On hearing that the Folk Choir in which I sing, does folk carols and the the Castleton Carol wasn't on the rosta, I decided to have a go at an arrangement. The carol dates back to something like the 15 century, when it was first mentioned; however the tune was not written down until the early 20th century, where it appeared in The Journal of the Folk Song Society, vol.IV (14) 1910, p.63, as Over Yonder's a Park. This contained Ralph Vaughan Williams and Ivor Gatty transcribing of Mr. Hall, of Castleon Derbershire singing the song in 1908. A transcribed copy of this tune is:
Ralph Vaughan Willams commented "(He) Districtly heard 'purple and pall' when sung to him". Also "The tune above given is, I think, traditional, but has probably been slightly “emended” (see F♯ in bar 6), by generations of Castelton carol-singers (of whom Mr Hall was one), who were in the habit of adding “basses” to their carol-tunes. It must, however, be said that Mr. Hall gave me no bass for this one."
Ralph comment on "purple and pall", just about every version I've heard sings these words, and so thats what I adopt. On the chorus, and almost every uses "And I love" and not "But I love" so we stick with that.
A greater question is what is the carols about? What is clear is that it was based on an earlier carol, the Corpus Christi Carol. This carol is very rarly performed, I only know Jeff Buckleys version myself, which is in turn probably strongly based on the Benjamin Britten version. However if you look at the verses, they are almost identical to the Castleton Carol, whilst the chorus is totally different.
The Corpus Christi Carol is clearly about the Fisher King, a character from Arthurian legend, the Fisher King was a holder of the Holy Grail. The Fisher King, had a leg/groin wound that meant he was not mobile. The wound did not heal, but the King was sustained by being the holder of the Holy Grail.
With this background the words of the Castleton Carol make more sense, but there is no mention of the king in the lyrics - this makes some think the carol is about Jesus. Here though, given the basis on the Corpus Christi Carol - we borrow a verse. The verse in modern english "And in that bed there lies a knight, His wounds bleeding day and night;" - which we add as a whole new verse.
The other verse which differs, is the stone that stands by the bed side. The Corpus Christ Carol says in modern english "And by that bed’s side there stands a stone, 'Corpus Christi' written thereon." This makes it clearer that the stone is the Holy Grail, than the line that "the virgin mary knelt upon". Hence I replace with the line from the Corpus Christi Carol.
The other verse which needs explaination, is the thorn which blooms. In Arthurian legend, Arthur when injured in battle, recovered in Avalon. So it is possible that the Fisher King was also in Avalon. Now a possible location for Avalon is Galstonbury Tor. Galstonbury has a thorn, which somewhat unsually flowers twice a year, and can flower on Christmas. Traditionally a branch is sent to the queen, and this continues till today. So the blooming thorn in the Castleton Carol is almost certainly a reference to the Glastonbury Thorn.
Otherwise I take my lyrics almost straight from Mr Halls lyrics, a few minor rewordings is all.
So before starting on harmonisation, what tune is the taken as the basis on which everything is based. The simple answer would be to base it on Ralph Vaughan Williams transcribe of what Mr Hall sang; but since then many versions have been recorded, and the song, as all folk songs has evolved.
However the general tune hasn't changed, on just about every version has "purple and pall" with the "pall" flattened a semi tone from "and" to be out of key. Now the chorus, lines 2 and 4, the first bar and two notes is repeated, but this gives a question mark on the first note of the 2nd chorus line; a natural accidental of E. This seems odd, it subrtracts from end of the previous line, shortening its duration. So here line taken by many is used, "And I" is taken as two semi quavers on the D, this was suggested by my choir master - from what he sung as a child, and I agree with him totally . This makes the start of both lines of chorus identical and keeps last note of each line as a crotchet.
On timing, nearly always the last note of each line has a some what unspecified rit. In the last line of the chorus, in a few recordings there is a serious rit on the "bove" of above in the last line - this brings the verse to a false conclusion which is resolved through "anything" which takes the tune back to the tonic. This we make explicit by delaying "anything" to start on the second tripplet of the bar.
