How choristers and organists improvised on chant during the Tudor period in England
Author: Jane Flynn
I'm a music historian and performer (organ, harpsichord, piano) whose research interests focus on the Tudor period: 1485-1603. I'm especially interested in the evidence for the the training of choristers and organists during that period that allowed them to produce quite complex multi-voice (polyphonic) music using only chant books or other sources containing only a single notated line of music.
This is the fourth of a multi-part post. (Please see previous post below for earlier instalments.)
The main way that Tudor organists allowed more dissonances in their descants than were available in ‘text-book’ sixteenth-century counterpoint was through maintaining or repeating a short point throughout a descant. As Morley points out, the old descanters used this repetition to “excuse” certain dissonances of short rhythmic duration (including “two discords together”) that might occur against the melodic intervals in the plainsong when they “wrested” in the point (Introduction, pp. 76, 118). Example 5a is a version of an organ verset by Burton on Tibi omnes, in which the plainsong is presented in equal semibreves (except in bar 3), rather than in the plainsong figuration in his verset (Ex. 5b). The fact that the descant works contrapuntally against a simple plainsong, as well as with the notated plainsong figuration, suggests that a high degree of flexibility is inherent in it. This aspect of points will be discussed further below.
The descant of Burton’s Tibi omnes comprises the repetition of a point that lasts two semibreves; over the first note of the plainsong it consists of a crotchet rest and three consonant crotchets that include a leap (of a 3rd or 4th, whichever is consonant at the time); and over the second note of the plainsong, two descending stepwise crotchets and either a repeated note if this is consonant, or if it is dissonant, another descending step (to a consonance). Because this point includes a rest that occurs at the same time as a note of the plainsong, the number of progressions is reduced, making voice-leading easier to manage for beginners. Note, however, that there is a close connection between this descant and 4:1, since it comprises mainly crotchets. The minims at the end of each point could easily be replaced by two crotchets consisting of a lower neighbour (the notes g and a in the first point).
Example 5a A version of Burton’s Tibi omnes on a simple plainsong:
In Ex. 5b, Burton’s Tibi omnes, the plainsong begins as simple, undecorated semibreves, but in bar 3 the rhythm is changed from two semibreves into a dotted semibreve—minim. This is to avoid the voice-leading error of a ‘hidden 5th’ (created by similar motion to a perfect consonance) that would occur with two semibreves. This solution is incorporated into Ex. 5a, but otherwise the point works well contrapuntally against a simple plainsong, if rather dull! Burton’s other alterations to the plainsong add rhythmic and melodic variety: for example, in bar 4 the second semibreve (B-flat) is broken into two minims (B-flat—A); in bar 8 there is a popular rhythmic device in Tudor organ versets, in which two semibreves are replaced by the syncopated rhythm, minim—semibreve—minim; in the last bar the plainsong is broken like the descant, having two occurrences of the point itself. Of particular interest is bar 10, in which the simple plainsong note (B-flat) is broken into two minims, A—B-flat, creating a dissonant 7th in the descant that is neither prepared nor resolved ‘correctly’, but is excused because the point is being maintained. Likewise, the last bar includes three discords together: an accented passing-tone, an escape tone (es, a leap from a dissonance), followed by another accented passing-tone. Maintaining points and making plainsong figuration are ways in which descanters could create good voice-leading and manipulate dissonance, either by adding dissonance or reducing it.
Example 5b Burton, Tibi omnes (Te Deum, 29996, fols 22v-23r):
Because points tend to be very flexible, they can often be applied to any given plainsong: as mentioned above, the interval of any leap in a point may be altered; they can be applied at any pitch (not only at 4th, 5th, or 8ve); although they tend to be short, they often occur in shortened form or are broken up or “dissolved” (Morley, Introduction, p. 95), especially towards the end of a verset. In Ex. 5c I apply Burton’s point to the Miserere and play repeated notes in the plainsong (except in bar 1) in the rhythm minim—semibreve—minim (as in bar 8 of Ex. 5b).
Example 5c Jane Flynn, Miserere 3: One way of using Burton’s point:
This is the third of a multi-part post. (Please see previous post below for earlier instalments.)
