Singing polyphony

Author(s): PETER PHILLIPS

Source: The Musical Times , WINTER 2014, Vol. 155, No. 1929 (WINTER 2014), pp. 7-18
Published by: Musical Times Publications Ltd.

Stable URL: https://www.jstor.org/stable/24615654

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Musical Times Publications Ltd. is collaborating with JSTOR to digitize, preserve and extend
access to The Musical Times

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
 PETER PHILLIPS

Singing polyphony

to life? I think the answer lies in a general misunderstanding of its

Why do nature,
singers findthis renaissance
and the demands makes on its performers. polyphony so difficult to bring
For me the litmus test of this difficulty is experienced almost every time
I ask a singer to sing some plainchant. No matter how well-trained, he or
she will have routine problems identifying each note and ending precisely
in tune (the last two notes of a phrase are often pitched too low). I hear my
colleagues say that singing chant is the ultimate test of their craft - not only
the tuning, but the speed with which to deliver it, the appropriate colour
of voice, the phrasing. To do it well requires a naturalness of delivery, a
gentle melodiousness of sound underpinned by a complete confidence in
the placing of each note, which stands at the opposite extreme of what the
traditional voice-training programmes still to be found as standard in the
conservatoires throughout the western world are likely to teach. If you study
voice in one of these establishments it will be assumed that you (really) wish
to sing opera, which means projecting your voice over an orchestra into a
vast, dry space. This in turn means singing loudly, which leads to vibrato.
The precise pitching of notes is subsumed into a splashy soundworld in
which the singer is expected to project a strongly individual personality,
predicated on an emotional tension which has to remain heightened to
remain interesting.
All this represents the reverse of what good chant singing requires.
Church music before the advent of baroque performance practice was sung
quietly, or at least gently. There was no orchestra to raise the decibel levels
(there were no instruments of any description most of the time, and certainly
not when singing chant), and most of the specialised singing took place in
smaller spaces than one might imagine - either private chapels, side-chapels
in cathedrals and abbeys, or in their quires, where the monks or nuns were
gathered for reciting the hours. There was rarely any need to fill a cathedral
nave with sound, and when this was done, on the highest days, the chant
would still be sung gently, its clean contours audible in the clear, if not always
reverberant acoustics which obtain in most large stone buildings. And to
sing gently inevitably removes much of the crisis associated with vibrato,
the dreaded V word which haunted my youth. Voices do not vibrate so
obviously when kept within natural bounds, though to eliminate all vibrato
is also not entirely natural, leading to a 'white ' sound in performance which
goes against modern standards of taste even in the singing of chant. (What

the musical times Winter 2014 7

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
8 Singing polyphony

was thought of it in the renaissance period history doesn't relate, though I

would give much to find a reference to a monk or nun being reprimanded
for singing in some way inappropriately - too loudly, too self-importantly,
too vibratofully).
The main problem with vibrato, it should go without saying, is that it
is inherently out-of-tune: something which can only be tolerated in sing
ing chant within a tiny spectrum. Why badly out-of-tune singing should be
tolerated in any branch of the singing profession is a matter of wonder to
me, but it is part of my thesis that the extreme vocal discipline required to
sing chant well is still regarded as something beneath much contemporary
vocal tuition. For most professors of singing it is too self-abnegating, the
emotional response too hidden for the modern world, too miniature an art
form for those with wide eyes for the big stage. However, it is also part
of my thesis that the hidden emotion which can be felt from chant and its
derivative, polyphony, is the exciting part of this corner of western classical
music, and something we can make our own. Current thinking can easily
assimilate an understated artform: writers like Robbe-Grillet; painters by
the dozen; architects like Mies van der Rohe; minimalist musicians working
in the contemporary sphere like Part or Reich, have been giving us exteriors
which are formally perfect, but cool and even superficially uninviting, in
conformity with van der Rohe 's axiom that 'less is more '. Yet somehow the
best of them draw us in. And once drawn beneath the surface there is an