The last bar also minor variance on the notes, in particular should "a-ny-thing" be 2-3-1 or 4-3-1, 2-3-1 would keep it the same as previous lines. Both are seen, but I prefer going up to the fourth, to keep a descending run to the final tonic.
The key of the carol does change. G minor as shown by Mr Hall is common, but offen you see E minor and A minor. Here I chose E minor, as its the key that I'm most familiar the song being sung in.
This gives the Cantus Firmus on which the harmony is based as:
The song has the chorus intersprsed as every second line, with the verse in between. The verse is telling a story, and so ideally this should have clarity in presentation, the chorus though does not need the same clarity in presentation. This suggests having just a single part on the verse, with the chorus in harmony.
The first, second and fourth line, after the first quaver (on lines two and four) all have the same tune for a bar and a half, with only the end of line four changing.
Line three brings the interest, as its different, goes higher that the other lines, and in the last bar decends by thirds, rather than a note
That the lines are similar means harmony may well break this. Each of the main lines starts (after the quaver) starts on the tonic, and so this sets up how harmisation will work. The end of line four, the second chorus line, decends to the tonic, and so this will drive chords.
In doing an harmony, I have in back of my mind, the rules of counterpoint. These rules are great when setting out against an existing tune, the cantus firmus. The aim is to give each voice its own tune, that has its own interest. The rules say things like, no parallel octave or fifths, aim to have contary motion (so when one voice rises, another falls).
Counterpoint though has no knowledge of chords, and chords are a basis of modern music. The fifth in a chord is the third harmonic of the note on which a chord is based; the fifth harmonic is a note just below a major third. So this is the basis of triads, a note plus a third plus a fifth. The cantus firmus is in E minor, and so this suggests minor chords, these with the third a fair distance below the fifth harmonic, have a bit of an edge to their feeling (in just intonation the fifth harmonic of a minor third, is the sixth harmonic of the tonic). This is the nature of the minor key, and the chords emphasise this. Hence the question of which minor chords are avaiable in a minor key, of E:
Now many song will be based on just one or two of these chords, the tonic plus dominant is common. However to establish the key of E minor, the subdominant chord is also need, its how the note of C is accessed. This is actually a subtle point, the cantus firmus contains no C note, so how can it be said to be in E minor, when in B minor there would be no changes (B minor has a C♯ whereas E minor has a C). Here we take it as E minor, as the song starts in E, and finishes in E - but we can emphasise this by using the subdominant triad.
Chorded songs often follow chord progression, e.g. I-IV-V-I. Starting on the tonic triad, give a stong hint at key, and that the song ends on an E suggests finishing on the tonic triad. So we aim for a progression that starts and finishes on a tonic, but is broken in the middle. The first line of chorus ends on a F♯, which can only be accessed in the Dominant. So as a whole this suggest a general progression I-IV-V in the first line of chorus, which is then reversed in the second line of chorus, V-IV-I. The latter is a tad of a blues progression, but we use it anyway.
Now using chords everywhere, and the music rapidly becomes chord heavy, the intervals uncomfortable to sing. So instead aim to only use chords on the key notes, in line with the chord progression. So on the first line, use a tonic triad on two of the E options; a subdominant on "them" becoming a dominant on "ring" (which breaks the rule on contary motion, but only slightly).
Between the chords we use general rules of counter point, give as much motion as possible. As the folk choir is most comfortable in the central octave, also avoid too many excersions. To give enough room, this forces all parts to at some stage sing the same note, which in the first chorus line is done on "I".
Now the second chorus line, starting on a B, and the dominant triad can be accessed without inversion. The next bar becomes primarily the subdominant chord. The final bar is best keeping it all in the tonic, with an inversion on "bove" which is resolved on "thing".
The second chorus line quite easilly becomes chord heavy, so in order to break this, on "Je-" have everyone sing a D, on "-sus" adding in a fifth; to only introduce the third and the full chord on "a-", as lead in to the final bar.
So this rapidly gives the arangement:
This shares the verse out between the voices, with it slowly going round the whole choir, to end in unison on the last verse.
Its worth saying a few words on how to fit the words to the notes. The music is written in 6/8 which means two groups of three quavers per bar, and so is counted "1-2-3 1-2-3" with the emphasis on the first note of each triplet.
