The vielle or medieval fiddle was the most popular instrument in its heyday for religious and secular song accompaniment. It first appeared in western Europe in the 11th century and continued to be played until the middle of the 16th century, flourishing in the 12th and 13th centuries. There is a wealth of vielle iconography, which can tell us a great deal about the variety of its form and the context of its use. There is a medieval source for its tuning, Jerome of Moravia in the 13th century, who gives 3 tunings, leaving us with some puzzles as to what exactly they mean in practice. Our only renaissance tuning source is Johannes Tinctoris in the 15th century, which isn’t entirely clear in its meaning.
This article provides a detailed discussion of the different ways in which we can make sense of historical fiddle tunings and, in the light of that, a closely argued case for the relationship between the vielle and the crwth or bowed lyre, demonstrating that they were identical in style, having more in common with the hurdy gurdy family than modern bowed strings.
There are two editions of this article. This one included detailed analysis. For an introduction for the general reader, go to On the (medieval) fiddle: a short introduction to the vielle.
The importance of the vielle
The importance of the vielle in mid to late medieval European culture is clear from surviving historical artefacts, paintings and manuscripts. From the 12th century on, the vielle was associated in literature with both amateur and professional players (along with the citole and harp), as an instrument to play dance music; to accompany others’ secular songs; to self-accompany secular singing; and as an accompaniment to religious song.
The vielle found favour in royal and noble households. The seal of Bertan II, Count of Forcalquier, France, dated 1168, has him armed with a sword and shield on horseback on one side, and sat playing a vielle on the other; King Ottokar II of Bohemia, writing in the early 13th century, gave the names of 17 fiddlers at King Manfred of Sicily’s court, implying that there were many others; and Duchess Isabella, wife of Duke Philip the Good in 15th century Burgundy, employed two blind Portuguese lutenists who also played vielle and other “soft”, i.e. indoor, instruments.
The medieval fiddle is a regular element in decorated manuscripts and paintings of the period, and features in religious paintings played by angels or humans in attendance of the holy, sometimes on its own, often among other instruments.
Minstrels, musicians in the employ of wealthy households, used the fiddle to accompany others’ singing; and jongleurs, French musician-poets who performed both in the service of the wealthy and freelance, used it to accompany dancing and their own singing of chansons de geste, songs of heroic deeds, a popular genre in the 12th and 13th centuries.
The troubadours, French Provençal lyric poets and singers of the 11th to 13th centuries, now have the reputation of having the vielle as their favoured instrument, as do the trouvères, the epic poets and singers of northern France in the 11th to 14th centuries. This idea is repeatedly cited but is not backed up by the evidence, which does not tell us how troubadour and trouvère songs were accompanied, or indeed if they were. Possibly some song types in their respective repertoires were accompanied and others unaccompanied, but this is an argument from silence. What seems most likely to have happened is that some modern writers have extrapolated from minstrels and jongleurs and presumed troubadours and trouvères to have had the same performance practices.
Distinctive structural features
Looking at the consensus of the many available images, a modern bowed string player would immediately notice the vielle’s distinctive structural features.
The first is the option, on a 5 string vielle only, of a bourdon. Harps and psalteries had bourdons, too. In their case, it was a low pitched string, sometimes pitched several steps below the next string in the sequence, that acted as a drone string. In the case of the vielle, it was a low-pitched string placed off the fingerboard and attached to the side of the pegbox by passing through a hole, so it could only be played open. As we see from the angle at which the bourdon leaves the bridge to reach the pegbox on the angel in green painting on the right, this made it possible to play the bourdon by plucking with the left thumb. Dominican friar and music theorist, Jerome of Moravia, writing in c. 1280, made it clear that the bourdon could also be bowed.
The second and third features are a variety of body shapes, some very wide, and a variety of bridges. The bridge height was either less than its width or equal to its width (unlike on a violin), making for a low action (the relative height of the strings to the instrument). The iconography suggests most strongly that the bridge was often flat, unlike the marked arc of a violin bridge, severely limiting the angle at which strings can be bowed and positively necessitating drone playing on all strings at once. Of course, even on a completely flat bridge, string slots in the bridge could be made of various depths to create the string arc violinists are used to, enabling single string playing; but the iconography and the logic of the tuning system do not suggest this.
