Hearing is believing?
“During my apprenticeship I did the assistant’s work – that is, I had to do my best to teach the Choristers as I myself had been taught”
A H Mann, 1899
Perception and measurement
In a book that is going to incorporate both scientific and historical perspectives on boys’ singing, we will need a good appreciation of the fact that perception and measurement can tell different stories about voices. The last half century has witnessed a phenomenal growth in voice science. Computing power has given an exponential boost to a process that began in 1855 when Manuel Garcia first began to use an adapted form of dental mirror. Digital technology has placed instrumentation of a sophistication and capability unthinkable even thirty years ago within reach of even the smallest singing studio.
Vocal pedagogy, and more specifically the teaching of boys to sing, has a long history. If we wish to know how singers were being instructed during the eighteenth century, then we have Pier Francesco Tosi’s Opinioni of 1723 and Giambattista Mancini’s Practical Reflections of 1774 to consult. They are substantial works. In them can be found the principles of bel canto that were influential in the later development of “English cathedral tone” by significant pioneers such as Zechariah Buck, long serving organist of Norwich cathedral from 1817 until 1877 (Kitton, 1899). Tosi and Mancini were both castrati, a fact of importance that cannot be overlooked, and both advocated the joining of vocal registers that they called “head” and “chest”. These archaic terms are still in quite common use today and unpicking what people understand by them in boys is going to occupy some space in this volume.
Both Tosi and Mancini were preceded by Giovani Maffei whose remarkable1592 treatise may be regarded as the first real attempt at vocal science. I draw on Maffiei’s work in later chapters when I look at what boys may have been like in the sixteenth century. The word “empirical” refers to the basis of the knowledge that has been handed down by these early pedagogues. An empiricist is one who derives knowledge from the senses, which would have to include the practical experience of working daily with singers. This definition sets the empirical approach apart from modern science, which is ever more dependent upon measurements of increasing accuracy. The ability of scientific instruments to penetrate normally inaccessible parts of the body takes us well beyond what we can observe unaided with our senses. To the extent that even simple measurement is seldom much used in daily work with boys (the present volume being an exception), it would not be unreasonable to describe most of today’s choir training as empirically based.
This has led to the kind of voice teacher/choir director conflict I described in the previous chapter. Even singing teachers can be divided into those with empirical leanings and those who base their work more directly upon scientific measurement. The former might be more likely to state that if a boy whose speaking voice has deepened reports no strain or discomfort, and no fatigue is observed in him, he can continue to sing as high as he likes. The latter might be more likely to state that boys sing too high for too long, basing their opinion on the fact that scientific measurements reveal that the boy may be at risk of injuring his technique if not his physical apparatus. I examine how such evidence is obtained in later chapters, and the question of how high a boy should sing once his speaking voice has deepened will occupy many words.
It is something of a moot point as to whether Manuel Garcia the younger (1805 – 1906) was an early modern voice scientist or more of an empiricist. His famed laryngeal mirror permitted increased use of the senses for observation, but it made no actual measurement. Of the first century of vocal science (1850 – 1950), the century of the laryngoscope, Nielson wrote:
Unfortunately, the ‘new plans’ . . . failed to result in a cohesive ‘government’ to preside over the domain of voice teaching. In fact the “scientific” literature on the voice that was published during the century following the laryngoscope’s invention does not even begin to approach the unified pedagogy of the bel canto era. On the contrary, it contains some of the fiercest controversies in the history of the field (Nielson, 2005).
This was undoubtedly true and distrust of modern voice science has not gone away in significant sections of the choral community. A problem with science is that it is always provisional. Beliefs that are not subject to the possibility of modification or complete refutation are not scientific, however sophisticated the measuring equipment may have been. We like to think that voice science has now come of age, but infallible knowledge belongs to the realm of scientism, not science.
