Grounds for optimism!
Research undertaken during 2017 is suggesting that it may be easier than first imagined to recreate a “period singer”. Historical anthropometry is the study of past human body forms and shapes. Perhaps surprisingly, much of this work takes place in the field of welfare economics where anthropometric data, particularly that relating to stature, is extremely useful as a surrogate measure of prosperity. A leading exponent in the field is Richard Steckel. Drawing on his work and that in the UK of Gregori Galofré-Vilà, it has been possible to make a good estimate of how a sixteenth century boy chorister compared with one of today. They were not that different and it is possible to find living boys who might be very similar to boys who died 500 years ago.
The instrument exists. What we need to discover is how to play it!
Early Twentieth Century Boy Voices as “Period Instruments”
Between 2005 and 2015 my research interests have included the boy soprano circa 1880-1940. We know a lot about this historic voice and how it was produced. Perhaps one of the most exciting finds during the time of this research have been living boys who could reproduce this voice.
It has, of course, been possible to talk to these boys and measure their vital statistics. Physically, they are the same as any other living boys. The difference lies entirely in the way they have been taught to sing. In other words, for the 1930s boy soprano, we have access to the instrument and we know how to play it! It’s all explained in this essay:-
Here is a recording made in 2013 of a boy one month short of fourteen (as featured above) singing like a fifteen or sixteen year old of the 1930s. He is singing the first verse of the Walford Davies setting of O Little Town of Bethlehem. I have used this as a test piece several times, making comparison against a reference standard recording made by Denis Barthel in 1932.
So, can this be done for the sixteenth century?
The most common singing range for an unchanged boy voice is A3 – F5 (similar to the mezzo-soprano range). In theory, then, any living boy should be able to replicate the range of C4 – D5. This, in today’s notation and equivalent pitch, is the range of the mean – the most common boys’ part between 1550 and 1600. Alas, it’s not that simple! Sometimes it helps to start by discounting what you have grounds to believe was not the case.
We might want to discount, for example, trained boy choristers of today on the grounds that, even though popularly called “trebles” they are actually conditioned to sing soprano parts. The soprano tessitura is some way above mean and the top reaches involve extending the range to places most boys’ voices don’t go, at least to G5 and in extremis the high soprano C6. Moreover, the lowest part of the boy range, particularly the critical C4-E4 region where presence is required in sixteenth century polyphony, is commonly neglected in the training and repertoire of present day “trebles”.
Equally, we might want to discount boys in primary school choirs (where they exist) on somewhat the opposite grounds. Much of the music published for primary schools is pitched too low to develop what is called the child’s “singing voice”. In consequence, children use only their speech register. The result is a harsh tone, increasingly strained at higher pitches and the unmistakable timbre of – well, primary school singing!
It is hard to believe that the boys who sang for William Byrd sounded like a primary school choir – not least on account of the number of times the word “sweet” is used by sixteenth century writers when describing singing of which they approved. Other possibilities need to be considered – the impact of puberty, for example, or the engagement if only partially of the upper portion of the boy range in a form of “blended” registration.
Few choir directors currently working with English boy choristers seem inclined to experiment with sixteenth century pitches. A recent exception has been Bill Ives at Oxford’s Magdalen College. Byrd’s O Lord, Make Thy Servant Elizabeth is in six parts, the highest being the mean part ranging an octave from C4-C5. The boys of Magdalen recorded at a pitch thought by current research a semitone sharp of the original, actually singing therefore between Db4 and Db5 – still within the mean range. Try to ignore the counter-tenors and listen to the boys in bars 8 – 10 on the phrase “give her her heart’s desire”. Here they have to dip down to the bottom of the mean range, sharing notes with the contratenor parts. The boys’ timbre is both quite clearly distinct from the falsettist sound and strong enough for the voices not to be lost behind other parts. This, surely, is what we need from means.
Seeming to miss the point about pitch and parts, reviewer James Manheim wrote “part of the unearthly effect comes from the unique sound Ives coaxes from the choir’s boy “trebles” (sic) – this may be a disc for those who think they don’t like choirs with boys on the top part”. An interesting comment! Could it be suggesting that “sweet” might be preferable to “shrill”? Might the Magdalen boys on this recording be giving us one example of what means might have sounded like?
Testing the low range
Now, at last, we have some choirs actually participating in mean voice research. The first of these is Romsey Abbey under the direction of George Richford. Their excellent recording in the Convivium Tudor Choir Book series is now available. Pitch is set by their use of the historic St Teilos 5′ organ. Here is a short extract of Byrd’s Verse (or Second) Service, sung at historic pitch from Andrew Johnstone‘s latest scholarly reconstruction of Byrd’s manuscript.
A highly profitable visit to Romsey the week after this was recorded yielded some extremely useful voice samples and measurements that are helping with the construction of future studies. Sam Hudson of Blackburn Cathedral kindly made some boys available for further testing of the methodology developed for Romsey. A preliminary essay entitled English Choirboys Ancient and Modern can now be read on the new Fearest Voyces pages of this site. The essay is illustrated by sound samples from the research.
High tenors, contratenor altus and haute-contre
It is not only boys’ voice types that have been lost. Of at least equal importance is work on the type of adult voice that would have sung the part immediately below the boys. Today, of course, we expect some kind of falsetto voice, whether a trained counter-tenor or a bass singing in falsetto (or indeed, increasingly a female contralto in an otherwise all-male choir). Historical scholarship is now of the opinion that we should be looking for a high modal tenor, similar to but not necessarily the same as the Haute Contre.
Short sample of tenor voice quality in alto range (Valdemar Villadsen)
Of course, Romsey were presented with this problem. They chose to use Convivium’s own Adrian Green to sing the Byrd verse that in old high pitch editions would have been sung by a counter-tenor.
Andrew Johnstone and I have now formulated a plan for “Quire Pitch Evensongs” in which the boys sing as means, the lighter tenors sing the contratenor parts, the baritones sing the tenor part and the true basses really show the world what they can do! You can read more about our plans here.
Research can be a long slow, painstaking process. “Long, “slow” and “painstaking” are not words that readily spring to mind when considering the frenetic lifestyle of the modern chorister – to say nothing of the additional difficulties of finding a chaperone every time you want to re-record one phrase!
Wouldn’t it be wonderful if at least some research could be conducted with “digital boys”? Well, it now can. Pueri Epiacum is my new digital choir – created with the rather amazing software produced by EastWest sounds. The system uses sampled voices as the basis of a sophisticated programme that allows the manipulation of formants and the insertion and modification of articulations – in other words the voices can be made actually to sing. The process is not for the faint-hearted. You don’t just type words in and get a choir! It almost certainly takes longer to create a digital performance than to rehearse a live choir and teach the notes to live boys – but therein lies the point. For research, a huge amount can be learned through the process.
Here is Pueri Epiacum’s recording of the first part of the Sanctus from Taverner’s Mean Mass. Not a counter-tenor anywhere to be heard!
The boys whose voices were sampled were American – trained on three-register singing as opposed to the English “head voice only” approach. You can hear their presence in the lower part of the mean range where English choristers in some choirs at least might struggle to be heard. Listen particularly to the word “Deus” and compare with the live boys of Magdalen above.
(Page updated October 2018)