sensorimotor difficulties that hindered them from tuning in or hitting the correct intervals, while at the same time they could have extensive vocal ranges and unstrained sounding voices. Songs with repeats apparently helped stabilize sensorimotor memory within this group.
However, both of these groups benefitted from songs with restricted complexity and tranquil tempi.
Attention and focus
Characteristic for the PPS was the challenge of staying within a consistent tonality. Often this seemed to connect with loss of focus. A change in tonality could be caused by the event of breathing combined with a leap between two phrases or, in fact, with any feature that demanded more attention. When a phrase started on the ending note of the previous phrase, it helped this group of students stay on the track. Repeats and musical forms with repeats (ABA, AABA) also appeared to be good features for mastery, as did tranquil tempi and rhythms.
Pickups were characteristic stumbling points both for PPS and other students, perhaps because of their short duration and the fact that they were the first note of the context (whereas short notes in the middle of a phrase would be easier to place), or because they appeared on an unaccented beat or subdivision of a beat.
Wide intervals and frequent leaps were demanding both in terms of focus and sensorimotor coordination.
Descending lines and repeats of the same note proved difficult for some students.
Tonality
Songs with a clearly defined tonality characterized by many tonics and fifth scale degrees appeared to be easier to sing, especially when these scale degrees appeared on the heavy beats. Tonic triads also helped the PPS establishing and staying within tonality. Songs with key changes turned out more difficult to sing.
Sequences
Sequencing is a common feature in many kinds of music, and is often considered to be easy to remember and, as such, easy to sing. Stepwise upwards sequencing seemed to stabilize tonality with the PPS studied.
However, the analysis revealed that sequencing downwards could be difficult when a leap occurred between the ending of the first phrase and the beginning of the second phrase. In such cases, the phrase was repeated from the same scale degree instead of being sequenced.
Familiarity with the song
The initial analysis of 2012 suggested that PPS performed better at songs they had known for a long time. However, in 2013 this assumption was contradicted by some PPS who performed familiar nursery rhymes poorer than other songs. Moreover, in 2013 two PPS recorded one novel song each from their last year of study. These songs were not easy in the sense of rhythmic and melodic structures, and were presented and taught in class as demanding songs. These songs were the best performances in the portfolios of the two students.
Discussion
The PPS described in this study are in the landscape between contour singing, more precise interval matching, and correct performance in terms of tonality, which equals the last three phases of children's singing development (Welch, Reference WELCH1998). However, the stronger performances within their portfolios suggest that these students are not fixed in this landscape but are still developing. The contour-interval-tonality processing hierarchy referred to in the MMP (Peretz & Coltheart, Reference PERETZ and COLTHEART2003) resembles the model of children's singing development (Welch, Reference WELCH1998). Even if it may be relevant to ask whether there are deficits in the music processing of the PPS, their stronger performances could also indicate that factors that are unrelated to biological and developmental factors play a vital role.
The melodic structure of the songs seems to be of importance when it comes to staying in pitch. The singers are more likely to stay within the same tonality throughout a song if the melody has a clear tonal scaffold of tonics and fifths, or tonic triads. Stepwise sequences in upward motion seem to have a stabilizing effect on tonality. This could be explained by the fact that sequencing is a kind of repetition, and as such is less attention demanding. In this study and in my own teaching, I have observed that singing descending lines in tune often proves to be more difficult than singing ascending lines. One reason, based on personal experience, may be that bodily tension decreases when singing downwards, while tension and focus increase in the opposite. The examples of difficult downwards sequences from the music analysis included leaps between the phrases, in which the first phrase started on one of the notes of the tonic triad. The second phrase, which was supposed to start from the scale degree below the first, was often sung as identical with the first phrase. This could possibly be due to a desire to stay within the strong tonal scaffold, perhaps combined with a weaker focus on the task because of the descending melody line. In their study, Krumhansl and Kessler (Reference KRUMHANSL and KESSLER1982) found that the chromatic notes were perceived as the most distant notes to the tonal context. This could explain why the PPS had problems with key changes, since chromatic notes are the means by which modulations are carried out.