A major early work of voice science that is particularly relevant our quest to understand boys better was that of Lennox Browne of 1896. In this paper are the visionary words:
My object has been to draw attention to the fact that, valuable as has been the laryngoscope in a physiological, as it is undoubtedly in a medical sense, it has been the means of making all theories of production of voice too dependent on the vocal cords, and that the importance of the other parts of the vocal apparatus in their actual influence on the pitch has been overlooked (Browne, 1896: 103)
This remains the case even today. Browne wrote with considerable understanding about the nature and malleability of the pharynx and the mobility of the larynx, but he did not have at his disposal the means to make measurements of what we now call formants (see later chapters). The laryngoscope had enabled investigators to note that the vocal folds (more often called “cords” or even “reeds” at the time) appeared to operate in “thick” and “thin” fold modes of vibration, the former associated with the vocal sensations commonly called “chest” and the latter with what some called “head” and others “falsetto”. As we shall see in a later chapter, the more recent work of Hirano (1981) has resulted in a more sophisticated understanding of vibrational modes that is largely what we accept today. Moreover, modern measuring instruments can readily detect the patterns of crycothyroid and thyroarytenoid antagonism described by Hirano. The larynx, however, is but one third of the vocal mechanism. Given the spectacular increase in the size of the male larynx at puberty, it is unsurprising that this phenomenon has occupied a disproportionate amount of attention in understanding boys’ voices. However, we now know that the pharynx undergoes similar transformation towards and beyond the end of puberty and the implications of this have been surprisingly neglected. We need to look, as Browne envisaged in 1896, to the other parts of the vocal apparatus and we shall do so in this book.
Empiricism reigns?
Perhaps most significant of all empirical beliefs is that boys’ voices are at their best just before they break. We saw in the previous chapter how Aled Jones and Andrew Nethsinga amongst many others hold to this belief. Though a common belief, it is an empirical belief. Voice science paints a more complicated picture. If we are to unpick this thorny matter, we must devote some space to the meaning of the word “best”. In the philosophy of value judgement, the word “best” has more than one meaning. At its most subjective “best” can simply be the highest in an ordering of personal preferences. More objectively, it can be a value judgement arrived at after careful consideration of competing options, each with some degree of merit. Most objectively, it an efficiency judgement, and quantifiable at that. An empirical judgement that such and such a voice is “the best”, or a very good one, will likely be a personal preference. Several judges in an eisteddfod may agree that a particular voice is “the best”, in which case there will be criteria beyond any individual preference, but these may not meet measurable scientific criteria for efficiency. The eisteddfod judges are therefore working with an empirical framework. John Cooksey made this important statement about a boy’s unchanged voice at the “height of the prepubertal period”:
Full, rich soprano-like quality. Reaches its pinnacle of beauty, power and intensity. Range expansion of the childhood voice at its maximum. (Cooksey, 1992: 55).
It sounds here that he is talking about a boy’s voice “just before it breaks” when according to empiricist observers it is at its best, but he is not. He is talking about the voice of a 10 – 11 year old, identifying “optimum period” as “grade 5 (near end) to grade 6”. That would be Y6 to Y7 in the English system. His words were written in 1992. If puberty has advanced a little since then, (see Chapter 4) then the optimum period may have correspondingly shifted perhaps to nearer the beginning of grade 5. In the English system, therefore, most of Y6 and the beginning of Y7 is when a boy’s voice is at its “best”. I shall explain and demonstrate in later chapters that Cooksey has objective, readily replicable criteria that show how a well-trained eleven year old voice is indeed working at a level of optimal phonational efficiency. This has demonstrably declined by the age of thirteen but thirteen, for the majority of boys, is still the age “just before the voice breaks”, so we have an issue here.
Alan Thurlow was director of music at Chichester Cathedral from 1980 until 2008, and a notable advocate of boy choristers – to the extent that he became a vice president of the erstwhile campaign to keep girls out of cathedral choirs. I came across an interesting piece by him in the journal of that campaign and I think it summarises very well the empiricist understanding. It is well worth reproducing at length. Interestingly, he uses the term “change” rather than “break”.