extra frisson of understanding, which can make more obviously presented

performances seem disappointing. Whether you put opera into this category
of shortfall is not my concern. The problem on hand is that the psychology
of singing is more rooted in old-fashioned concepts than any of these other
artforms and, as a result, at least one part of the profession associated with
it - pre-baroque singing - has been hobbled. And since there has been
little opportunity for the performance practice of these repertoires to fully
mature, so the public has not been encouraged to fully appreciate them. I
would say singing-culture and choir-culture are languishing at least 50 years
behind instrument-culture, an immaturity sponsored and maintained by the
dominance of the extraordinarily powerful world of opera.
Whether the composers of renaissance polyphony would have recognised
any of these ways of thinking is of course unknown. Hiding emotion
probably wasn't what they were doing deliberately, and the excitements
available to us from discovering vibrancy beneath a surface - most likely
to be appreciated by contrast with some sort of opposite like opera -
almost certainly wasn't available then. At least their opposites were both
fewer and of a different kind from the countless ways modern society has
invented of dressing up the surface of things: they were more preoccupied
with perceptions of moral chaos, of the workings of the devil, of sharp

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
definitions of good and evil. Beauty for its own sake was highly problematic,
especially for a certain type of religious extremist on both sides of the
Catholic/Protestant divide. Yet the music of that world is there for us to
make what we can for ourselves, in our own context. We are entitled to do
this — for example to sing Requiems in 3000-seat concert halls to a paying
public — in the same way that we string up the paintings of the period next
to each other in galleries. It is not authentic, but then neither is the choral
sound we make (I assume). However this giving up on strict authenticity
is not an invitation to do whatever we like; in fact quite the opposite, as so
many choirs have discovered. The music must be respected, and it demands
difficult things both of our singers and of our powers of interpretation. It
is a new music for us, for which we need to learn new techniques and undo
old preconceptions, which serve it so badly.
I opened with the singing of chant in trying to identify these new
techniques because polyphony so obviously and directly descends from
chant. Not to proceed from this starting-point when trying to find a suitable
modus cantandi is to make a common but fatal error: if a style is judged by
what came afterwards, rather than by what went before, it will inevitably be
seen through a distorting lens. It has been the easy habit of musicians from
the 17th century onwards to view renaissance polyphony as a primitive form
of baroque writing, regrettably lacking in certain basic commodities which
had been established as essential for a full expression - like instruments and
coloratura soloists - but which nonetheless had some value both in itself

and as the parent of the splendours which were to follow. By the time we get
to the 19th century there were those who, in an antiquarian way, were keen
to identify what might have been individual about the renaissance style,
but in romanticising it redoubled the aesthetic connection it had formed
with operatic singing. The scene in Pfitzner's Palestrina where a choir -
standardly the opera chorus, singing as if the music were by Verdi - is asked
to sing an extract from Palestrina's Missa Papae Marcelli shows at every
performance I have ever come across precisely how little has been learnt
about what makes this music moving. In the 20th century, despite the huge
burgeoning of interest in polyphony as an independent repertory, the mind
set, or perhaps I should call it the throat-set, of most professionals has been
to tack its performance practice onto baroque practice, but self-consciously
simplify it in a way which is seen as fitting for a simpler style.
The reason why it has been so hard for modern performers to apply the
necessary unpicking of later styles in singing polyphony is because the
world which sustains them professionally came into being with the baroque.
If they could visit Italy even as early as 1610 they would know what was
required of them to make a living. If they were to go back to Italy in 1550
they would find it much harder: they would effectively have to relearn their