Example 4a is a Sanctus by Avery Burton (c.1470-c.1543, of the Chapel Royal). It has eight notes to one of the plainsong, and, like the 12:1 Miserere (Ex. 3f), includes many groups of four stepwise quavers. But though the dissonances in the Miserere are the same as those already discussed: passing-tones and neighbors, which move stepwise, Burton’s verset includes an example of what Morley describes as “two discords together” (in this case parallel 11ths or 4ths at plainsong notes 4-5). Morley condemns this practice, although he acknowledges that “harsh allowances” of dissonance “was the fault of the [earlier] time” (Introduction, pp. 118-19). Ex. 4a also includes appoggiaturas, which are forbidden in ‘textbook’ 16th-century counterpoint because they involve a leap to a dissonance before resolving down by step. In Ex. 4a, the first appoggiatura consists of a leap to an E-flat, which is a 7th above F in the plainsong; it resolves down to D, which is a consonant 6th.
Example 4a Avery Burton, Sanctus (Te Deum, 29996, fol. 23), 8:1, pts, ln, un, and appoggiaturas:
Burton’s verset demonstrates that a descant can consist of a repetition of a group of four (or more) stepwise descending notes each of which ends with a consonance against a note of the plainsong. Furthermore, any group of descending notes can begin on a dissonance, because it will resolve down by step.
The beginning of Ex. 4b, Christe rememptor omnium by Blitheman, can be used as a model of how to practice this technique, because all of the descending scales lead to a consonance (many of them a 10th). If the first note of a scale happens to be a dissonance, this is not a problem, because it resolves stepwise downwards as an appoggiatura (as we saw in Ex. 4a).
Example 4b Blitheman, Christe rememptor omnium (30513, fols 103v-104v):
In Ex. 4c I demonstrate a rudimentary way of practicing this technique by playing a descant that comprises groups of three stepwise descending notes. Ex. 4c has a contrapunctus primarily of parallel 10ths against the plainsong, and therefore by playing a group of three notes starting a 12th above the next note of plainsong I play a 10th at the same time as the note of plainsong. It does not matter whether the first note of the group is consonant or dissonant, because any that are dissonant will either be upper neighbors or appoggiaturas. In Ex. 4c, for variety when there are repeated notes in the plainsong the consonance sounding at the same time as the note is a 12th rather than a 10th, and this can be sighted as a 13th.
Example 4c Parallel 10ths above Miserere, each preceded by two stepwise descending crotchets (12ths for repeated notes in the plainsong):
In order to play groups of four notes, you can practice sighting a 6th (or 13th) above each upcoming note of the plainsong, and therefore play 10ths above them. This is a little more difficult than playing groups of three notes only because you have less time to look ahead in the plainsong as you are playing. For variety you can introduce other consonances and begin at the appropriate interval above the upcoming note of plainsong: descanters tried to develop what Morley described as a “quick sight” of descant (Introduction, p. 120). In Ex. 4d I demonstrate this technique using 6ths and 8ves as consonances, and in addition, begin after a rest, as in Burton’s Sanctus (Ex. 4a). This creates a syncopated effect that Morley describes as “a driving waie … odded by a rest” (Introduction, p. 89), and it also means that the consonances do not sound at the same time as the next note of the plainsong, but a note afterwards. Of course, this is not a problem because the notes are still descending stepwise to a consonance. There are four places in which there are parallel 4ths, i.e. “two discords together” (see Ex. 4a).
Example 4d Jane Flynn, Miserere 2: Sighting consonances of 6th and 8ve above Miserere, and playing 4:1 in descending crotchets, odded by a minim rest:
This is the second of a multi-part post. (Please see previous posts below for earlier instalments.)
Example 3a, a Miserere by John Redford (d. 1547, master of the choristers at St Paul’s Cathedral, London), is an example of Sextupla or 6 descant notes to each one of the plainsong. As in Exx. 2a-b, all the leaps in the descant are consonant, and the rest of it moves stepwise; but in addition to passing-tones and accented passing-tones Ex. 3a includes lower and upper neighbors (or auxiliaries). These are indicated below the staves; for example, at the beginning of bar 2, the second note in the bass, E, is a dissonant 11th (or 4th) below A in the plainsong; it is a lower neighbour because it is preceded and followed by the note F which is consonant against the A.