Now this is also used to chose where to split a word, to make a line fit the music. Almost always this should be split a word over the first and second quaver. So verse 3 is sung "At the bed s-ide there st - ands a stone.", whilst verse 4 is sung "And in that b-ed there li-es a knight.", so the bed differs in each line, due to different placement.
The tricky line is "With “Corpus Christi” written thereon." where both Corpus and Christi must each be spread across a triplet, and so "C-orp-us Chr-is-ti".
So it is worth looking at the arrangemnet, to see how well it has worked. The High and Low voices, have reasonable independence, and a not unreasonable tune. The Mid voice though is more constrained between the High and Low. Where though the mid voice has a horible time is going into the second line of chorus, getting the worst of the resolution of the D♯ "pall"; followed by 3 notes on A, a tricky note out of the run, then three notes on B. So what can be said - many of these notes are constrained by the structure contained, e.g. condsider the B note, this starts as the fifth of the chord, then the root, then back to the fifth. Hence the line is left as written for the time being, but will the knowledge that it may need rexamining when sung in choir, the mid line is not fun.
But how does it sound, well so far only synthesised piano has been used. This is played without any feeling, but as a slave to the music, it make no mistakes from what is written!
Discussing with the folk choir master, and the fourth line for the mid voice would probably be beyond the choir ability. In particular the drop of two semi tones over "and pall, And I" is always hard for a voice, then the jump from A to D would put the mid voice at the top of their range and so not be comfortable. As a singer with many years experience, I can understand these points, but with respect to the semi tones, would comment that what needs to be grasped is that "pall" is a note out of key, as such it is meant to have an edge to it; however as soon as the tune returns to key, the singer needs to think in key, and so "And I" is sung in key, a tad below "pall". Singing in key is something to feel in the bones, though years of tunes.
The choir master point though is taken, and so a second version has been created. The sharp note "pall" is an essential part of the tune, as Ralph Vaughan Williams says - it is there through the tradition of folk carol singers in Castelton. So this note is kept, the other option the D♮ "And I", then the B minor chord can not accessed at the start of the fourth line, a pity as this is the only place an uninverted dominant chord can appear; but what is must be, and so the chord is sacrified for the same start as line two, a fifth and an octave.
Now the fourth jump for the mids, came about as building the chord on "Jesus above", so this build needs to go. Still though it would be good to keep the chord progression V-IV in this bar. Another complication is that the F♯-C interval is a diminished fifth, a tritone. Tritone I prefer to avoid, its a nasty interval, so any F♯ can only have an inverted chord that does not have a fifth. This is easy on "my", by using an inverted dominant chord, which was wanted at the beginning of the line. The "sus" of Jesus, can be made subdominant chord which was wanted on the end of this bar. "Love" and "Lord" and fifths want to be avoided (otherwise we have parallel fifths); what gives a good pattern for each voice is doing parallel sixths, which is allowed in counterpoint for a short period. This we can use to have the mid and high in unison on "love", spliting to give the dominant chord, before joining again on "Lord" - this gives a good singing line. Now turning to "Je", to fit in with the notes around, and avoiding voices moving together, and a fifth plus an octave works well. This leaves only the hard F♯ on "a" of "above". Fifths can't be used, an inverted dominant chord wouldn't be appropriate at this point - as its used on "my"; so the options are closing down. What was chosen, was almost the only option, an inverted B7 chord, missing a third, this is the dominant chord again, but without the third doesn't seem so, and gives a good clash before the last bar harmony. This doesn't give an ideal lead into the final bar, but there seems little other option.
This gives the music:
And to as how singable it is, line four has the same start as line two, which simplifies the singing. The mid voice still goes up to a C, which is hard to avoid. The mid line, is certainly more singable, so hopefully more acceptable. The high voice, has an intesting tune, often moving in opposite to the low, but keeping chords as wanted.
And the tunes:
The arrangements done here are © David Summers 2019. However the music has been written to be sung. I can't say if the music is any good, but I'd be glad for people to sing it. So if you want to use the music for your choir, all I ask is you drop me an email at music (at) davidjohnsummers (dot) uk, and that you don't make any changes to the music without asking me.
You can download pdf of the music, by clicking on any of the images above.