Above we see the three types of bridge on a vielle. In all these cases, there is a separate string-holder which has a particular effect on string pressure, creating downward tension on the soundboard, whereas a fixed and glued bridge without string-holder creates an upward pressure on the soundboard.
On the left, from Hans Memling, Angel Musicians, Antwerp, 1480s, we see a flat bridge and therefore strings of equal height. Often this would be a simple wooden block. In this case and in some other iconography, the block is castellated, possibly to allow tuning of individual strings. Luthier Chris Doddridge tells me that castellation also has a beneficial effect on string vibration and overall timbre. Castellated bridge or not, strings would generally have to bowed as one humming block.
Top right is a detail of a vielle from the San Francesco Church, Giffoni Valle Piana, Italy, early 15th century. Here the strings disappear into the string holder on a supporting post (seen more clearly on some paintings, for example, Domenico di Bartolo’s Madonna of Humility, shown further down the page). On such iconography, the strings again appear to be flat, though they could technically be made as an arc. On this design, the string holder is on the edge of the instrument, as we see here. On some iconography, the holder covers up to half the body of the fiddle. It is difficult to see how this could work in practice, so it appears to be iconographical license.
The third bridge type, bottom right, shown next to the bow, is seen on the anonymous A ball in Augsburg, c. 1590-1595. This curved bridge is seen from the 14th century, allowing any string to be played individually, as on a modern violin.
Overall, historical images show us a range of sizes; a variety of shapes, most typically having a body shaped as an oblong box with rounded edges or with a gently curving waist; the neck was sometimes fretted with gut, more often unfretted; string numbers were usually 4 or 5, but sometimes 3 and possibly 6; it was played on the shoulder, or against the upper chest, or on the knee facing forward like the rebab or viola da gamba; and played with bows of various sizes, lengths and curvatures.
Left: Two details from The Way of Salvation fresco in the Spanish Chapel, Florence, by Andrea di Bonaiuto, 1365. Here we see the low, flat bridge and the use of a bourdon on a 5 string vielle. The position of the thumb implies the bourdon is plucked while the other strings are bowed. The illustration bottom right, from the Boethius manuscript, De Musica, 14th century, strongly suggests this, too. Compare the shape and the low height of the vielle bridge from the green angel in the Immaculate Conception Altarpiece above and from The Way of Salvation fresco with that on the detail of the baroque violin, top right, from Orazio Gentileschi, Young Woman Playing a Violin, c. 1612. The low, flat vielle bridge would facilitate playing all the strings as one sonic block, not possible on the curved violin bridge, and would make it much fuller-sounding but quieter than the violin.
A variety of sizes, shapes, strings, and bourdon / fret choices (click picture to open in new window).
From left to right:
Lincoln Cathedral, 13th century: 5 strings, no bourdon, fretless.
Gloucester Cathedral, c. 1280: 3 strings, fingerboard not visible.
From the Chapel of San Nicola in Tolentino, Italy, after 1305: 5 strings, bourdon, fretless.
Manuscript in Austria, c. 1300–1350 (Universitätsbibliothek Graz 32, fol. 106v): 4 apparent strings but 6 pegs, no bourdon, fretted. There are occasional depictions of 6 string vielles. Does this image suggest 6 strings in 3 or 4 courses, or is this artistic license we shouldn’t take literally?
Hans Memling, Antwerp, 1480s: 5 strings, no bourdon, fretted.
Luckily, we have more than iconography to tell us about the medieval fiddle: there are a few extant instruments, remaining in various degress of wholeness.
Surviving instruments I: the Elblag “fidel”?
In 1986–1989, during excavations in Elblag / Elbing, Poland (it’s been in both Polish and German territory, and so has two names), several instruments were found in the latrine of a wealthy house: two bone pipes and a recorder from the 14th–16th century; a koboz dated 1350–1450 (misidentified as a gittern, the western European counterpart to the eastern koboz); and, from the same era, an instrument carved from a single piece of linden wood with a spruce soundboard, described by excavators as “a very small fidel”, measuring 33.3cm long.