Example 3a (recorded up 4th) John Redford, Miserere (30513, fols 10v-11r), 6:1, passing-tones (pt), accented passing-tones (apt), lower neighbours (ln) and upper neighbours (un)
This Miserere by Redford can be used as a further demonstration of how beginners could practice playing ‘species-counterpoint’-style descants against a simple plainsong. Ex. 3b is a note-against-note contrapunctus of Redford’s Miserere, with just two alterations to the consonances: these are notated as black semibreves on the note G, to avoid the melodic tritone between B and F; and some notes are put up or down an 8ve. (The ‘hidden 8ve’, created by similar motion to a perfect consonance at plainsong notes 21-22, is being allowed to stand as it is not ‘forbidden’ in Tudor style.)
Example 3b Contrapunctus (1:1) of Redford’s Miserere:
Examples 3c-e are Dupla (2:1), Tripla (3:1), and Quadrupla (4:1) descants based on the contrapunctus in Ex. 3b of Redford’s Miserere. The dissonances are limited to passing-tones and neighbours.
Example 3c A 2:1 descant based on the 1:1 contrapunctus of Ex. 3b:
Example 3d A 3:1 descant based on the 1:1 contrapunctus of Ex. 3b:
Example 3e A 4:1 descant based on the 1:1 contrapunctus of Ex. 3b:
Example 3f is an anonymous Miserere with 12 notes to 1 that Thomas Mulliner copied into his book right after he had copied Redford’s Miserere. If you compare the first six consonances between the notes of the descant that sound at the same time as notes of the plainsong in Ex. 3f, you will see that they are the same as the consonances in the contrapunctus of Redford’s Miserere in Ex. 3b (and of Ex. 3c-e). Perhaps Thomas Mulliner wrote this verset himself, using Redford’s Miserere as a model. The only notated dissonances are passing-tones and neighbors, as in Ex. 3b, but in bar 9 there are double-stroke ornament signs drawn through the stems of six crotchets. As such signs are used only in this bar in the verset, Desmond Hunter’s comment (regarding Virginalist music) is apropos: “embellishment implied by the signs should feature as an integral part of the music rather than simply adding surface decoration”. I suggest therefore that the crotchets with the double-strokes may be realised as two quavers (the note above the pitch of the crotchet, and the pitch of the crotchet itself), so that the long continuous run of quavers is maintained. In this case, the use of the signs could be an acknowledgment that four of the six crotchets create additional dissonances (two appoggiaturas and an unprepared or ‘fake suspension’, which will be discussed in more detail below, Ex. 4a and Ex. 7a, respectively).
English descant (counterpoint) treatises from the mid fifteenth to the late sixteenth centuries demonstrate various ways of learning and of teaching students how to perform counterpoint against a cantus firmus vocally and/or on the keyboard. These include, for example, Leonel Power’s (1375/80-1445) treatise, which has the heading ‘This tretis is contrivid upon the Gamme for hem that wil be syngers or makers or techers’ (GB-Lbl Lans. MS 763, fols 105v-113); a manuscript copy of William Bathe’s (1564-1614) A Briefe Introductione to the True Art of Musicke (London, 1584), and Thomas Morley’s (1557/8-1602) A Plaine and Easie Introduction to Practicall Musicke (London, 1597), dedicated to his master, William Byrd (c.1540-1623). They begin by teaching ‘simple’ or note-against-note (1:1) counterpoint—how to make a series of consonant intervals between a ‘plainsong (the term they used for a cantus firmus, even if it was unrelated to liturgical chant) and a new melody modelled stylistically on liturgical chant. Simple counterpoint provides the contrapunctus or framework on which various kinds of contrapuntal “ways” could be made: descants that introduce more dissonances than are typical of most other Renaissance musical styles and descants that comprise the repetition of short ‘points’ (motifs), among other techniques. An experienced organist might also choose to elaborate the plainsong itself, at the same time as playing a descant against it. This is one of the four techniques Morley described as ‘plainsong figuration’ (learned by choristers at Cathedral and College choir schools), known as ‘breaking’ the plainsong (Introduction, p. 90); in other words, a (solo) singer applied the same procedure to the ‘simple’ notated chant as to the simple sighted (imagined) descant sung by another (solo) chorister. This could result in performances in which the singers effectively contested against each other over ‘who should bring in the point soonest, and make hardest proportions’ (Morley, Introduction, p. 120).