Most of the PPS performed best in songs they had known for a long time; for instance, the most common Norwegian nursery rhymes. This is understandable due to the number of times the students have heard these songs, either just listening or singing along. Frequent exposure over a long time may facilitate the development of an inner representation of a song that will act as a support for their memory and as an adjustment tool for their own production.
However, some students performed these well-known songs poorly – so why would that be? Karen J. Wise (Reference WISE2009) asks a similar question in her study of self-declared tone-deafs. In a test of singing skills, they were asked to sing Happy Birthday; one reason for choosing this song was because it would be familiar to most people. They also sang a song of their own choice, which they performed better than the assigned song. Wise suggests one possible explanation could be that Happy Birthday was learned at early age while the singer's singing skills were still developing, so he/she may not have been fully capable of performing the song in terms of voice production and cognition. I would like to add that it is also possible that the conditions under which a song is learned can be unfavorable. In my experience, quite a number of adult role models in kindergarten or school tend to choose tessituras for singing that suit their own voices rather than the voices of the children. This will of course affect the way the child learns the song. If the child is not completely able to match his own singing to that of the adult role model, the outcomes may vary considerably. Some children may be able to reproduce the song from another starting point. Others may adopt different strategies such as contour singing, even when production of correct intervals or tonality may be within reach from a developmental point of view. Another question is whether singing activity in kindergartens or schools is so highly valued that correct performance is a goal. If it is, the adult role models would ensure that all conditions for learning are favorable, such as sufficient focus, sufficient repetition and appropriate tessituras. If not, it is very likely that the child will sing happily along believing that s/he knows the song, even when, in fact, s/he does not. Ultimately, s/he will grow up with an incorrect representation of the song and incorrect motoric schemata for it.
Lyrics are connected to rhythm. When a melody has many notes, there are often many words connected to these notes. The PPS performed the songs with rapid rhythmic passages more poorly, which could be explained by motoric and sensorimotor factors. It is more difficult to coordinate the body by eye or by ear when the tempo is fast. When words are added to fast melodic passages, this is likely to make the whole performance even more difficult. As mentioned in the introduction, occasional singers sing better when using non-lyric syllables than when using lyrics, whereas those who are congenitally amusical benefit from having lyrics. It may be that the musical function of the latter group is so poor that lyrics facilitate support from the brain's language centre, thus improving the performance. Non-amusics may benefit from a situation where the amount of information is stripped down by using syllables like ‘la’ instead of lyrics. At the same time, personal experience indicates that, once a melody has been learned with lyrics, it may demand a lot of attention to remove the lyrics and sing the melody on ‘la’. Regardless of how the issue of lyrics is treated, the amount of attention required for the given task seems to matter. For instance, one of the PPS said in her interview that she had previously focused more on the lyrics and not given the melody itself much thought. Going back to the discussion about the age at when a song is learnt, I would suggest that learning a song with many or complicated words requires a bigger amount of attention for a young child than for an adult.
Last, but not least, I believe that mental aspects of singing need to be taken into consideration when it comes to promoting or hindering development of singing skills. Sloboda et al. (Reference SLOBODA, WISE and PERETZ2005) found that people's concept of tone deafness was ‘couldn't sing,’ and that it was an innate feature. In other words, there would be no use trying to improve weak singing skills. This corresponds with my experiences from practice. Years prior to the time span of the present research I tried to get in touch with students with singing difficulties for action research, but only one student from the target group attended. Singing difficulties is often connected with shame (Schei, Reference SCHEI1998; Persen, Reference PERSEN2005) which may also be an explanation why hardly none showed up. In the present research, 25 students were offered remediation. Eight students attended the informal talk, seven agreed to singing mentoring, but after all only two participated. This suggests that low musical self-concept is hard to change. Still, three students repeated their singing exams even when two of them did not have to. In the informal talk I pointed out that since the PPS were able to do some songs at a higher level, they might be able to do more songs equally good, and encouraged them to repeat their examinations at no risk (the best try would be chosen for the overall music grade). Bandura (Reference BANDURA1997) claims that a person can build expectations of future mastery (self-efficacy) through, among other things, authentic experiences of mastery and verbal persuasion from others.
