They don’t make ’em like they used to

Chansonnier Cordiforme (ca. 1470)
Chigi Codex (1498-1503)
Capirola Lutebook (1515-1520)




Chansonnier de Zeghere van Male (1542)


Neue Lauten-Früchte (1676)

Dowland’s Lachrimae

Some things never change. Lachrimae (‘tears’)by English lutenist John Dowland, was one of the greatest hits of its time, and those who hear it now immediately understand why. When I first discovered this music (i.e. ‘early music’, so called, especially Renaissance music), one of the biggest surprises was how easy it is to understand. There is nothing arcane about it, no cryptic musical language to decipher. Lachrimae is as unmistakably beautiful today as it was in Dowland’s time.

The piece began its life in the 1590s as a lute pavan. It can be found in many manuscripts, but the first to have a certain date is William Barley’s A New Book of Tabliture published in 1596. A song version, Flow my Tears, was later included in Dowland’s Second Booke of Songs (1600). In 1604, he published Lachrimae, or Seaven Teares, a collection of seven variations (all pavans) on Lachrimae for five viols and lute. (This was only the third book of consort music in England, and the first music to be specifically written for five viols and lute.) Eight years on, Dowland’s last book, Pilgrimes Solace (1612), ends with a Lachrimae galliard.

These are but a fraction of the manuscipts available. Lachrimae was very popular abroad too, even after Dowland’s music had gone out of fashion in England. According to Peter Holman, author of a book on the 1604 Lachrimae collection, Dowland’s music survived longest in Netherland, particularly the songs Now o Now I Needs Must Part and Come Again, sweet love doth now invite, the Earl of Essex Galliard, and of course Lachrimae. Dowland’s music became popular abroad for two reasons: one, he spent many fruitful years on the continent as a virtuoso lutenist; and two, English dance music was very fashionable in Northern Europe in particular. There is also plenty of evidence of composers after Dowland using the ‘tear motif’ from Lachrimae — the four-note falling sequence that starts the piece — though they may well have been unaware of its provenance.

So when we refer to ‘Lachrimae’, it’s not clear exactly what we mean. There are many ‘Lachrimae’s, and no where is there a definitive or original version. One of the problems is that Dowland, and indeed most musicians contemporary to him, exercised little control over the arrangement, distribution and performance of his music. We know this annoyed Dowland. In 1597 he complained that his pieces had ‘lately printed without my knowledge, falce and unperfect’. Performances would have included some improvisation, certainly with regards to ornamentation. (And indeed when I play it on guitar, I fiddle around with the divisions, maybe change a chord a bit to make it more playable.) Some of these versions have survived thanks to personal manuscripts. A fun example is to listen to the Fantasie P. 1 found in Paul O’Dette’s John Dowland: Complete Lute Works Vol. 5. It is a wildly ornamented arrangement written in the early 17th century. (Unfortunately it’s not on YouTube, but it is easily found on Spotify etc. If you read lute tablature, you can find the manuscript here — the Fantasie is on page 27.) Part of the appeal of the lute repertoire is that it gives the player so much freedom. The tablature is skeletal, with no expression markings and seldom anything more complicated than three ornament signs. And we aren’t even always sure what the ornament signs indicate!

Here’s a performance on lute using the manuscript Dd.2.11 (1590s); I think it’s image 158.

Some of Dowland’s contemporaries made arrangements of Lachrimae. Here’s William Byrd’s:

And see this video, at 2:39, to hear Thomas Morley’s 1599 arrangement for consort:

In Lachrimae, or Seaven Teares, Dowland writes: ‘And though the title doth promise teares, vnfit guests in these ioyfull times, yet no doubt pleasant are the teares of which Musicke weepes, neither are teares she alwayes in sorrowe, but some-time in ioy and gladnesse.’ A rare sign that Dowland wasn’t entirely the miserabilist he seems. This will be the last video, Lachrimae Verae, the final Lachrimae from that book. But quickly before that, Peter Holman again: ‘Today, Dowland’s music is more popular than at any time in the 400 years since it was written; indeed, Lachrimae [referring to the 1604 book] is probably the most recorded and performed collection of instrumental music before the Water Music or the Brandenburg Concertos’.