This blog gives a brief introduction to note-against-note counterpoint and then examines the descant techniques used in seven rudimentary two-part Tudor organ versets (Exx. 1-7). These versets are notated using note-values that resemble the originals, but with ligatures indicated by square brackets, and on five-line staves with modern clefs. The dissonances and voice-leading are indicated below the staves as they are indicative of Tudor musical style. The descant techniques in each verset are then further demonstrated by being applied to the Miserere, which is an antiphon for the prayer office of Compline that was often chosen as a plainsong in pedagogical sources. The blog concludes by referring to eight additional organ versets (Exx. 8-15) that either include two or more of the techniques used in the rudimentary versets, or demonstrate easy methods of adding a third part, in order to demonstrate some of the ways that the more experienced Tudor organists could have improvised.
The Miserere (see Fig. 1) has a small range (only a 5th), it moves mostly stepwise, with some repeated notes, and includes the leap up and down of a 3rd (a-c and c-a). During Compline, the choir would sing the Miserere antiphon first, followed by the psalm (sung antiphonally), but the repetition of the antiphon could be played by an organist (replacing the choir): he or she would read the simple (undecorated) plainsong melody from a book of chant, and play one or more descants above or below it. When the Miserere is sung monophonically the rhythm will usually follow the stresses of the words (perhaps with a short break in the middle), but when it is used as a plainsong by a Tudor organist (or by a choir for a solo singer to descant on) each note is given the same length of time, except perhaps at the end, when the penultimate and antipenultimate notes might be doubled, and the last note held.
Figure 1 Plainsong Miserere mihi (Have mercy upon me O Lord and hear my prayer):
Leonel Power’s descant treatise demonstrates a method used by beginners, especially in England, in order to become proficient at improvising descant. This method is to create an imaginary note-against-note contrapunctus against the plainsong (which they are reading) (a process known as ‘sighting’), using only the consonances (shown in Fig. 2). The consonances are subdivided into perfect (unison, 5th, and their octave equivalents) and imperfect (3rd, 6th, and their octave equivalents) because they are governed by different voice-leading rules, the main rule being to avoid producing consecutive parallel perfect intervals (5ths or octaves).
Figure 2 The consonances above and below the note G: unison (= 8, 15), third (10, 17), fifth (12, 19), sixth (13, 20), octave (15, 22):
Once the performer is comfortable performing simple counterpoint, he or she can choose to perform a ‘broken’, ‘diminished’, or elaborated version of a ‘sighted’ contrapunctus.
The first organ verset to be examined here is Dignare from a Te Deum by John Blitheman (c.1525-1591, organist at Christ Church, Oxford, and the English Chapel Royal), in which the organ replaces the choir for the odd-numbered verses. This Dignare (Ex. 1) has three notes for every one of the plainsong (3:1).
 The ‘plainsong’ in the organ versets of Blitheman’s Te Deum consists of the ‘faburden’; see Part 7 of the blog for a discussion of ‘on the faburden’.
As Morley points out, the old descanters used the term Tripla ‘when for one note of the plainsong, they make three blacke minimes’ (Introduction, p. 90). In Ex. 1 the plainsong is in the top part notated in semibreves, and there are three black minims to each one in the bass descant (separated in the manuscript by dots to help the eye). The figures below the staves indicate the consonances. An unusual (but not unique) feature of this verset is that all of the notes of the descant are consonant (also see Exx. 12, 13, 15). Below the figures are the letters c, o, s, and p, which indicate the kind of voice-leading that exists between the two parts when, as discussed by Bathe (Briefe Introduction, ed. Karns, pp. 124-26), they both approach the next note of plainsong (unless both parts repeat their notes, indicated as R). For example, the first note of the plainsong, F, moves up to the note A, while the third descant note, D, moves down to the note C that sounds at the same time as the A in the plainsong: the parts are progressing in contrary motion, and therefore marked ‘c’. Likewise, ‘o’ indicates oblique motion, in which one of the parts repeats the same pitch; ‘s’ indicates similar motion, in which both parts move in the same direction, up or down (to an imperfect consonance, a 3rd or 6th), but with different melodic intervals; ‘p’ indicates parallel motion, which is used with imperfect consonances in Tudor style, but not with perfect consonances (5ths or octaves). Note that the voice-leading is mainly in contrary motion: c=10; o=3; s=3; p=3; repetitions of consonances (i.e. no motion)=5.