the musical times Winter 2014 9

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
io Singing polyphony

trade, which is what modern specialists also should do. The reasons why
they may not be keen to do this are twofold: this kind of modest singing
attracts less money than solo performance; and, more subtly, modern concert
conditions may require polyphonists to project into big symphony halls,
which in turn means they will have to sing louder than I think was normal
in the 16th century while still presenting the contrapuntal lines without
distortion. This type of singing is not only counterintuitive for us now, it
also requires lengthy training of a unique kind. Most singers smudge it.
What most singers tend not to do is rethink themselves as performers
of chant. If they were prepared to take that risk they would learn many
things, both stylistic and intellectual. On the intellectual side they would
come to understand how the words relate to the music, both in chant and
polyphony. One of the distortions bequeathed to renaissance music by the
baroque is the idea that every word must have a passionate meaning, and
that any setting worth performing must reflect that meaning. Admittedly
during the High Renaissance (from about 1575 onwards) there was a gradual
move towards this way of writing, through madrigals and the intermedii
to the first operas, and the composers who made the initial experiments
(Lassus, de Rore, de Wert, Marenzio and Gesualdo amongst others) are
the ones who are most referred to even now when renaissance sacred

music is discussed. But although these composers are historically important

and endowed with a certain glamour, the vast bulk of the greatest polyphony
is by more traditionally-minded men, like Josquin, Palestrina and Tallis, in
the style I'm describing, which carries directly through from the past.

through the entire 24-hour cycle of the monastic hours, especially

To geton to the bottom
an emotionally-charged oflikethis,
feast-day there
Good Friday. Clearlyis
the no better education than to sit
words have the utmost significance - none ever had more, if you believe
in what you are singing — and yet, in the countless hours of presenting so
much music in eight separate services, it is not possible or desirable to accent
every important word. One soon learns that chant is the ideal medium for
preserving the dignity of the texts while moving through them. There is
intense meaning, but it is not played up or dwelt upon because to do that
would not only condemn the participants to completely sleepless nights and
days as a result of the sheer length of the liturgies in question, it would also
deprive these services of an accumulative effect of meaning. The music
and the texts combine in these services to create an atmosphere which is
intensely moving while being understated: which is what much of the best
polyphony can do. It follows, then, that it is this understated atmosphere
which needs to be found in modern performances, and it is most likely to be
found by starting with the procedures of the monks and nuns.

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
 To understand why religious communities sing chant as they do it is
necessary to understand what makes chant up. This changes little from
service to service, or from text to text: its essence is melody with no implied
harmony and very little rhythm. It is a melodic art-form, sometimes syllabic,
sometimes with soaring melismas, but always with a dignified tread and a
sweep of phrase which manages to be both expansive and efficient. The
words are the essence of the matter, but they are not paraded as such. They
have been subsumed into the melody, which carries them further and more
beautifully than reading them does. The two have become one.
Some of these melodies, usually the more melismatic ones, are extremely
long, which calls for special techniques from the singers. The essence of
this is to maintain a good legato, so that the sweep is maintained, which is
managed partly by staggering the breathing between the singers so there are
no perceptible gaps in the sound; and partly by singing through each note,
so that each acquires its own proper weight. The result should be a flow
which betrays little obvious shape, apart from swaying up to the high notes
and away again to the lowest ones, an endlessly plastic song filling an often
reverberant space with waves of sound. Compositionally there will be little
to emphasise apart from the subtlest internal references, and there is often
no very predictable beginning or end.
Good chant singing will therefore depend on singing through the lines
with a sense of direction towards the next natural break. In addition, it is
always useful to understand how the melody has been constructed: any
repeating material, like sequential motifs, should be phrased naturally and
sympathetically. Maintaining the pitch of the final (or key note) through a
protracted sequence of melodies is as essential as it is difficult, requiring
pin-point accuracy of tuning on the first, last and penultimate notes which
need to be sharp enough, especially the whole tone which often forms the
last interval to be sung. This interval is often too wide, a problem which
is made worse if the penultimate note is also sung too flat. Apart from
this the top notes of the phrases will need to be high enough, along the
usual lines of minor thirds and sixths ringing just sharper than usual. The
words should never be pronounced in an explosive way: the vowel-sounds
unified and made clean, the consonants smoothed in a way to help the legato
flow, except to say that the natural accents of the words should be made
clear, since these will give the melodies extra shape. A good chant singer
will instinctively move towards the next natural accent, which often is on a
higher note than the rest of the phrase.
From these disciplines proceeds good polyphonic singing. The con
nection between chant and polyphony is not merely that one follows the
other chronologically, but that one fed into the other directly: there have
been few such smooth progressions from one stylistic period into another in

