The beautiful Boissart mandore, part 1 of 3: The pre-history of the mandore

The history of a stunning 17th(?) century instrument, observations on its lutherie, and questions over its dating.

The Boissart mandore, dated by the V&A to 1640. (As with all pictures, click for higher resolution view.)
The Boissart mandore, dated by the V&A to 1640. Photograph by Ian Pittaway, included courtesy of the Victoria and Albert Museum, London. (All pictures, click for higher resolution view.)

In the family of renaissance plucked instruments, the mandore is the result of a union between two mediaeval string families: the oud and the lute on one side, and the gittern on the other. The resulting offspring is a small instrument with a musically significant (but alas now largely unplayed) surviving repertoire. Some actual instruments survive, and there is no doubt that the most exquisite of these is the beautiful Boissart mandore in the Victoria and Albert Museum. This article and two to follow will: (1) trace the pre-history of the mandore; (2) examine the V&A’s beautiful Boissart mandore and attempt to reconstruct its personal history for, as far as I know, the first time; (3) describe the making of a new mandore based on the Boissart model.

When Paul Baker emailed to say we had an appointment at the Victoria and Albert Museum for a private viewing of the Boissart mandore, I was delighted. The Boissart instrument is dated to 1640 by the V&A and, at the time of our viewing (January 2015) was to be retrieved from storage specifically for us to photograph, measure and drool over its beautifully carved back, neck and peg box.

Paul Baker is an early instrument maker and early musician based in Kingswinford, West Midlands. He makes gitterns, hurdy gurdies and a variety of other strings; plays mediaeval and renaissance music on wind and gurdy; and creates interactive installations for museums. I am a musician a few streets away from him in Stourbridge, specialising in mediaeval, renaissance and baroque plucked strings – lutes, cittern, gittern, bray harp, etc. – and had asked Paul to make me a mandore based on the Boissart model. ‘So’, he thought, ‘we have to go and see the original, then.’

Seeing it up close was a shock of the most thrilling kind. In person, one realises its beauty cannot be fully appreciated from the photographs on the V&A’s website, good though they are, one reason being that the skill of its carved back is even more impressive when one sees just how small the instrument is.

More of that soon. First of all, what exactly is a mandore?

The pre-history of the mandore

The mandore combines the characteristics of two families of mediaeval instruments: the lute, with its bowl back made from several glued ribs and peg box bent back from the neck at an obtuse angle; and the gittern, with its bowl back carved from a solid piece of wood and curved, sickle-shape peg box.

For the ultimate origin of the lute family tree, we go all the way back to the gut-strung oud, which has been played for so many centuries that its origin cannot be securely dated. The earliest depictions of instruments we can say with some certainty were ouds are depicted in the art of the Sassanid era of Iran, the last Iranian empire before the rise of Islam, from A.D. 224 to 651. Oud, or al-ūd in transliterated Arabic, means the wood, possibly so named originally to differentiate it from instruments made of gourd or with a resonating table of stretched skin. The quill-plucked oud entered Europe in the 9th century via the Moorish occupation of Spain. By the second half of the 13th century it was played by Christians as well as Muslims, as attested by the beautifully decorated Cantigas de Santa Maria, a richly-illustrated paean of praise to the Virgin Mary. This song book was compiled in 1260-1280 during the reign of Alfonso X (King of Castilla and other regions in modern day Spain and Portugal), clearly showing several Christian oud players.

By around 1300, Europeans had developed the oud into the 4 course mediaeval lute, also strung with gut and plucked with a quill. (A course is a group of strings in unisons or octaves to be plucked as one, so a course can refer to 1, 2 or occasionally 3 strings to be played as one.) From the beginning, lutes were made in a variety of sizes and therefore pitches.

Left: Ouds as depicted in the Cantigas de Santa Maria, 1260-80. Be careful about the number of strings and pegs shown: mediaeval portrayers of stringed instruments sometimes showed the general outline and characteristics and then included a random number of strings and pegs. The oud on the left has 9 strings and 9 pegs; on the right we have 9 strings and 12 pegs. Mediaeval artists had also not yet got to grips with dimensions and perspective – that would come in the renaissance – so the pegbox, instead of being at an angle back from the neck, is shown helpfully flattened upwards so we can see it. There is a hint of 3D, though: I love the way the musicians are standing with one foot on the frame, the other foot coming towards us beyond the frame. Right: One of the first depictions of the transformation from oud to mediaeval lute, c. 1300, embroidered on the Steeple Aston cope, England.
Left: Ouds as depicted in the Cantigas de Santa Maria, 1260-80. Be careful about the number of strings and pegs shown: mediaeval portrayers of stringed instruments sometimes showed the general outline and characteristics and then included a random number of strings and pegs. The oud on the left has 9 strings and 9 pegs; on the right we have 9 strings and 12 pegs. Mediaeval artists had also not yet got to grips with dimensions and perspective – that would come in the renaissance – so the pegbox, instead of being at an angle back from the neck, is shown helpfully flattened upwards so we can see it. There is a hint of 3D, though: I love the way the musicians are standing with one foot on the frame, the other foot coming towards us beyond the frame.
Right: One of the first depictions of the transformation from oud to mediaeval lute, c. 1300, embroidered on the Steeple Aston cope, England.