The lutherie of the instrument is extremely rough: the back of the asymmetrical body shows traces of hewing to carve the hollow bowl and the uneven back of the neck has what can only be described as hacking marks. It was not smoothed or polished. It must be significant that the area around Kowalska Street, where it was found, was inhabited by some of the poorer people of Elblag. The instrument, having fallen into the latrine and found 5 or 6 centuries later, is damaged but otherwise well preserved. It is missing the lower part of body, part of the soundboard, the bridge, and the peg heads. It has 4 broken peg stumps and 4 grooves in the nut of roughly equal spacing for 4 strings.
The excavators’ description, “similar to a baroque dance master violin”, likens it to a 17th century kit or pochette, similar in size and sometimes the shape of the rebec design. So is this a rebec or a vielle? I will conclude that, in all likelihood, it is neither, but before we arrive at that conclusion it is worth briefly exploring some iconography to explain why the distinction between vielle and rebec is not always easy to make.
The first extant evidence of the word rebek is in an early 12th century table of Arabic and Latin terms (Paris, Bibliotheque Nationale, Ms. lat. 14754). However, there is no illustration. Bearing in mind the slippery problem of medieval instrument naming – several instruments with the same name, the same instrument with several names – we cannot be sure what rebek meant in this context, and the rebec as we know it does not appear with any certainty until the 14th century. All the evidence suggests that rebecs always had 3 strings, whereas vielles had 3, 4 or 5 strings. It follows that the two church elders carved on the Church of Sainte Pierre in Moissac, France, dating from 1115–1130, pictured above right, are holding/playing vielles. Though the shape is that of the later rebec, the 5 tuning pegs (4 extant, 1 missing in both cases) indicates medieval fiddles.
The earliest apparent example of an identifiable European rebec is in the Catalan Psalter, c. 1050 (bottom left), 65 to 80 years before the Church of Sainte Pierre. A similar instrument (bottom right) is in the English manuscript, BL Arundel 91 (f. 218), a Passionale (Lives of the Saints) originating in Canterbury in the first quarter of the 12th century. Both are apparently single strung, as rebecs were. Though they look like rebecs, it is most likely that they are both vielles in the shape that was later to be used for the rebec, due to the lack of any substantial corroborating evidence for rebecs at this point. Both are in one of the vielle shapes, and the instrument on the right has 4 strings upon closer examination, though whether this is to be believed is a moot point, since it has 3 tuning pegs!
The point about shape alone not identifying the ‘rebec shape’ as a true rebec is made clearer by the fact that the rebec never had a bourdon, whereas vielles often did. The first three illustrations below show 5 string vielles with bourdons; the last, if it is to be taken literally, shows 3 strings plus a bourdon (all other evidence shows it should not, as only 5 string vielles had bourdons). All vielles are in the shape later used for the rebec.
From left to right: a detail from an English psalter, first half of the 12th century (Bibliotheque Municipale, Lunel, France, MS. I, f. 6); a sculpture by Bennedeto Antelami, c. 1180, in the baptistry of Parma Cathedral, Italy; from the Lambertus Treatise, French, 13th century (King David with musicians on f. Av of Bibliotheque Nationale, Paris, MS. lat. 6755(2)); and a 13th or 14th century illustration added to 10th century manuscript of De Musica by the Roman senator and philosopher, Boethius, originally written c. 510 (Biblioteca Ambrosiana, Milan, Italy, Cod. C. 128).
Distinguishing with certainty between the rebec and vielle must therefore take into account a variety of factors. Both instruments were carved from solid wood. The vielle had a variety of body shapes and the rebec only one. The rebec had 3 strings and the vielle had between 3 and 5. The 5 string variety of fiddle often had a bourdon, the rebec never did. Vielles were sometimes fretted, often unfretted, and rebecs were always unfretted. This means that, in some cases, it would be only the tuning – not visible in iconography – and the rebec’s arced bridge which distinguish it from the vielle, a distinction made even finer by the fact that some vielles also had an arced bridge and the bridge, in any case, is often not visible in depictions. Christopher Page observes a further distinctive factor of instrument names built upon r-b – rubeba, rubeb, rubeba, rebecca, rebeccum, rebec: their small size. Jerome of Moravia in the 13th century, and Jean Charlier de Gerson and Johannes Tinctoris respectively in the 15th century, all describe the “rebeba” / “rebecca” / “rebecum” as small, or smaller than the vielle.