the musical times Winter 2014 11

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
12 Singing polyphony

the history of music, this one taking unusually long since it was necessary
to establish the basics of how to notate the new music. In fact this transfer

presents us with the most extreme example of a revolution in musical

composition - the possibility of writing part-music - while carrying on the
old way, which was to rely on chant as the basis for everything, including
the new polyphonic style.
The singing techniques of the monks and nuns can and should be
transferred to the kind of polyphony in which the lines are allowed to flow
in an uninhibited way, essentially until that point in history when barlines
became a necessity. This lack of inhibition characterises all medieval and
most renaissance polyphony until the High Renaissance, though one needs
to go to Monteverdi's Missa In illo tempore (1610) really to find out how
barline mentality could be applied to the prima prattica, as renaissance
composition was called by then. But for the most part the requirements are
similar because the make-up of the lines is so similar, proceeding largely
by step with the occasional carefully prepared leap, all within the same
restricted vocal compass within each part. This make-up suggests singing
through a phrase with a sense of direction towards the next cadence, which
may be quite long delayed; in order to maintain the necessary legato the
singers on that line will need to agree a system of breathing whereby there
is no audible break, by snatching a breath at different times; the lines should
ebb and flow up to and away from the highest notes, all the parts now
contributing to this effect.
It is worth adding that, except in specially designated passages, the
monks or nuns of the past would not have sung chant solo. To obtain the
desired legato they would have had several, perhaps many people contri
buting, which meant that staggering the breathing was no problem. In the
professional chamber choirs of today, where every singer costs, there is
more of a problem. The breathing of long legato lines with only two singers
to a part is difficult to manage, requiring careful agreement and rehearsal
as to where each will briefly fall silent. To have three or more to a part
makes this procedure simpler though in my experience it makes pin-point
tuning less likely. Two-to-a-part polyphonic singing has become established
both in terms of cost-effectiveness and as the optimum number of singers
able both to sustain a good legato and deliver good tuning. One-to-a-part
performances will work best in buildings which are reverberant enough
to carry the sound of the ensemble through the inevitable gaps caused by
breathing.
The rules of tuning are also closely linked to the realities of singing
chant, especially so with the first and last notes which will need to be sharp
enough - for example, in those codas where one part (often the soprano) is
asked to hang onto a note for several bars without rest. The impression that

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
 Ex.ia: Plainsong antiphon Ave regina coelorum, opening (beginning on G)

Ex. ib: Palestrina: Missa Ave regina caelorum, opening, showing the first phrase of the plainsong
(marked with *) presented in imitation

Cantus
ft

Altus
Altus

Ky

* *

Tenor
**
J- &
eJ
Ky le - i - son,_

* *

Bassus ■•*»■rr|T^
ir^
Ky

3^^ r ir r r
('
Ky

le -
le - ii -
- son,_
son,.

Ky - ri - e e Ky - rie_

Ky

the musical times Winter 2014 13

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
14 Singing polyphony

Ex.2a: Plainsong antiphon Ave regina coelorum, opening (beginning on F)