Alongside the oud, mediaeval lute and the early life of the renaissance lute, there was the gittern, which looks similar to the smaller members of the lute family except for 3 key characteristics:
(1) the backs of ouds and lutes are made of several ribs glued together, with the neck being a separate piece of wood attached; whereas the bowl of the whole gittern, including the neck, is carved from a solid piece of wood;
(2) the strings on a lute are attached to the bridge, which is glued to the soundboard, whereas gittern strings are attached to hitch pins on the edge of the instrument and pass across a ‘floating’ bridge, which is not glued and is kept in place by the pressure of the strings;
(3) the pegbox of a lute bends back from the neck at an obtuse angle, whereas the gittern pegbox has a gently curving sickle shape.

The gittern was strung with gut, had 2, 3 or 4 courses for most of its life, 5 courses for the last 50 years or so, was played with a quill, and could have a variety of entertaining carvings on the headstock, such as the head of a man, woman or animal, or an acorn (the latter as seen on the Hans Ott gittern below, one of only two surviving). The earliest depiction appears to be that painted on a pillar of the Bayeux Cathedral crypt in Normandy, dated to the 11th century. Like the oud, gitterns appear in the Cantigas de Santa Maria of 1260-1280. Certainly by the 14th and 15th centuries it was an important and popular European instrument, loved by all classes of people: late mediaeval Italian royal courts recorded the hiring of several gittern masters; Charles V of France’s court in the 14th century owned 4 gitterns; and Geoffrey Chaucer mentions the gittern several times in The Canterbury Tales, written between 1387 and 1400, describing it being played by people who sing and play music in taverns and as a popular instrument to play duets with the lute which, like the gittern, was played with a quill at this time.

Left: Possibly the earliest depiction of a gittern, painted on a pillar of the Bayeux Cathedral crypt, 11th century, Normandy. Does this instrument have 2 courses or 4 single strings? It’s difficult to know the reality, since the anatomy of both angel and gittern in the drawing is clearly an approximation. Right: Gittern on the left and rebec on the right, as portrayed in the Cantigas de Santa Maria, 1260-80. The gittern and rebec were essentially the same bowl-back instrument carved from a solid piece of wood, but the gittern was plucked and fretted and the rebec bowed and unfretted, which meant they had different bridges. If this representation is entirely accurate, then both instruments at this point shared 3 single strings and D shaped sound holes. As noted under the Cantigas ouds above, mediaeval artists could be rather approximate with numbers of strings and pegs, but the similarity in gittern and rebec here is so striking that I’m inclined to believe it.
Left: Possibly the earliest depiction of a gittern, painted on a pillar of the Bayeux Cathedral crypt, 11th century, Normandy. Does this instrument have 2 courses or 4 single strings? Does it have a carved rose or just a soundhole? It’s difficult to know the reality, since the anatomy of both angel and gittern in the drawing are clearly broad approximations.
Right: Two gitterns, one being played, the other tuned, in the Cantigas de Santa Maria, 1260-80. According to this representation, gitterns in this time and place had 3 single strings and D shaped sound holes. This makes it essentially the same design as the rebec, an instrument that was named in the 14th century but which had been evidenced in Europe since the 11th century. Rebecs had 3 single strings, D holes, and, as with the gittern, were carved from a solid piece of wood into a bowl-back instrument, the chief difference being that gitterns were plucked with a quill and fretted and the rebec was bowed and unfretted.
Left: By c. 1330 the gittern had a carved rose instead of D holes and now had 4 courses, as shown in this fresco by Juan Oliver in Pamplona Cathedral's refectory, Spain. Right: Gittern by Hans Ott, dated to c. 1450. Hans Ott made instruments in Nuremburg from 1432 to 1463. This instrument is now in Wartburg, Eisenach, Germany, one of only two surviving gitterns. As we see, the gittern now had 5 courses.
Left: By c. 1330 the gittern certainly had a carved rose instead of D holes and now had 4 courses, as shown in this fresco by Juan Oliver in Pamplona Cathedral’s refectory, Spain.
Right: Gittern by Hans Ott, dated to c. 1450. Hans Ott made instruments in Nuremburg from 1432 to 1463. This instrument is now in Wartburg, Eisenach, Germany, one of only two surviving gitterns. As we see, the gittern now had 5 courses.
Lute and gittern playing together from Pere Serra’s Virgin of the Angels, 1480s (detail), a section of an altarpiece now in the National Museum of Art of Cataluña.
Lute and gittern playing together from Pere Serra’s Virgin of the Angels, 1480s (detail), a section of an altarpiece now in the National Museum of Art of Cataluña.