In my experience there were two ways in which the change could occur. Either, as Donovan Peters expressed so well, the voice moved gradually towards ‘a pure head register with an immature young man’s fundamental register under it’ (in which case they usually kept a complete range of usable notes), or the voice stayed more or less the same but with the upper notes gradually disappearing one by one and without any obvious sign of the emerging falsetto. In the first way, and unless the tone became too ‘hooty’, it was usually possible for a boy to sing on and to continue enjoying the experience. Some managed it so well that the change was hardly appreciable to the listener. The most apparent difference was the loss of power in the lower registers of their treble compass (the ‘chalumeau register’ as I used to refer to it). The rich and powerful low notes which they could previously sing gradually became weaker, and more difficult to produce, as they moved down towards the emerging ‘break’ between what was becoming their falsetto voice and the lower, natural, tenor notes that were developing. The younger voices in the choir had plenty of power in that range, so in the overall ensemble you were not really conscious of the loss of power in the older voices. However, it did of course restrict the kind of solos that these older boys could still sing…….Reverting to the second way of the changing voice, I used to describe this (particularly to concerned parents) as ‘the ladder’. The rungs represented the notes of the voice range. For years the chorister had been used to climbing up the ladder with confidence, but one day he discovers that the top rung is missing. Gradually the lop rungs start to disappear one by one; this means that he now begins to tread cautiously as he approaches each of the remaining high rungs, in case this time they are not there, and this in turn starts to affect his confidence in his voice. Generally speaking choristers were happy to stay on when their top note was, say, still up to treble G. They could happily mime the odd ‘a’ flat and above, and it was interesting to see how they could still, by pure dint of confidence and leadership, bring their side in reliably on a high-note lead whilst not actually singing the note themselves.
A significant change since Thurlow’s day has been the arrival of the ubiquitous voice coach. Every choir now, it seems, even those in parish churches, must have a singing teacher or voice coach who works with individual choristers. These were not at first universally welcomed by those of Thurlow’s generation. Roy Massey, late of Hereford Cathedral, confided the following to me a few years after his retirement:
I just produced the sound that I found the technique of doing it and I always produce the sound I wanted to hear and half the time it was just the way I liked it really, there’s no more theory than that . . . It’s a great shame it all started getting bogged down with theories really and people were brought in as voice coaches. We have voice coaches these days, they worry me . . . Because I’m sure there’s some very good voice coaching going on but are they producing a sound of the choirmaster’s creation or are they producing the sound of the voice coach? I wouldn’t have wanted anyone near my boys because they might spoil it, they might alter what I want to hear (Massey in Ashley, 2008: 97).
The contrary singing teacher’s stance is put forcefully by Janice Chapman who subscribes to the view that cathedral musicians are trained in organ playing and possess little serious knowledge about singing. She is troubled by her experience of young adults with defective technique which she attributes to their years, first as choristers and then as Oxbridge choral scholars. She has been unreservedly forthright on this issue in personal communication and is not constrained in print.
The approach contained suggestions such as ‘stand straight’, ‘use energy’, and ‘be alert’ . . . the choral director took almost all rehearsals and services though were sometimes deputized for by the senior organ scholar. The choral director was basically an organist with choral skills but not personal vocal proficiency or training. TUNING was always a priority but never seen as a result of voice production . . . rather as an aural or mental problem. The college provided a singing lesson each week (1 hour) for choral scholars when individual vocal problems were meant to be addressed. The recommended voice teachers were often from a similar academic and musical background (Chapman, 1995).
The implication here is that tuning and a great many other matters might be improved through the application of scientific knowledge of voice production, as well they might be. The difficulty is communicating scientific ideas that may sometimes sit uneasily with a thousand years of empirical tradition. The problem is twofold. Knowledge and understanding gained by voice scientists has first to be accepted and absorbed by the teachers and conductors who work with boys. If a change in time-hallowed practices is implied, there may be significant challenges. If these teachers and conductors accept the challenges, they must find ways of communicating the knowledge to the boys they work with. Much of this will be indirect, in the form of the games, exercises and metaphors that are used in choir training and the teaching of singing. Skilful mediation between the language of voice scientists and what the twelve-year-old mind can assimilate does need to be part of the training of musicians working with boy singers.