Ex.ïb: Palestrina: motet Ave regina caelorum, opening, showing a different working (marked
with *) of the same chant; the imitation here supported by extra harmony notes.
* * * * *

r » r r ir -^r
Cantus

* * * *

Altus

Tenor

Quintus
m

Bassus

JU3JJ U ■CT ZT

* * * * *

li J3J
* * * * *

ve

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
a piece has been sung flat throughout may come merely from the tuning of
the last chord, or the last few bars, which suddenly sag, as can happen so
easily in chant. Or it may, more seriously, have come from the tuning of the
first note or chord, from which it never recovers. Very often what happens
in performance between the first and last notes is perfectly reliable. Apart
from this the top notes in all the parts need to be high enough, especially, as
noted above, ringing bright. Mean-tone tuning should also be judiciously
applied in block chords and at cadences, with the fifths as pure as possible,
the major thirds lower, and the minor thirds higher than as represented on
the modern piano.
Similarly, the words will almost never be pronounced in a marcato
way, the only possible exception being in homophonic passages which are
deliberately underlining an important word. More generally, the vowel
sounds need to be unified, made clean and bright, the consonants mid
phrase clearly audible, so that a word ending with an's' is not automatically
elided with the beginning of the next word when it too has an 's'; and the
final consonants of a phrase sung lightly and carefully together. The natural
accents of the words should be made clear, since these will give the melodies
extra shape. A good polyphonic singer will instinctively move towards the
next natural accent in the text he or she is singing which, in the hands of
a master, is often on a higher note than the rest of the phrase. The words
are the essence of the matter, but they are not usually paraded as such. A
polyphonic web is a complicated and sophisticated construct which, as the
Protestant reformers noticed, is not ideally suited to projecting texts. The
words have been subsumed into the web, just as they have in chant: with
clear and unanimous pronunciation they can help both the impact of that
web as music, while being heard better by the public. But they will never be
as clear as they can be in monophonie solo singing, or when read.

of voice, born as it was in a completely different environment with

Much ofdifferent
thisimperatives.
style Modern
will singers
be unfamiliar
will struggle to undotothecontemporary students
mannerisms which have formed in the centuries since the 16th, and in fact it
is not desirable that they should undo all of them. As I said earlier, however
much we may want to respect the notes of the polyphony in front of us,
and hone a style to express them, in the end we are doing it for ourselves,
now, usually in circumstances a religious community would have thought
totally inappropriate. We do have to adopt modern practices to appeal to
our audiences. However, there are, or should be, limits.
Baroque vocal lines, whether sung solo or in a contrapuntal web, are not
constructed like renaissance lines. The 'revolution' in compositional practice
which took place around the year 1600 was rooted in a more sophisticated

the musical times Winter 2014 15

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
16 Singing polyphony

use of harmony than what had gone before. With harmony came barlines;
and with barlines came the possibility of block-based phrasing, whether in
two-, four- or eight-bar periods. These blocks are controlled by repeating
harmonic patterns, which often end with a perfect cadence. In conjunction
with this new practice came a more extrovert approach to setting words.
If these were sacred words - which was no longer the given it had been,
since the new interest in opera and its secular love language quickly took
the limelight - the feelings behind them might be played up more obviously
than had been customary in renaissance polyphony, through the use of
expressive devices such as chromatic harmony, disjointed rhythms and
dramatic turns of phrase. In general, there was an ever-clearer marriage of
individual syllables to individual notes, which might well come down to the
potential impact of every single note.
The techniques needed to sing this kind of writing as a soloist with basso
continuo is essentially the same for both the secular and sacred repertoires
and, because written for soloists, does not have very much application to good
part-singing. It is worth remembering that this kind of solo performance
did not exist before the invention of recitative around 1600. It remains true,
however, that if a voice sticks out above the others in any kind of equal
voiced counterpoint - in a choral fugue by Bach as much as in a renaissance
motet — the balance of the writing will be unsatisfactorily distorted. Of
course there are times when one voice (or part) needs to be heard above
the others, but when this should be remains a spur-of-the-moment decision
of real delicacy. Nothing is worse than a soloist thinking as a soloist for the
entire length of such a piece.
Many of the choral disciplines identified above for singing chant and
renaissance polyphony well are the same as for singing baroque counterpoint
well. But time did not stand still. Elsewhere in baroque writing wholly new
methods of performance were being tried out: ornaments which were not
meant to be performed rigorously in time; the use of rubato in phrases
which specifically involved unequal notes with the intention of making
melodies more affective; the regular use of cadences, to indicate where the
singer (and the player) can breathe. These are some of the expectations
which modern singers instinctively bring to the performance of renaissance
polyphony, and they all reduce the expressive potential of this music.
The most difficult one to undo is the notes inégales. The instinct to pull
around and embellish in the hope of finding greater expression is deeply
rooted in all of us. Even in Bach's vocal counterpoint it is not out of place,
since his phrases, although longer than most other composers' of the period,
still have a more packaged, harmonically-based scaffolding than anything
written by Palestrina or Lassus. But there are other differences.
The use of instruments to accompany voices in counterpoint changes