Rebec or fiddle? First on the right are details from a late 15th century manuscript, Le livre des échecs amoureux moralisés (The book of chess lovers moralised), produced for Louise de Savoy (1476-1531), mother of King Francis I of France, now in the Bibliotheque Nationale, Paris. The mermaid’s instrument is shown with 3 lines for strings, clearly meant as 3 courses since there are 5 tuning pegs. This is a 3 course vielle with an arced bridge in the shape also associated with the rebec.
The next two images were painted by Melozzo da Forlì in c. 1480 in the Basilica dei Santi Apostoli, Rome. First, an instrument with an arced bridge and very much in the rebec shape, but with 4 strings and frets. Judging by the criteria established above, this is therefore a 4 string vielle. This is a particularly interesting example, as on the above ‘rebec-shaped vielles’ it has not been possible to see the shape of the back: this shows it to have a bowl identical to the rebec.
The next instrument is a clearly-depicted 4 string vielle with a flat bridge. The juxtaposition of these two Melozzo da Forlì instruments from the same source illustrates the contemporaneous multiplicity of vielle shapes.
Finally, the form of this 3 string fiddle, a detail from the painting of baby Jesus’ bath in Köln Cathedral, Germany, 13th–15th century, is the shape and larger size we would ordinarily expect of a vielle, and the typical shape luthiers now use for their reproduction instruments.
On this basis, the rebec is distinguished from the vielle of the same teardrop shape by having a combination of no frets, an arced bridge, a relatively small size, and 3 strings. The Elblag instrument was carved broadly in the rebec shape with no evidence of frets and no possibility of frets, due to the very rough and uneven shape of the neck, and it is only 33.3cm long. These factors in isolation might indicate a rebec, but other factors throw up a range of insurmountable problems preventing clear identification.
(i) The Elblag instrument has 4 nut grooves and 4 pegs for 4 strings. Together with the fact that it has a single sound hole in the centre of the soundboard, unknown on any rebec, it cannot be a rebec and is too small to be a vielle.
(ii) Identification of this “fidel” as a bowed instrument is uncertain. Neither an indicative bridge nor a bow was found with it. This does not mean it was not bowed, just that identification as a bow or plectrum instrument cannot therefore be made.
(iii) The instrument was found in Poland. As Alexander Lyczywek of the University of Gdansk laments in his study of the Elblag findings, research into musical history and instrument organology in eastern Europe is sparse. This, I have to conclude, is what led Polish researchers to identify the eastern koboz found in Elblag as a western gittern, an error repeated in all literature I have found on the Elblag instruments. The “fidel” may be a gudok, an eastern European bowed instrument similiar to a rebec, but its appearance in Elblag does not match gudoks played today, and I have found no research into its historical evolution, so this identification is similarly impossible.
In conclusion, we can identify criteria for distinguishing between the sometimes visually similar rebec and one form of the vielle, but we lack any reliable criteria for identifying the Elblag instrument, even whether it was played with a plectrum or a bow.
Surviving instruments II: Saint Catherine de Vigri
Saint Catherine de Vigri (1413–1463) was first lady-in-waiting to Margherita d’Este, and later joined the Order of Poor Clares of Corpus Domini at Ferrara before becoming Abbess of their first convent in Bologna, Italy. Canonised in 1712, Saint Catherine was a musician, and the vielle now on public display in the Corpus Domini Church at Bologna belonged to her. The body and neck of her 4 string vielle, 49cm long, are carved from a single piece of wood, and it has a flat bridge. Not only is the fiddle and bow on display in a glass case, so is she, sat on a chair, her skin now blackened with age.
Surviving instruments III: the Mary Rose
The only other two fiddles known to have survived are English. Ironically, they were not preserved through being in someone’s private collection, but because they sank to the bottom of the Solent – the strait between the Isle of Wight and mainland of England – on board Henry VIII’s ship, the Mary Rose, on 19th July 1545. It wasn’t until 1971 that the wreck was discovered, and not until 1978 that the initial excavation work began. Unsurprisingly, then, neither of the found fiddles have survived intact; but the remains do show that these vielles were also carved from solid wood.
Together with the other surviving fiddles and observations from iconography such as that from the Church of Sainte Pierre in Moissac, and from the Gate of Glory, Santiago de Compostela, shown above, this strongly suggests that all vielles, like rebecs and gitterns, were hewn from solid wood rather than made from wood glued together in separate sections, like lutes.