Example 1 John Blitheman, Dignare (Te Deum, 30513, fols 75v-76r), 3:1, all consonances:
The descant in Ex. 1 includes many leaps because it is entirely consonant. A more usual method of playing Tripla is given in Ex. 2a, another verset from Blitheman’s Te Deum. In Ex. 2a all the leaps are consonant but most of the descant moves stepwise, and the notes that are dissonant (2nd, 4th, or 7th) can all be described as passing-tones. In addition to using passing-tones in a metrically weak position, Tudor musicians allowed accented passing-tones (which are forbidden according to ‘text-book’ sixteenth-century counterpoint rules):
Example 2a Blitheman, Tu rex (30513, fol. 73v), 3:1, passing-tones (pt) and accented passing-tones (apt):
In Ex. 2b I demonstrate how to apply the technique of Tripla with passing-tones and accented passing-tones in the descant to the Miserere plainsong. Note that the voice-leading between plainsong notes 9-10 and 21-22 is contrary motion (6th to an 8ve) in both places, and what might to us today be considered consecutive parallel 8ves (G/g to F/f) are not a problem in Tudor style, since, as Morley explains (Introduction, p. 80, another consonance (the 6th) occurs between them (also see Ex. 10 in a future blog).
Example 2b Jane Flynn, Miserere 1: One way of playing a 3:1 mostly stepwise descant above Miserere using passing-tones (pt) and accented passing-tones (apt):
 For more detailed discussions on the descanting techniques described in this paper see Jane Flynn, ‘Tudor organ versets: echoes of an improvised tradition’, Journal of the Royal College of Organists n.s. 3 (2009): 5-23; ‘To play upon the Organs any man[ner] Play[n]song’, BIOS: Journal of the British Institute of Organ Studies 34 (2010): 5-23; ‘Thomas Mulliner: An Apprentice of John Heywood?’ chapter 10 in Young Choristers, 650-1700, ed. Susan Boynton and Eric Rice, Studies in Medieval and Renaissance Music 7 (Woodbridge, 2008), pp. 173-94; ‘The musical knowledge and practice of expert Tudor descanters’, in Late Medieval Liturgies Enacted:The Experience of Late Medieval Worship, ed. Sally Harper, P.S Barnwell, and Magnus Williamson (Ashgate, forthcoming 2016).
 The only surviving version of this treatise is a manuscript copy by Andrew Melville (1593-1640), Scotland, University of Aberdeen Library MS 28, edited and discussed in A Briefe Introduction to the Skill of Song by William Bathe, ed. Kevin C. Karnes, Music Theory in Practice, 1500-1700 (Aldershot: Ashgate, 2005).
 I am very grateful to William Flynn for his assistance in preparing the music examples in this blog.
 I recorded all the music examples on 7-8 September 2013 on the Wingfield Organ built by Martin Goetze and Dominic Gwynn (www.goetzegwynn.co.uk) in 2001, for the Early English Organ Project. I wish to thank Andrew McCrea, Project Administrator of the EEOP and Director of Academic Studies at the Royal College of Organists, for his assistance, and Eleanor Goodison and John McCracken, church wardens at St Andrew’s Church, Wingfield, Suffolk, where the organ is currently resident, for their kind and generous hospitality. The organ compass is F to a2, without g#2, 40 notes; versets that have a range below F are transposed up 4th if their upper range is below e1 (see Exx. 3a, 3f, 10); for Ex. 11, which exceeds this range, the left-hand part from plainsong notes 24 to 34 is played up an octave. The sounding pitch of the Wingfield Organ is about one and a half semitones above A440, with tuning based on the Erlangen comma, a form of Pythagorean tuning.
The purpose of this site is to disseminate over ten years of research that demonstrates many of the improvisation techniques of the Tudor period. These are not only enjoyable to experiment with, but they will also help you to develop a deeper appreciation of the vocal and instrumental music of the period.
The first few posts will contain the text and music examples based on a paper I gave a few years ago at Amsterdam’s extraordinary Orgelpark—a university museum devoted to the organ’s past and future. I recorded all these examples in St Andrew’s Church in Wingfield (Suffolk) where enough of the timber of the original organ was found to provide some of the best evidence for reconstructing the small Tudor organ that is my featured image. All the examples were played on this organ in Wingfield, probably always in the large collegiate church that most likely supported a small number of choristers and allowed them to develop the basic musical skills outlined in the posts to follow.
Future posts will contain guided exercises that (if practised diligently) should allow you to improvise on chant using techniques admired and cultivated by Tudor singers, organists and composers.