We have no music for the gittern, or for virtually any other specific mediaeval instrument. There is only one piece of surviving evidence for 4 course gittern tuning, in a puzzlingly neglected source. The Berkeley Theory Manuscript was probably written by Johannes Vaillant, a 14th century Parisian music teacher, who died in 1361. His manuscript is a compendium of music theory, including drawings and tunings for the vielle (medieval fiddle), gittern, harp and psaltery. The drawing of a gittern gives a tuning of e b f’ c’’, which is most odd, and is clearly wrong, since it is contradicted by the text. The vielle is shown with a tuning of c d g c’, followed by the comment that the gittern tuning is the same, but with the 4th course loosened to a fourth below the adjacent course, clearly giving a tuning of A d g c’ which, given the size of the gittern, must mean an octave higher, a d’ g’ c’’. How was the mistake made on the drawing? If the e b f’ c’’ tuning is reversed to c f b e’ then we have a tuning entirely in fourths, as described in the text, if we also flatten the b and e. It appears that the author simply wrote the tuning backwards on the drawing and omitted the flats.

VielleGitternTuning
The vielle above and gittern below, illustrated with their respective tunings, in the Berkeley Theory Manuscript, written before 1361.

Iconography shows a smaller and larger gittern size and, at this point, absolute pitch was not fixed, so the important point is that the 4 course instrument was tuned entirely in fourths, as presumably were 2 and 3 course gitterns. This concurs with the anonymously authored Summa Musice, c. 1200 (probably French), which stated that fingerboard instruments were “tuned in the consonances of octave, fourth and fifth”.

This changed with the addition of a 5th course on the gittern, as it did on the related lute. In the early 15th century, the lute gained a 5th course. The evidence from manuscript witnesses indicates that some added the extra course at a higher pitch, some at a lower pitch. The adding of lower and lower pitches was a trend that was to continue for the next 2 centuries. By the time Johannes Tinctoris wrote his De inventione et usu musicae, 1481-1483, the gittern had also gained a 5th course (as we also see on the earlier Ott gittern, c. 1450) and the lute occasionally now had a 6th. Tinctoris stated that the gittern had “the same tuning and method of playing as the lute, though it is much smaller”. He described the tuning of 5 and 6 course lutes as fourths, except for an interval of a major third between courses 3 and 4, which was to be standard renaissance lute tuning. This, then, is also how the 5 course gittern was tuned. This means that both gitterns and lutes were tuned entirely in fourths until the advent of the 5th course, which added the ‘new’ interval of a major third between the third and fourth courses.

Iconography shows that the gittern thrived into the last quarter of the 15th century. As Tinctoris testified, the lute was now beginning to be played in a new way, with the fingertips, as well as the long-established quill technique. This reflected a radical change in music-making, from the quill playing of a single line, a melody probably with a drone, or two musical lines on adjacent strings, to the freedom of independent fingers now with the possibility of complex and delicate polyphony and counterpoint, 2, 3 or even 4 ‘voices’ played at once by a skilled lutenist. The gittern, it seems, did not make the transition from quill to fingers and was rapidly giving way to the rising popularity of the lute in this new style. The changeover was complete by 1500: there would be no more royally appointed gittern masters, as that role was now to be bestowed upon royally appointed lutenists across the courts of Europe.

Angels playing 5 course gitterns with quills, painted by the Master of the Lyversberger Passion (detail), 1460-1490, Cologne, Germany.
Angels playing 5 course gitterns with quills, painted by the Master of the Lyversberger Passion (detail), 1460-1490, Cologne, Germany. Except in this painting, gitterns always had floating bridges, but here we have fixed bridges. This may be a precursor of the mandore design, or simply a painter working from memory, getting his lutes and gitterns mixed up.

The arrival of the mandore

And so we come to the mandore, the first evidence for which is in Sebastian Virdung’s Musica getutscht. Published in Basel, Switzerland, in 1511, it is the earliest printed western treatise on musical instruments.

Lute and quintern (mandore) in Sebastian Virdung’s Musica getutscht, published in Basel in 1511 on the left; and the same two instruments on the right from Martin Agricola, Musica instrumentalis deudsch, published in Wittenberg in 1529.
Lute and quintern (mandore) in Sebastian Virdung’s Musica getutscht, published in Basel in 1511 on the left; and the same two instruments on the right from Martin Agricola, Musica instrumentalis deudsch, published in Wittenberg in 1529.