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
Ex.3: Caccini's Euridice, extract: an example of early baroque vocal style

che
che gl'al
È Jif r iprJ
gl'al-triaff-fre
-triaff-fre- naj
- naj
p ^r~^i?Jl
regreg
—p £
ge. del
- ge- ma del
mamio
tu_ mio
tu_do
do--lolo - rere
Sein-
Scin-til_
til_ la di pie

TJ

J r
Jpr^ ^pP i i P i
ta
tà non
nonsen
sen- rial'
- rial'
co co - re Ahi las so e non ram-men
non ram-men - ti

r r
r 'r r
r r

r
f ^J
j Jr rrJ||J JrrJ
f J'iJ r r J r 1 VJ
J r
Co-me tra-sigg'A-mor co-me tor-men - ti E pur fu'l mor - te dell'e-ter-noar-do - re

XT
TT

how we hear the music. In one way having instrumental acco

sort out any underlying tuning problems with a choir, since
organs tend not to lose pitch; but at the same time expressive
blend and tuning are made more difficult to achieve. If an id
a choir to sing polyphony is to have two people singing next
on each voice-part, working on making their vowel-sounds a
tuning systems as similar as possible, then having them
player who has no throat and who may be nowhere near the
is going to be crude to a degree. The problem for young
to polyphony today with this as their normal experience
performance practice is that they will have to develop an ear
of tuning they did not know existed, not to mention levels
with a fellow singer. The fact is that when an instrument is
performance of polyphony much of the responsibility for th
in it will be taken out of the hands of the singers. By this I
16th-century performances never used instruments, but I do
were avoided if possible and that a really good choir sing
today should avoid them too.

the musical times Winter 2014 17

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms
18 Singing polyphony

The controlling of the ego-driven solo mentality in good polyphonic

singing is the single aspect in all of the above which we will find most difficult
to realise. We are told from birth to be ourselves, to go for the thing we want
to do and be, and if possible make money out of it. No doubt the religious
men and women of the past were not entirely devoid of self-interest, but
their training was towards being modest. This was reflected in every aspect
of their daily lives, but none more so than in their singing where - to judge
from the music they sang - their job was to blend with others in an almost
constant hymn of praise to their maker. In this there was no place for the
ego; and if this is true for chant it is true for polyphony too. Nothing is more
destructive to the polyphonic web than a singer who wants to express him
or herself by singing out too far or, worse, by pulling the rhythm around.
Polyphony is at its most effective when it is sung perfectly in time, all the
notes exactly in their place like a clock ticking, which is the result of all
the parts of its complicated mechanism being in perfect synchrony. Any
individual singer who believes that rubato from him or her alone is going to
have a desirable interpretative result is thinking for themselves, and not for
the group as a whole.
But thinking like that is how the best singers of today have grown up.
And being proud of what they do, and confident of their abilities, is surely
part of what makes them good performers in the modern context, since
singers who want to hide behind others, as the monks and nuns of the past
could do, are not going to make the grade - one or two voices to a part - in
those symphony halls without amplification. There lies the challenge for
professional choirs singing polyphony around the world today; and there
the difficulty.

This content downloaded from

91.120.108.233 on Tue, 25 Oct 2022 21:51:26 UTC
All use subject to https://about.jstor.org/terms