The Orpharion

I was listening to the lutenist Paul O’Dette play John Dowland’s Mrs Winter’s Jump and thought, Well chop off my legs and call me shorty, that sounds like a virginal! Listen for yourself:

Turns out it’s an orpharion. Typical of the Renaissance, the name ‘orpharion’ is derived from Orpheus and Arion. It was invented in England at the beginning of Elizabeth’s reign, and became a fairly common alternative to the lute. It’s tuned the same way as a lute, and various contemporary books list it as an alternative to the lute, but there seems to have been very few books written specifically for the instrument.

While the lute was clearly the more esteemed instrument, the orpharion may well have been equally popular in the home: in 32 examples between 1565 and 1648 of household inventories which mention musical instruments of any kind, the bandora (its larger sibling) and orpharion occur as frequently as the lute.

An orpharion built by Francis Palmer in 1617 (see more photos here)

The orpharion differs with the lute quite significantly in terms of its shape, most notably the flat back. But the more important difference is its use of wire strings instead of gut. The subsequent brighter, metallic sound, and the more limited timbre and dynamic range it presents, are no doubt why I nearly mistook it for a virginal. Harpsichords, unlike the piano which uses hammer-action, are plucked instruments just like the lute and orpharion, so the method of attack is also similar.

Right-hand technique, however, differed between the orpharion and the lute because of the wire strings. William Barley published the first collection of music specifically for orpharion in 1596, and in it he wrote that

whereas the Lute is strong with gut stringes, the Orpharion is strong with wire stringes, by reason of which manner of stringing, the Orpharion doth necessarilie require a more gentle and drawing stroke than the Lute, I mean the fingers of the right hand must be easilie drawen over the stringes, and not suddenly griped, or sharpelie stroken as the Lute is: for if yee should doo so, then the wire stringer would clash or jarre together the one against the other

Here’s an orpharion being played. There seems to be very few videos on YouTube of the instrument — and half of them are by this guy:

Mille Regretz

Mille Regretz (‘A Thousand Regrets’) is a gorgeous and once popular song thought to be written by the Renaissance composer Josquin des Prez. Its fame is invariably linked to King Charles V, ruler of the Holy Roman Empire, who was said to be very fond of it, so much so that a good handful of contemporary composers wrote versions for him. This is the original Josquin version (well, some dispute its authorship, but for the sake of this blog post I’ll refer to it as his).

Luys de Narváez was a Spanish vihuelist during this time and intabulated the song for his instrument. The vihuela was a small guitar-like instrument particular to Spain, of which Charles was king, and was for all intents and purposes very similar to the lute. Narváez, who wrote one of the few extant books of vihuela music, gave the subtitle ‘la cancion del Emperador’ (‘the emperor’s song’) for his intabulation of Mille Regretz, the most explicit evidence we have of Charles’ fondness for the song. Unlike the Josquin version, there is more ornamentation as the vihuela, like the lute, cannot possibly hold a note for as long as a singer. This is a fantastic piece to play on the guitar too, and not at all difficult. Just tune the G string to F#, and maybe capo at the third fret, and you can play it from the original score. Here’s Julian Bream playing it on the vihuela:

Cristóbal de Morales was another Spanish composer who used the song in order to gain the king’s favour. Missa Mille Regretz is a parody mass — that is, not work of comedy but rather one that steals material, often from secular songs — and contrary to modern values this wasn’t looked down upon. The Kyrie from Morales’ mass setting:

Nicolas Gombert’s is the only other contemporary arrangement I’m aware of, and like Morales’ it is expanded to six-parts. I was not familiar with it until today, but it is very beautiful and seems to be the most intricate arrangement:

We can assume that there were also plenty of non-scored arrangements of this popular song, and if you browse on YouTube you will find plenty of modern ones. This is a particularly attractive version by the American early music group Voices of Music for organ and cornett:

King Charles V is perhaps better known to music lovers by his strange appearance in Verdi’s opera Don Carlo. Just at the end, he emerges from his tomb Commendatore-style as King Philip II and the Grand Inquisitor are chasing after Don Carlos, proclaiming

‘The sorrows of the world
follow us into the cloister.
The struggle in our hearts
will find peace only in heaven!’