Here we have a labelled woodcut of a rib-back 5 course “quintern”. It has the sickle peg-box and 5 courses of the gittern but the multi-ribbed back and glued bridge of the lute. This instrument had a variety of names: quintern, quinterna, mandörgen, mandurinichen, mandürichen, bandürichen, mandoër or pandurina in Germany; mandour in Scotland; mandorre in France; mandora or mandola in Italy.

That the same instrument was known by a variety of names was commonplace: different regions and nations had their own local names for the same instrument; or, to put it another way, sometimes the same instrument had different names in different places; and, confusingly, sometimes the same name was used for entirely different instruments. For example, the 4 course renaissance guitar was also referred to as the gittern or quintern in France and England from the 1550s, though it is entirely unrelated to the mediaeval bowl-back gittern or the renaissance quintern/mandore. By 1652, when John Playford published his Booke of New Lessons for the Cittern and Gittern, gittern was also the name for a small cittern. (Such potential confusion over names for the unwary musicologist is one explanation for Wikipedia’s and many others’ utterly muddled entries for the histories of the gittern and guitar.)

The mandore was made in a variety of sizes, but typically and on average had a vibrating string length of 30-40cm. This is a very similar size to the gittern, from what we know. We must beware of drawing broad conclusions from slim evidence, but paintings do appear to show two general gittern sizes, corresponding roughly to the two surviving instruments we have, one from the east and the other from the west of Europe. The instrument excavated from Elblag/Elbing, Poland (it’s been both sides of the Polish/German border, and so has two names), dated 1350-1450, has a vibrating string length of 30-33cm, depending on bridge placement (the bridge became detached on impact and was found beside the instrument), and is the fretless eastern European form of gittern, properly called a koboz or kobza. The western true gittern was made by Hans Ott gittern of Germany, c. 1450, and has a vibrating string length of 44cm.

The mandore flourished for nearly 200 years, starting from the early 16th century, just after references to the gittern died out. Some mandores were carved from a single block of wood like gitterns, others were multi-ribbed like lutes; some had the gitterns’ sickle-shaped pegbox, others the straight obtuse-angled pegbox of the lute; all had a fixed bridge. Stringing was varied: it could have 4 courses (a single chanterelle or top string with the rest double), 4 or 5 single strings, or sometimes 4 single strings with the lowest 5th string doubled with an octave.

Taken together, these facts may indicate that the mandore was a development of the gittern under another name, a new confluence that combined features of both the lute and the gittern.

One mandore tuning seems to indicate this. There were several tunings:

(a) a combination of fourth and fifth intervals to make ‘open tunings’ – for example, g’’ / c’’ / g’ / c’ / g makes an open C chord or d’’ / g’ / d / g makes an open G chord, reflecting the idea as far back as the Summa Musice, c. 1200, that fingerboard instruments were “tuned in the consonances of octave, fourth and fifth”;

(b) or this modified for some pieces with the highest course tuned down a major second or minor third (reflecting practice in mediaeval fiddle tunings);

(c) or the whole tuned at the same intervals as the first 5 courses of a renaissance lute, but at a higher pitch, being a smaller instrument, reflecting Tinctoris’ statement that the gittern has “the same tuning and method of playing as the lute, though it is much smaller”.

Mandore playing styles also appear to confirm confluence: either played with a quill, as was the medieval gittern and lute, or with fingertips, as was the renaissance lute. German musician and musicologist Michael Praetorius, in his Syntagma Musicum, 1619, described varied mandore playing techniques: with a feather quill; with a rapidly playing single finger; or with two or more fingers. To this, Marin Mersenne, in his Harmonie Universelle, Paris, 1637, added a fourth way combining both quill and finger technique: playing with the thumb and with a quill tied to the index or another finger.

The mandore was particularly popular in Germany, France and Scotland. Printed books and written manuscripts of mandore tablature provide evidence of its use and popularity, from a lost French book by Pierre Brunet, 1578, to other French, German and Scottish music with distinct national musical characteristics, through to the last surviving mandore music in the English Talbot manuscript (which also includes lute and cittern music) of c. 1695. We also have some beautiful surviving examples of 17th century instruments from France, Italy and Austria, of both the carved-back gittern and rib-back lute type.

A distinct Italian rib-back version of the mandore with its own tuning was developed from around 1650: the mandolino, i.e. small mandola (mandola being the Italian mandore), which was the beginning of the mandolin. But that’s another story.

All of which sets us up for a discussion of the beautiful Boissart mandore in the V&A in part 2