Practice is an important issue of mastery. The two students who handed in the difficult but well performed songs in 2013 must have practiced quite a lot and must have liked the songs since they put so much work into learning them. We have a Norwegian saying that practice makes perfect. This is obvious in most situations in life, but in everyday believes of many people it does not apply to singing (Sloboda et al., Reference SLOBODA, WISE and PERETZ2005). However, my experience with PPS is that practice does improve singing skills. One reason that many students have singing difficulties could simply be that they never started practicing or that they stopped practicing at an early stage.
Concluding remarks
In spite of the fact that PPS struggle with different types and combinations of challenges, I suggest in this paper that there are factors that may support vocal development and mastery.
Songs with tonal scaffolding help singers to establish and maintain tonality. For PPS, sensorimotor help is provided by tranquil tempi and the presence of repeats. This also applies to songs without physiological challenges such as leaps or register changes. Many PPS are untrained singers, so whatever is physically easy to sing is a better starting point for practicing and mastery. The issue of whether using lyrics makes singing easier or not, is a matter of whether lyrics require more or less attention, particularly with regard to whether there are many words or whether the language is unfamiliar. Along with the importance of structural features of the sung material comes the self-concept of the PPS or – as Numminen (Reference NUMMINEN2005) chooses to call them – developing singers. Authentic experiences of mastery and verbal persuasion from others may lead to increased motivation and further practice and progress.
These findings are derived from an assessment situation in which interesting data emerged from listening to a considerable number of recordings, and which led to a mixed methods study. They are also influenced by the fact that they are closely connected to educational practice since tools for self-practice in singing were developed based on the first findings and implemented in further teaching and learning activities. A possible link between the introduction of these tools and improvements in singing could have been an interesting topic for investigation. However, my main role was being the teacher and mentor of these students, helping build their skills and self-confidence so that they could qualify for their future work as kindergarten teachers. It would be unethical if my role as a researcher would intimidate the students and thus work against my role as a teacher.
It is difficult to draw generalizable conclusions from these findings because of the context, the bottom-up approach, and the methods. The fact that all of the students chose their songs themselves makes statistical comparison impossible. This also applies to the conditions under which the songs were recorded, since these varied with each student involved. However, this research may be more realistic from a musical perspective, since the student performances under these conditions may reflect their respective potentials better than performances in a strictly controlled experimental situation.
Nevertheless, the present research may contribute to the design of more effective standardized experimental studies in the future.
The findings can also be applied to and further investigated within the field of music education, both in relation to young children and adult PPS. Songs with easy-features are useful as starting points for training singing skills with both groups. Children are used to being in progress and doing simple tasks repeatedly. Adult PPS are often deterministic about their potential and need experiences of mastery in singing to be motivated for further practice. Both groups need the presence of strong vocal models and safe surroundings to be able to develop. For PPS sound files can act as models and offer the surroundings needed for undisturbed and unexposed practice.
Even if easy-featured songs should have their place in kindergarten singing repertoire, I believe it is also important to include songs that children or adults want to sing just because they like them. As Bjørkvold (Reference BJØRKVOLD1985) points out, children may very well be able to learn complicated songs, as was the case with the two PPS from 2013. However, I think this is only possible if good vocal models and sufficient practice is available.
Once a poor pitch singer, always a poor pitch singer? I would say that the findings of this research suggest the answer is no. This bottom-up study points out factors that may support singing development.