Three vielle tunings
Though we have plenty of iconography for the vielle, its manner of playing needs some decoding. We have one key medieval source for its tuning, Tractatus de Musica, written in Paris in c. 1280 by Dominican friar and music theorist, Jerome of Moravia, and another source, also Parisian, from before 1361, more of which below. Tractatus de Musica was written for inexperienced church musicians, for “the friars of our orders or of another”, to help them understand and perform ecclesiastical chant.
In the final short chapter he moved his attention to bowed strings and, for this reason, Jerome’s writing is very special for early music researchers, being the earliest surviving record of the construction and tuning of medieval instruments. He described the two-stringed “rubeba”, meaning either a rebab or rebec, stating that it is less important than the “viella”, the medieval fiddle. Jerome was moved to give instruction on the religious and secular use of the vielle, showing that in his day it was considered a suitable instrument for accompanying church music.
Jerome described the vielle as an instrument for singers to self-accompany; for heterophony, i.e. for playing simultaneously different versions of same melody so that, for example, at one point in the tune the voice may fall but the fiddle rise, creating a brief harmony, perhaps particularly at a point where that note is not available on the fiddle due to its tuning (as we’ll soon see); and biphonic music, i.e. a melody over a drone.
He wrote that the vielle “has, or should have, 5 strings.” The iconography shows that both 4 string and 5 string vielles, and some with 3 strings, were widespread. I therefore take Jerome’s “should have” to be his personal preference.
The first string on the 5 string fiddle is used as a bourdon in two of Jerome’s three tunings. Since there is no evidence of bourdon practice on the 3 or 4 string fiddle, perhaps Jerome considered this option to be integral to the instrument. His three tunings mark the vielle out as very different from other European bowed instruments, then or now, and the tunings he described help us make sense of the vielle’s flat bridge.
Before we turn to that, a quick note about the Guidonian hand, the theoretical range of available notes in medieval music, which will soon become important.
The Guidonian hand and medieval modes
Guido of Arezzo (Guido d’Arezzo, Guido da Arezzo or Guido Aretinus) was an 11th century Italian Benedictine monk seminal to the development of music, responsible for creating the first version of modern musical staff notation, which eventually supplanted neumatic notation. (For an explanation of neumes, see Bird on a briar: interpreting medieval notation, with a HIP harp arrangement.)
Guido saw the need for some musical education, as some monks found the melodies of Gregorian chants difficult to memorise. So he created the hexachord, a six note scale whose note names derive from the first syllable of the first six half-lines of the then-familiar hymn, Ut Queant Laxis, those syllables being ut, re, mi, fa, sol, la – which is, of course, the basis for the now-familiar eight note scale, do, re, mi, fa, so(l), la, ti, do.
The hexachord was later developed into the Guidonian hand, an ingenious chart of all available notes in his system, mapped onto the hand by associating each note with a joint, fingertip or the palm. You can see an entertaining video of the Guidonian hand in action by clicking on the picture below (opens in new window). Lovers of Sacred Harp music will notice a similarity.
A modern musician may notice something which appears odd about all the available notes in the medieval gamut: there are B naturals and B flats, but no other note is available either flattened or sharpened. This is not an omission, but integral to the way medieval modes worked.
All modern major scales work to the same musical principles, with identical gaps between the 8 notes of the scale, it’s just that they are pitched differently: so C major and D major sound the same, except that D major starts a tone higher, in this case making f into f# and c into c# to maintain the same pattern of tones and semitones in all major scales. It is not so with modes, which lacked sharps and flats. This means that the relationship between notes for a mode starting on D (dorian) is different to a mode starting on E (phrygian). In addition, each medieval mode has a returning note which plays a key role in the melody, this note known as the tenor, tuba, dominant, repercussa, or reciting note; and each mode has its own characteristic figures or melodic clusters of notes. Add to this the fact that some modes started and ended in the same place, known as authentic modes, and others started on one note and finished on another, known as plagal modes, and we see that the medieval conception of sound was unlike ours. And so the hypolydian mode, which starts on C, isn’t really like modern C major since it is a plagal mode, its finalis or final note being F rather than C, with characteristic melodic figures and a tenor or reciting note of A.
