voices at thirteen
On this page are audio recordings by boys who have reached that critical turning point of moving on from being treble. It is a data archive for the use of those wishing to study the adolescent male voice as it changes, and a resource for illustrating my books and papers.
1000 Voices across three phases (cross-sectional study)
The paediatrician J.M. Tanner proposed five stages of puberty that are still accepted as a normative standard today in clinical work. For his “eclectic” scheme John Cooksey inserted an extra stage, making six. Michael Fuchs preferred three phases in his work with the Thomanerchor of Leipzig. As a result of my own collaborative work with Professor Gary Butler, we have similarly adopted a three-phase system in Building Voices in preference to five or six stages. Here is a non-technical explanation. Below are several contrasting examples of each of the three phases taken from my cross-sectional study of 1000 boys’ speaking voices. (There are over 1000 of these samples, not a lot to be gained by posting them all, though of course they are all included in the quantitative analysis and statistics). See here for a table mapping phases and stages against definitive measurements of puberty, and here for a line graph showing the progress of one boy from the longitudinal study throught the phases.
Boys develop proficiency in singing throughout childhood. By the time the speaking voice pitch has fallen to anywhere between 220 and 260 Hz the boy is technically still a child, but voices that have been taught to access a singing rather than speaking register have been described by John Cooksey as optimal, and the "climax of beauty and fullness". Childish sweetness or "innocence" is still present, leading some listeners prefer the weightier and "fruitier" sound of the phase one voice (see below).
Pitch range of childhood phase 200-260Hz
Tanner stage 1.
The voice will be a little deeper than that of a young child, but this deepening is not at first attributable to puberty. Most boys will still be soprano/alto, but preparations for change need to be in hand. The term "peripubertal" can be used to describe this time when boys are likely to approach and just cross the threshold of medical puberty.
Pitch ranges of phase approx 195 - 215Hz
Tanner stage 1 , beginning stage 2
Initially, the voice sounds slightly deepened to most ears, but still retains a boy-like timbre. This is gradually lost as puberty is now having an impact. Ideally boys should have a cambiata part, but if not possible, low alto is likely to be nearer than tenor. Onset of instability marks the end of the phase, when the voice starts to deepen much more rapidly.
Pitch ranges of phase approx 170 - 199 Hz
Tanner stages 2 and 3
The voice contunues and accelerates its rapid descent and many boys at first have difficulty adjusting their singing. "Boy-like" timbre is lost from the whole range which now sounds "young-man" like, causing some listeners to say it has "broken". The term "emerging baritone" conveniently covers what most boys will be able to do.
Pitch range of phase approx 99 - 169 Hz
Tanner stages 3 and 4
Plangency, Sonority or Sweetness?
At what age do boys give their "peak performance"?
Claims that boys’ voices have unique qualities that cannot be replicated by similarly aged girls are regularly made, though equally regularly contested. Traditionalists have recently highlighted the age of thirteen, suggesting that it is the uniquely “sonorous” and “plangent” qualities of older boys that cannot be replicated by girls. However, a cathedral style choir is a mixture of different ages, thirteen-year-olds being a minority. Voices change in timbre between the ages of eight and thirteen, and boys mature physically at widely different rates. Thirteen is the age at which the greatest diversity in physical development is found. These are amongst possible reasons for the failure of traditionalists to make a convincing case. Read a more detailed argument here. This section of the website draws on some of the longitudinal case studies to illustrate that whilst any individual has a unique individual genotype of voice (much in the same way as the individual fingerprint or iris pattern are unique) the expression of vocal timbre as a phenotype changes as a boy grows and interacts with the environment that nurtures his voice.
Given the significance of the age of thirteen, I have selected a recording made at that age first. Other recordings made at younger and, where they exist, older ages are then presented for comparison. The extent to which each singer had progressed through pubertal voice change is known from the measurements regularly taken and, needless to say, this was a quality that varied significantly when the age of thirteen was held constant. The terms “prepubertal”, “peripubertal”, “in-puberty” and “completing puberty” ae used and explained and justified elsewhere on this site.
An unusually good and complete set of recordings exists for this cathedral chorister from the age of 08:06 to 13:11. Perhaps the Candlelight Carol, recorded at 13:04, was a peak performance? At that time, the completing puberty phase was just beginning.
This is the most complete longitudinal data set and more examples are found below. Here, the Tomkins alto verse anthem was recorded at age 13:03. The peak treble performance was a year earlier. This is an in-puberty recording that avoids the use of a falsetto necessary to retain the treble range.
