Orlando Sleepeth by John Dowland

Orlando Sleepeth may well be John Dowland's shortest work, if not in number of beats, then in temporal duration. The earliest version of this work is found in a late 16th-century lute book containing music copied by Mathew Holmes^[1]. Holmes was the Precentor and Singingman at Christ Church in Oxford and then later at Westminster Abbey in London. The first of Holmes's four lute books is catalogued as Dd.2.11 at Cambridge University Library. I have used the version made available by the Cambridge University Digital Library as the source for my transcription.

Holmes's lute tablature gives no indication of key or time signatures, leaving them for the transcriber to deduce. The time can be determined in part from the total duration of notes in each measure. That is how I concluded the first half of the piece was in ⁴/₄ and the second half in ⁶/₄. When transcribing old music, key signatures can be somewhat subjective. I believe there is a clear key change between parts and have notated it as such. Other transcriptions probably notate the whole piece in one key instead of as alternating between keys.

Orlando Sleepeth appears to have been written for the six-course Renaissance lute. The tablature uses six strings and the lowest note played on the sixth string is G. The six-course Renaissance lute was usually tuned from lowest to highest string as follows: G C F A D G. The intervals are similar to the modern guitar, consisting primarily of fourths, except the transition of a major third occurs between the fourth and third strings instead of the third and second strings. Retuning the third string on a guitar from G to F♯ would produce the same intervals (in fact that is how the vihuela was tuned). Although the difference in tuning between the guitar and lute does not prevent the music from being played on the guitar, it is necessary to transpose the music a minor third lower so that the same strings can be used and to preserve the playability of the piece. In order to play it in the same pitch range as the lute, you can place a capo at the third fret.

The lute tablature provides no indication of tempo. I believe a lot of lute music is played much slower than intended and therefore sounds quite boring. Julian Bream, in my opinion, plays this piece too slow^[2]. He also tacks on a da capo at the end to replay the first set of measures. There is no such repeat in the lute tablature. I believe the piece sounds much better played faster. If you don't agree with my choice of 140 beats per minute, play it at whatever tempo you prefer. Do note, however, that the tempo changes relative to the note durations (but not the beat) when the simple meter changes to compound meter. This is a common feature of Renaissance dance music.

I've notated a half-barre in measure 2 that requires clarification. You should not remove finger 1 from the second string to execute the barre. Instead, execute the barre to play the F♯ while allowing the A to continue to sound. If you can't pull it off, you're probably raising your elbow too high. As an alternative, you can try going straight to a half-barre at the beginning of the measure.

A Word About Transcription Accuracy

The open D in measures 13 and 15 is not an error. Other transcriptions notate a C, likely having copied secondary sources that applied an unwarranted correction.

Figure 1 shows the final two measures from Mathew Holmes's lute tablature. It is clear that the open fourth string is to be played and not the third fret of the fifth string.

Figure 1. Orlando Sleepeth Lute Tablature—Final Measures

Holmes crossed out a mistake in the penultimate measure, where he notated an open fifth string by accident. In both the mistake and the correction, he uses an a, which signifies an open string in French lute tablature. Had he intended the third fret of the fifth string to be played, he would have written a d on the fifth string as he does for the opening chord of the measure.