At which point he rescues Don Carlos.

Below is the final scene of Don Carlos. Charles emerges exactly 10 minutes in — he’s the one holding the enormous, ornate cross. It is a total what-on-earth-just-happened moment that suddenly comes in the last minute of this grand three hour opera.

What an ending!

Valentine Bakfark and Sixteenth Century Lute Music

The name Valentine Bakfark is now relatively obscure. A sixteenth century Transvylanina lutenist, he was among the most respected lutenists of the time, admired for his great technical innovation on the instrument. Five hundred years later, and the unfamiliar name of ‘Bakfark’ probably only conjures up images of a kind of foul regurgitation.

Bakfark himself

But regurgitation isn’t an entirely incorrect way to describe Bakfark and his contemporary’s music. Born sometime between 1507-1508 until his death, most likely from the plague, in 1576 (aged 69), Bakfark’s era was one long before the cult of originality and moreover before cults of personality. Even in the early seventeenth century with English lutenist John Dowland, who immortalised his famous melancholy image in song form, Semper Dowland, semper dolen (always Dowland, always doleful), such cults had yet to seriously develop. Musicians in this time and for much of history were not necessarily respected geniuses at the forefront of high culture. They could be people with modest backgrounds (Mozart, for example), and were often treated appallingly by the princes who employed them in what was virtually servitude. Composers from Monteverdi to Haydn were made miserable creatively and financially by ungrateful princes, even more so than many modern composers think they are by ungrateful audiences. (There is a good chapter on this in Tim Blanning’s The Triumph of Music.)

Furthermore, there was a clear hierarchy of music — and lutenists weren’t at the top. Hierarchies and divisions in music were particularly firm in the sixteenth century. Sacred music was indisputably of greater value than secular music. And instrumental music was considered inferior to vocal music. But this too was a sacred-secular division, and is why Palestrina composed only for voice. Within the hierarchy of instrumental music, the most lowly was dance music, usually homophobic, and the most respected was that which imitated vocal polyphonic music.

Unlike today, transcriptions were not considered of lesser artistic value. As Bakfark biographer Istvan Homolya explains, ‘the originality of ideas was not an essential aesthetic criterion. Artistic value was provided by the skill or ingenuity of elaboration’. The idea of originality, in fact, was yet to be born. The majority of Bakfark’s works were intabulations (transcriptions) of vocal polyphonic music. (And it wasn’t until the Baroque era that instrumental music really ceased to use vocal models.) Among instruments, the lute was the preferred one because it was portable and, moreover, capable of polyphony (playing more than one melodic line simultaneously). However, the lute is also thoroughly unsuited to polyphony. This may sound like an impossible contradiction, but it is the act of overcoming this contradiction that made Bakfark so remarkable.

A singer can sustain a note for a long period of time. On a lute, the note will die very quickly, in a matter of seconds. The way around this is ornamentation, so that you never linger on one note for too long. Bakfark was especially innovative when it came to ornamentation, and among other things he was one of the first pioneers of the mordent. He even used chromatic accidentals. Indeed, the dissonances and harmonies in Bakfark’s music can be quite surprising. In the third bar of the extract below, you have the unexpected use of a C♮ auxiliary note instead of a C# (in the key of B minor) that moreover sounds simultaneously with the F#, creating a tritone. I’ve noticed that in the Julian Bream recording, Bream ‘corrects’ the note to C#. Bakfark had a similar problem at the time, familiar to eccentric moderns like Charles Ives, in which the printers would try to correct these notes for him.

He also wrote much denser scores than his predecessors. While those like Francesco da Milano (1497-1543) primarily used the minima (minim, or half note) as the base rhythmic element, Bakfark primarily used the semiminima (crotchet, or quarter note). Shorter notes than that were ornamental rather than melodic (the shortest note at the time was the semifusa, the semi-quaver or sixteenth note). Naturally, this gave Bakfark’s work a much faster sense of movement.