My experience of training primary school teachers to teach science (see Ashley and Nicolson, 2008) as well as secondary music PGCE students suggests to me that part of the problem may be that a good many teachers and conductors simply do not have the knowledge themselves. This is confirmed by evidence that boys, particularly those aged between twelve and fourteen, complain that they are not taught about their changing voices. If they are, it is more likely to be in a biology lesson than in a choir rehearsal where the knowledge needs actually to be applied. Whilst it probably remains true that boys still learn the most simply through imitation of older boys in the choir, it is a serious failing to underestimate boys’ desire and need to know, and ability to understand.
The area where the most progress has been made is almost certainly that of vocal health. Olivia Sparkhall has recently written a text full of good advice to young people about how they should care for their voices. The information is accurate and the language, layout and illustrations are clear and at the right level for the readership (Sparkhall, 2023). Other texts such as Williams (2013) communicate necessary information at a level that is right for teachers and conductors. Difficulties potentially arise when the science of vocal health comes into conflict with the empiricism of choral practice which, as we saw in the previous chapter, it can do over the question of how high boys should sing.
How hard is measurement?
If we accept that voices change throughout the life course, and that adolescence is a time of accelerated change, particularly for young males, it ought to be possible to measure that change. If we can measure changes and we have a good awareness of what any measurements mean for our teaching or choir direction, we might act on them. We should then see results in the way boys sing. How hard, therefore is it to make the necessary measurements? We shall see that some measurements are in principle easy for one whose primary qualifications are in music to undertake, though there may be practical difficulties of administration in a busy choir. Other measurements require moderate investment in a tool such as Voce Vista. Yet others are sufficiently difficult either to defeat scientists altogether, or to cause scientists to use surrogate measures of arguably dubious provenance. No measurements are of any value unless they are related to a body of knowledge that links theory and practice.
Although we know a lot about the change of voice during adolescence, there is still a lot that we either don’t know, or know only on the basis of a small number of studies which in turn have been of only a small number of subjects. There are two measurements that are both critical and difficult to make. The first is that of a boy’s progress through puberty and the second is that of the growth and development of a boy’s larynx and vocal tract. I shall explain here why these measurements are difficult and in subsequent chapters describe attempts that have been made and what the consequences are for vocal pedagogy, choir training and historically informed performance.
Progress through puberty is difficult to measure because it requires an intimate physical examination that can only be performed by a suitably qualified paediatrician. The reference clinical standard is that of the volume of the testes. I shall refer to this measurement by the abbreviation TV. Unless a paediatrician able to make such measurements is willing to collaborate in a study, approximations, estimates or surrogate measures are going to have to be relied upon. Measurement of testosterone concentrations offers an alternative but correlations with TV have not always been good. A well-established system that has been in in common use for some time is that of five stages devised by J. M. Tanner (more details in the next chapter). This requires not actual measurement but visual observation of a boy’s genitalia and can be carried out by a GP if there are genuine concerns about precocious or delayed puberty. Again, this is an intrusion into privacy that boys will resent and find highly embarrassing, so it is unlikely that such inspection is going to be made in a study of voice, still less in day to day choir work or singing teaching. A surprising number of studies across a range of disciplines requiring correlation of one accurately measured parameter with pubertal status have therefore resorted to a system of self reporting based upon the five staged drawings produced by Tanner. The reliability and validity of self reporting (or reporting by parents) has been shown to be poor (Coleman and Coleman, 2002). It is important in the evaluation of a study to determine how claims about the pubertal status of subjects have been arrived at. Other non-intrusive measurements that I shall describe shortly are probably better than self reporting of Tanner staging.
