home ' about ' albums ' the harp ' ap Huw pages ' contact
the manuscript ''' interpretation ''' tuning ''' historical context
interpretation/ rhythm
advanced search


So far, we have discovered how the movement symbols tell us the order in which notes are played, and the detail of finger technique, but we still have no clear idea of how the music is to be barred and allocated specific rhythms. The key to this aspect of the music lies in the system of "mesurau" which underpins practically every aspect of the construction of this music, from its rhythmic detail to its overall form.

On p. 107 of the manuscript we can find a list of the twenty-four "approved" measures: their names and their specific patterns in a binary notation. They can be best understood as chains made up of two distinct harmonic entities, which can be visualised in the form of chords. They can be most easily seen in the Clymau Cydgerdd section of the manuscript, from p. 23 to p. 34. The first two lines of text on p. 23 follow the measure Mak y mwn hir; its digital notation at the top of the page is given as 1111 0000 1010 1111 0000 1011. This pattern can be clearly seen in the lower part, where the symbol 1 is represented by the chord G-C-E, and the symbol 0 by the chord B-D-F. The above explanation of the upper part movements should also make it clear that the undamped notes in the upper part also belong to these chords.

Analysis of the music text shows that all the music apart from the pieces from p. 56 to 65 is based on one or more of these measures. Unfortunately not all the pieces in the manuscript have their particular measure patterns specified in the available documents, but enough can be learned from analysis of those pieces whose measure is specified to allocate measures or measure-types to the remainder.

The measures are particularly valuable to us because they have not only harmonic but also metrical, and hence rhythmical significance. Peter Greenhill makes the fundamental point that for the system to make any sense, each element or digit (1 or 0) of the measure-pattern as used in a particular piece should be allocated an equal time-value. He further notes that the measure-patterns used in the music notated in the manuscript tend to fall into groups of four elements, or digits. This can clearly be seen in the Mak y mwn hir pattern above. This tendency to group in fours is found to permeate the metrical and rhythmical structures of the music.

The first stage in applying this to the music itself is to go through the entire text, identifying the cyweirdant (1) and tyniad (0) elements of the measures. This is not necessarily a straightforward procedure, since a single element or digit can be spread over several columns of text. Thus in line 4, p. 39, there is an example of a single cyweirdant (1) element which spans eleven columns of text. Once this process is complete, however, we can move on to barring the music and allocating rhythmic values to individual notes.

For a simple example of this procedure we should refer to the first piece in the manuscript, Gosteg Dafydd Athro (p. 15). The measure involved is a commonly used one: Corffiniwr, which has the digital notation 1100 1011 1100 1011. Again the grouping in fours can be seen. For the purposes of a preliminary transcription into staff notation it is proposed that each digit of the measure should correspond to a bar of music. The next stage is to find the section of the piece which is most saturated with notes in the upper part, to see how long the bars need to be.

The section of greatest note saturation proves to be the final one, section 10, which is exactly the same as the first, but with every movement (plethiad y pedwarbys) doubled, so that it contains twice the number of notes in the upper part as the first section. The section is not notated in the manuscript, but replaced by a verbal instruction. In the original notation the start of it would look like this

A reasonable starting point would be to place the chords and single notes of the lower part on the main beats of a two-beat bar:

Given that the eight notes in the upper part of the first digit are to be played consecutively, the next question is how to place them in relation to the main beats. Peter Greenhill finds that generally the first note of each plethiad movement should be played just before the beat. If we follow the principles for the movements outlined above, we find that plethiad y pedwarbys consists of four notes (in this case F-G-F-G). The fingerings are: 3 (damped by 4) — 2 — 4 (damped by 3) — 1. Treating each note of the movements as of equal value for the moment (this will not be the eventual outcome), we find that they will fit as follows:

It follows that the remainder of the piece, which has a lower density of notes, can be comfortably accommodated within this bar-length. It has long been accepted that the damped notes in each movement should be short, since the undamped notes have the function of melody notes which tend not to change from section to section within a piece. Thus the equivalent part of the first section of Gosteg Dafydd Athro might look like this:

In order to establish a consistent method of fitting all the notes of all the movements into this format, Peter Greenhill further divides each of the beats of the bar into four (semiquavers in 2/4). The placing of the movements in relation to the main beats which is thus developed can then be applied throughout the manuscript.

This two-beat bar format works equally well for Caniad y Gwyn Bibydd, but for most of the remainder of the music in the manuscript it is found that the density of the notes in the more saturated sections demands a four-beat format. Exactly the same principles apply for the placing of the various movements in relation to the main beats, however, as the following example from section 9 of Caniad Marwnad Ifan ap y Gof demonstrates:

This is an unusual passage in that the normal position of the hands is reversed. It will be noticed that some chords in the lower stave which are missing or misplaced in the tablature have been restored. The digits of the measure (0011 0011) are here marked in green; the words "bis dechre" at the end of the first line mean that the first half of the line (up to the spiral mark) has to be repeated.

The movements which are numbered in red in the facsimile above are worthy of special mention:

1 Plethiad y wahynen — a single backstroke with the middle finger.

2 This figure is a scribal error: it should look like a letter z, and is a single backstroke with the ring finger.

3 An extended version of crafiad sengl, which normally consists of only three notes. The two highest notes are struck together.

4 Tafliad y bys ("throwing of the finger") - a composite movement which takes its name from the fact that the second finger has to play three notes in all, requiring it to leap from one string to another.

5 Ysgwyd y bys — a single backstroke with the index finger.

In order for the reader to follow the allocation of note lengths in the above transcription, two principles should be borne in mind:

1. The first note of each plethiad movement comes immediately before the beat.

2. The single-note (backstrike) crychiad movements: plethiad y wahynen, the ‘z’ movement, and ysgwyd y bys come immediately before the plethiad movements they anticipate.

The various movements in the upper part fit together like pieces of a jigsaw puzzle; Peter Greenhill has noted that certain combinations of movements occur very frequently, and it is difficult to imagine any other way in which they could be pieced together so comfortably. In the transcription above, to illustrate the fingering, where two numbers for fingering are placed one above the other, the upper number is for the finger which strikes the string, and the lower one for the finger which damps it, as the next string is struck. Notes which are struck with the back of the nail are marked in the transcription with a bar (|) above the note. The fingering technique works in such a way that the hand rarely has to shift position, and then usually between phrases.

While the transcription given above shows the order in which the notes are played, and their positions relative to the main beats, the note values should not be taken literally. Peter Greenhill’s study of the rhythms of Welsh poetry, and those of some British traditional music, such as Lancashire hornpipes, have led him to the conclusion that the shorter notes should not be played evenly. Pairs of semiquavers in the transcription should be played long-short, as quaver-semiquaver triplets. This gives the music a spring, or bounce; if played as written, these passages would sound rather dead. This is one element which gives Paul Dooley’s performances their lightness and vivacity.


back to top

  ©pauldooley.com 2004
last updated 17/04/06