Another cathedral chorister, this time winner of the BBC Chorister of the Year competition. A busy schedule of concert appearances, including a Balshazzar's Feast at the Albert Hall followed . Danny Boy, given in his home city when aged 13:08 is chosen as a peak performance .
Look at the World was recorded twice, at ages 09:04 and 12:09. The nine year old recording is that of a prepubertal child, whilst that at 12:09 is late peripubertal at the very cusp of in-puberty. These two recordings may be regarded as definitive. The first sung competently and with assurance, but with clearly childlike timbre. The second may well be recognised as the mature voice of a boy chorister, quite possibly the “more plangent tone” of which Giles writes, yet both would have contributed to the collective sound of the boy’s choir.
The last treble recording was made at age 13:11, well into the completing puberty phase. The recording of Lift Thine Eyes is multi-tracked, demonstrating a full range from alto to first treble. The is an be an “expanded” range as described by Henry Leck .
What actually constitutes the peak treble performance is a subjective judgement. The Dowland was recorded at age 12:06, which was the last full "treble" session when the voice was late peripubertal. The Britten was recorded at 11:10 before any growth spurts had occured - a prepubertal voice that had had time to develop enough musicianship to cope with an "art song" in the style of Pears. There are other examples on the CD.
By the age of 12:11, it was clear that falsetto was necessary to maintain the full treble range. The Darke carol was recorded in the most comfortable key for the voice - A major (lowest note G#3). This transcript was made at the time.
Recorded at age 12:06 in this case, another multi track demonstration of the expanding voice. This is the time when the practised singer gains new notes at the bottom of the range without losing the top.
Sadly, there are no recordings of the voice at younger ages available. As is often the case with competition wins, opportunities opened up, including a major recording contract involving "six of the best young singers of the UK" - three boys and three girls. The boy was 14 by the time this happened and well through the "in-puberty" phase. The two short extracts below allow comparison of a boy's voice at this stage of adolescence with that of a similarly aged girl. It would be hard to disagree with Giles's assertion of "plangency" and "sonority" against "sweetness", though conclusions cannot be drawn from a sample of two!
Much activity took place during that year. The last two samples show changes to the treble timbre between the ages of 13 and 14. The first is of the Britten Hodie recorded at the same concert as the Londonderry Air. The second is a studio recording of the same piece made for research at age 14:02, by which time the speaking voice pitch had fallen to 178Hz.
Finally, two early recordings, that made at eight being a child with obvious promise whilst that made at ten is of a young chorister making a significant contribution to his choir. If the notes are accurate (which they are!) should the younger voices not also be acknowledged? Choirs are places where boys both learn and perform.
In complete contrast, the earliest recording by this singer, of a piece he already knew well through his choir membership. This, like the similar recording opposite, illustrates my point that a cathedral style choir is a mixture of ages, and therefore a mixture of both sweetness and plangency.
Fuller details, including spectral analysis of a single vowel are given here.
This is another comprehensive data set for a boy who became head chorister of his cathedral at the age of 14. Though a regular soloist, no recordings had been made by the cathedral by the time the growth spurt marking the transition from peripubertal to in-puberty had occurred, so a studio session was arranged to record some of his favourite pieces, of which the Britten Corpus Christi carol was one.
The next two recordings, as with Case Three allow a comparison of the same piece at different phases. Perceptually, the in-puberty 13:03 version sounds once again sounds edgier or perhaps more focussed than the peripubertal recording made fifteen months earlier. The two recordings, being made under identical controlled clinical conditions, are almost the same in total amplitude. A spectrograhic analysis is provided here.
It is well established that all boys pass through the same phases and stages of puberty in the same order, but that the time spent in each stage can vary significantly. The peripubertal phase for this boy was longer than for Case Four opposite. No in-puberty growth spurt or significant speaking voice deepening had occurred when the Minstral Boy was recorded at almost the same age.
Unlike Case Four, this choir recorded several times and this boy was featured as a soloist, at the age of 12:02 in a CD featuring the whole choir, and at 13:11 in a Christmas CD featuring the two boys who had for some time been the choir's leading soloists. The 12:02 voice is early peripubertal and it might be thought that leaving a commercial recording until 13:11 might be risking pushing the voice through the in-puberty phase, as in Case Three. This was not the case here. The growth spurt commenced at 13:08, but the medical in-puberty threshold was not crossed until 14:03. A close run thing!