The other measurement that is difficult is that of larynx and tract development. These measurements, if anything, are even harder to make than measurements of the testes because they require invasion of the body, and because the body parts that require to be examined are tiny and complex. Many texts on singing refer to the length of the vocal folds, the inference being that voices deepen during puberty because the folds increase in length. This is an easy concept to understand and can be explained to boys perhaps through analogy with stringed instruments. A violin sounds high because it has short strings and a ‘cello sounds lower because it strings are both longer and thicker. I have been “guilty” of such over-simplification myself in the past (Ashley et al, 2008). The truth of the matter is that our knowledge of the growth of the vocal folds is less than we think. Until fairly recently, all our knowledge came from a very small number of cadaver studies (Kahane, 1978; Hirano, 1980, see next chapter). A more recent study carried out on anesthetized living children has demonstrated that assumptions that have been made about vocal fold growth during puberty appear to be misleading and in some cases just wrong (Rogers et al, 2014). It would appear that the pubertal growth spurt of the vocal folds is less than previously thought. Other changes in fold composition about which we currently know very little still need to be researched. Meanwhile the actual measurement of a living child’s vocal folds is a procedure so invasive, delicate, and requiring of such highly specialised skill that it is likely to be some time before such a study is replicated. We need, in the mean time, to be perhaps a little more circumspect in what we say about the growth of the vocal folds.
The other difficult but necessary measurement is the growth of the vocal tract. We do know that an adult male pharynx is significantly longer than that of a child or a woman, but pharyngeal growth is the poor relation of laryngeal growth in many texts on adolescent singing. This strikes me as odd since timbre is such an important topic in the discussion of boys’ voices and the pharynx plays such a significant role in resonance and timbre. Again, it is a difficult measurement to make. X-rays have been used in the past but such methods could not be countenanced today. In the next chapter, I describe in a bit more detail the few studies that have actually measured growth of the vocal apparatus and puberty.
The better news is that there are much simpler measurements that are non-invasive and adequately informative for most purposes in boys’ singing. If what we are trying to understand is related to growth during adolescence, the simplest way to measure growth is through measurements of height and weight. Provided such measurements are made accurately, they can be sufficiently informative. For example, good correlations have been shown between increase in height and increase in pharynx length (Fitch and Giedd, 1999). Of course, to measure growth systematically it is essential to take measurements at regular intervals. Doing so opens a quite powerful window, that of growth velocity. Most people are aware that there is a growth spurt during puberty, but I shall show in the next chapter that if velocity, the actual rate of growth, is mapped systematically it can prove a surrogate measurement arguably more reliable than self-reporting of Tanner stages. The other change that we can readily measure is that of voice deepening. The range of a boy’s singing voice can be highly misleading for reasons I describe in detail in Chapter 5. However, the much more limited range of pitch across which he habitually speaks has been shown to correlate well with other measures of puberty, including growth velocity. Provided we know how to measure it with sufficient accuracy, and provided we fully understand what these measurements actually mean, there is no need to fret about intrusive measurements of puberty. We have all we need. I devote the next chapter to a study I undertook in collaboration with a consultant paediatric endocrinologist and explain in more detail the relationship between speaking voice pitch and other measures of puberty.
Conclusion
There is little doubt that, before the dawn of modern voice science, some boys at least sang with great beauty under the direction and tutelage of those I have called the empiricists. It would be foolish to ignore the accumulated knowledge and wisdom of centuries. The writings of the pre-scientific theorists are also worthy of study. At the very least, they open possible windows into historically informed performance. We can also trace through them a course of development in vocal pedagogy and, to the extent that people still use terms such as “chest” and “head” voice today it is valid to understand what the theorists of previous centuries believed about topics such as vocal registration. To plough on blindly as though tradition were an infallible guide to how things ought to be done certainly cannot be right, but if we cannot make a distinction between perception and measurement, that is what we are destined to do.
