John Dowland: Farewell (P4) (An in nomine)

What a challenge! Hard to understand, harder to arrange, even harder to play. 

Well, Autumn has arrived again here in Wales, so I'm back in uke transcription mode.

I made these arrangements from a version in: Poulton and Lam, 1981, ‘The collected lute music of John Dowland’, Faber & Faber. I would never have managed transcribing the MS off my own bat.

First, a description of the piece by the great Nigel North:

“Dowland's Farewell on "In Nomine" stands in a unique position within his lute music. By his curious choice of musical materials in this piece, Dowland is locating himself at the absolute center of English instrumental music. A century before, the Englishman John Taverner had composed a popular Mass on the plainsong Gloria tibi trinitas. The Benedictus of Taverner's Mass circulated seperately and quickly spawned a huge complex of instrumental pieces emulating it. These pieces became vehicles for composers to show off contrapuntal mastery and clever invention. Every important English composer in the sixteenth century seems to have written at least one piece of music for instrumental consort that uses a long-note cantus firmus taken from Taverner's "In nomine" section; later composers such as Christopher Tye and William Byrd left numerous "In nomines." This lute work is Dowland's only contribution to the genre, but is the only surviving "In nomine" for it. 

“Dowland [mostly] places Taverner's cantus firmus melody baldly in the highest voice, and weaves a fantastic net of counterpoint beneath it. Indeed, though he titles the piece "Farewell," a name he uses for only one other lute solo (a highly chromatic and difficult one), it properly belongs with his many Fantasias. The lower "voices" of the lute gradually shift through a number of different contrapuntal gambits and textures -- different imitative motives, syncopation and chromaticism, a dancelike change of meter, hocket, culminating in virtuosic passagework at the end. Dowland took a generation of music for five people and proved he was capable of playing it with five fingers.” [allmusic.com]

I have checked with Taverner’s "In Nomine" (available on IMSLP) and the theme (mostly in the upper voice) in Dowland’s piece agrees very closely with Taverner’s cantus firmus or superius, but the notes on the lute are struck twice in each bar. Both pieces set in Dm (which transposes to Em on the uke). The main deviations, are in bar 6 (Tav. has C, Dow. has A) and bar 55 (silent in Tav.).

An extract showing two particularly complex bars / measures. The cantus firmus is obvious in the notation as white notes. The second and third voices progress with their own rhythms, and in the 2nd bar the lower voice has a semiola rhythm. You can see how difficult it has been to fit all this onto a uke!

In transcription, some bars are much as in the lute original; these are: 1, 3, 4, 16 – 23, 27 – 29, 37, 38 and 54. In other bars much use is made of strings 5 – 7 on the lute; in some cases I have raised these notes by an octave if they contribute to the performance with little distortion of the music or of the fingers. 

The 1st voice or cantus firmus is generally evident by being a white note, whilst most of the others are black. It is not always possible to hold notes for as long as shown in the notation.

In many places, in a single bar, the 2nd and 3rd voices are syncopated to different rhythms (see image above), which it is not always possible to represent within the limited scope of the ukulele. My practice has been to substitute an appropriate note where there was one in the original, to maintain the rhythmic feeling. 

After much thought I have mostly combined the 2nd and 3rd voices in the notation to avoid a confusing forest of stems and make the timing clearer – but thereby obscuring which voice is which. When you see the original tabs as illustrated above, you will see that one is only told when to pluck a note, not its length or where it fits in the voicing, so in this transcription we can sympathise with the original lutenists.

It might well be a good idea to hold the notes in the lower voices for as long as feasible to maintain continuity. I have tried to indicate appropriate notes with a tie symbol immediately following the note. (This also improves the smoothness of the MIDI playback.) The main exceptions are in bars 44, 48 and 52 where Poulton & Lam found a hemiola rhythm in the 3rd voice, and which I have tried to indicate by stem direction.

I have made 3 versions of P4 for the ukulele.

The first arrangement [click to download] has been made for a ukulele with the 3rd string tuned a semitone lower than usual, to B3, giving the same intervals as between the upper four strings of a lute. I find that, not surprisingly, this places the notes more conveniently under the fingers than in the conventional tuning, although presenting some unfamiliar chord shapes.

The second arrangement [click to download] has been adjusted to the normal low-G uke tuning, but is a bit harder to play.

The third arrangement [click to download] has been raised from Em to Gm, to give a bit more scope for the lower voice. I have tried to indicate the voices by stem direction. If you would like the easier (?) to read unvoiced version, please ask. In any form, it's a real finger twister.