This cathedral choir permitted boys to stay beyond the normal choir school leaving age and he was still singing a falsetto treble at age 14:07 when the speaking voice had fallen to 145Hz, a value associated with the early completing puberty phase. By his own admission, his voice was “straining” at this time, but this he regarded as a price worth paying to remain in the choir as long as he possibly could. All of the boys in this choir were recorded for another investigation when he was 13:09. The Vaughan Williams This is the Truth was the piece chosen.
An alternative could be to move to alto, but this is very rarely done in English cathedral choirs where adult counter tenors are preferred. The recording of Gibbons' Record of John made at age 14:01 is a "might have been" that can be compared with the Tomkins verse anthem in Case 2. The boy himself, though willing to cooperate, regarded the whole exercise as a little quaint and eccentric!
Soprano, Treble or Cambiata? (longitudinal studies)
Longitudinal studies of boys’ voices are relatively rare. Much of what it is written about adolescent vocal growth is derived from cross-sectional studies because, as Hollien et al remarked in their exceptional longitudinal study “longitudinal research of the type to be reported is immensely labor intensive”. In consequence “the data provided are averaged on a group basis rather than by individual”. It is rare, therefore, that the life course of an individual boy will fit norms derived from data that are mainly cross-sectional. Many studies reporting general growth have been “ been based on protocols where the target factor was measured only once or twice a year”, - not really adequate for understanding voice. Cooksey’s most famous study was longitudinal over a three-year period with monthly sampling. However, it suffered from the drawback of taking an inadequate cross-section of all the different approaches, techniques and styles in use around the world. My own studies have attempted to build on Cooksey’s by addressing this potential shortcoming. The albums below are not presented for the primary purpose of listening or music appreciation, but for what can be learned from boys taught in different ways and followed through their early adolescence. Some tracks are made available on that understanding - that they are indeed "voices for science" and a learning resource. This account should be studied carefully before listening to the tracks.
Louis Alexandre Desire, from Paris, is the boy with the lute on the front cover of my post-doctoral monograph. Louis had quite a career as an exponent of bel canto boy soprano technique. As a soprano, he recorded quite an extensive discography extending from childhood to mid-adolescence. In 2007 he gave a song recital as part of my inaugural professorial lecture – possibly a unique occasion! Four pieces from that recital were issued on the disc Il Passagio. Recorded at the same time was a demo disc called Swing and Sadness – an attempt to add a bit of “swing” to classical repertoire, but sad because the boy soprano voice flowers only briefly. During his visit from France, I “fixed it” for Louis to be the soloist in Hear My Prayer sung at evensong in Bristol Cathedral. It was his ambition at the time to sing the solo made famous by Ernest Lough with an English cathedral choir! The tracks below chart the developmental course of this boy soprano voice.
Monday Afternoons stands in direct contrast to Il Passagio. It attempts to chart the course of a typical English “treble” of the early twenty-first century, if there is such a thing. Max Matthew was recruited at aged 10 – his singing future at the time largely an unknown. Uniquely, samples of a wide range of parameters were taken at monthly intervals. Demo recordings were also made regularly to capture any subtle nuances or differences resulting from growth. We decided to issue these recordings as a professionally produced CD, which is still available. The story took an interesting twist as Max’s career became split between the National Boys Choir of Scotland and the RSCM Northern Cathedral Singers. The approaches to dealing with young adolescent voice could hardly have been more different, if illustration of the fact that there is no such thing as a “typical treble” were needed. The tracks below continue beyond Monday Afternoons into Monday Twilight, right up to age 16.
According to Irvine Cooper, the youthful Wayne Newton was a perfect example of a cambiata voice. Here in the UK, I still cite Dominic (Inigo) Byrne. Dominic was originally a chorister at Chester Cathedral, but his ambitions for a singing career led to a disagreement between the choir director and the London vocal coach. When faced with the choice between a professionally coached commercial recording and the regime of the cathedral choir, he opted for the former. Paradoxically, Dominic never knew he was a cambiata, though the fact that he was coached to be a perfect example by a teacher who did not use the term speaks volumes. Dominic’s identity change was studied in minute detail, though unfortunately no physical growth measurements were taken. Accordingly, I have included other boys from the longitudinal study whose voices came to be of similarly high cambiata quality. Dominic features in the BKS film Riding the Changes.