Lute music was written in tablature form, not unlike guitar tablature today. This poses another problem as tablature, unlike musical notation, cannot notate the continuity of voices. It wasn’t possible to notate two musical lines separately. For example, these are the first lines of Bakfark’s Fantasia I in modern notation:


And here it is in tablature form (excluding the last bars from the above extract):

bakfark-e1494359774615.png(I’ll come onto how to read the tablature in a bit, though much of it should be plain.)

In the 5th bar, whoever transcribed the piece decided (rightly, I would say) to hold the B until the next bar. In the tablature, that corresponds to the 2 in bar 5 on 2nd string down. As you’ll notice, there is no indication that it’s to be held. Maintaining the integrity of the voices would have been up to the judgement and skill of the performer.

The lute is also capable of fewer independent lines. The lutenist only has six strings and four fingers (on each hand) at his disposal. Furthermore, the separation of parts will be much less distinct on a lute than in a vocal ensemble, or even in, say, a viol ensemble. Before Bakfark, lute music would usually be in two or three parts. Quite astonishingly, Bakfark managed to faithfully intabulate even six-part polyphony on the lute. Yet despite pushing the lute to do things that were the least lute-like, he nevertheless, as lutenist Daniel Benko has commented, ‘was able to produce perfectly the correct polyphonic transcription of vocal works’. Here’s Benko performing one of the two extant Bakfark intabulations using six-part polyphony:

One is struck of the wonderful musical diversity between nations and cities in the sixteenth century. And when it came to the lute, this meant, among other things, that there were various national styles of tablature. The main types were French, Italian and German. There are further variations — the English for example put notes in between, rather than on, the lines, and might notate the rhythms differently, but otherwise maintained the French style by and large. The facsimiles of Bakfark’s work that I’ve come across use both the French and Italian styles. I assume the books were printed in each region according to the prevailing style. Lutenists were not expected to be familiar all the varieties of tablature.

French lute tablature is probably the easiest to read. The lines of the stave, or the spaces in between (or below and above) the line, correspond to the courses on a lute. The lower on the stave, the lower the course, and vice versa. The fret positions are written using letters: ‘a’ means the open string, ‘b’ the first fret and so on. (Italian tablature used a numbering system instead and, confusingly, reversed the order of the strings so that the lowest string would be at the top of the stave. The excerpt included a few paragraphs above is an example of this style.)

The score doesn’t indicate precise notes. There were many different lutes, each with differing numbers of courses and tuned to different pitches. Moreover, players would often just tune the top string as high as it would go and then tune the other strings by it. (John Dowland for some reason did the same but with the lowest string.)

The flags above the stave notate the rhythm. A vertical line is equivalent to a semibreve, and each horizontal line halves the length. So the first three beats are: semibreve, dotted minim (you might have to enlarge the score to see the dot), and crotchet. The only other detail is the diagonal line, for instance in the 3rd bar, second row, that specifies the length of time that note should be held.

Hopefully I didn’t lose anyone. Just thank me for not attempting to explain the bafflingly cryptic German system!

Here is Bakfark’s Fantasia I again, but using French tablature:

Bakfark Fantasia No. 1

Although Bakfark wrote fewer works than most (and indeed, destroying much of what he did produce, leaving us with about 40 extant works), he took an unusual personal interest in his work being published. It would seem he was very protective of his work. The two books he published, Lyons and Krakow, were carefully crafted and published at Bakfark’s own expense. His predecessors did not care so much about having their work published, and indeed much of it wasn’t.

Here’s Daniel Benko again playing the Fantasia I. Though reasonably short, it’s hard to take in at first, each section blended together with no hard cadences to help negotiate its structure. (At least, that was my experience listening to it.) You can find the entire score — the entire book in fact — here. IMSLP also has a version in Italian tablatureFellow guitarists can easily play it from the score by tuning their G string to F#.