The gamut notes in this system were known as musica recta or musica vera, meaning right or true music. It is precisely in these terms that Jereme of Moravia described the notes on the vielle. This implies that any changed notes – flattened or sharpened from the gamut notes – were wrong, and this is just what the church taught: each mode had a particular sound and was associated with a particular mood, so changing notes would sully the nature of the mode or mood. We know that singers and musicians did change notes in performance (rather than in writing) because the church frowned upon it and called those altered notes musica ficta or musica falsa, meaning feigned or false music.
There was one exception, accepted by the church, for an audibly obvious reason: the strategic flattening of B when it appears against an F. This interval is called a tritone, i.e. three tones (or three ‘whole steps’). If you try playing or singing any two notes together that have an interval of a tritone, like B and F, you’ll notice that it’s has the sort of audio clash you might hear in old horror film music, so it’s no surprise that it gained the name, ‘the devil’s interval’. In any other case but this, musica ficta was frowned upon.
Of course, what the church taught and what people actually did are two different prospects and, as we’ll soon see, Jerome of Moravia’s three vielle tunings take account of this.
Tuning 1: d (bourdon drone), G, g, d’, d’
In Jerome’s day we are a long way historically from fixed or standardised pitch, so the important aspect of these tunings is the relationship of the notes to each other, rather than the modern idea of absolute pitch. What sets the theoretical relationship is the note G, the lowest note of these tunings and the lowest possible theoretical note in the gamut.
On modern bowed instruments, we tend to count strings from the highest to lowest sounding: Jerome works from the other side. Two of the tunings he describes are re-entrant, meaning they do not strictly follow the pitch order of high to low or, in this case, low to high. Since one string does not follow this low to high pattern, the next string has to re-enter the sequence, hence the name.
In the first tuning, the first string d is a bourdon, plucked or bowed open. The second string, tuned to G, can play only G, A, B, c, according to Jerome, thus assuming that the player will not move out of first position, which appears to be a universal assumption in this period. The third string is tuned to g, so plays the notes an octave above the second string: g, a, b and c’. This means the notes e and f are missing, supplied by the fourth and fifth strings at the octave above. The fourth and fifth strings are both tuned to d’ and so, Jerome states, they can play d’, e’, f’ and g’ and, with the application of the little finger, a’.
This begs two related questions: In what kind of music is it permissible to have notes missing in the sequence which can only be played instead at the upper octave? And what sort of music is intended with a bourdon that can only pluck or bow a drone? Jerome implies an answer when he states that “a vielle tuned in this fashion can play all the [church] modes”. So if, as Jerome implies, this was a tuning for church music, therefore accompanying voices singing in different octaves, then it doesn’t matter if two missing notes in the sequence are transposed up an octave. Whether this tuning was used in other contexts or exclusively in a church setting is impossible to know.
Tuning 2: (no bourdon) d, G, g, d’, g’
Strings 1, 2, 3 and 4 are tuned the same as in the first tuning, but string 1 is on the fingerboard, i.e. there is no bourdon. String 5 is now tuned up to g’, not in unison with the fourth string.
The second of Jerome’s vielle tunings is, he states, “necessary for secular and all other kinds of songs, especially irregular ones, which frequently wish to run through the whole hand.” A little decoding may be necessary: “irregular” songs are either those employing musica ficta not approved by the church, or else songs that do not follow the ecclesiastical schema of modes, hence “secular and all other kinds of songs”; and “the whole hand” is the Guidonian hand, the whole gamut, as described above.
In this tuning, therefore, there are no missing notes: all notes are available from low G in a range of two and a half octaves up to d’’ (so not quite the whole hand, as e’’ doesn’t seem to be possible).
There are two possible interpretations. This tuning, with its series of d and g notes in different octaves, is ideally suited to interweaving the melody between constantly droning strings. Additionally, the combination of the tuning, the flat bridge, and the low action could have been used to play parallel fifths or fourths on adjacent strings, a technique also used in church music, known as organum. On the other hand, the arrangement of pitches, meaning that all notes are available in first position to “run through the whole hand”, makes this tuning alone the ideal candidate for the curved bridge.
Tuning 3: G (bourdon), G, d, c’, c’
Jerome did not state that the first string in this tuning is a bourdon, but his contemporary Pierre de Limoges owned the only complete copy of the Tractatus, and added a note to this effect. Jerome didn’t state what kind of music this tuning is designed for and, if Pierre de Limoges knew, he didn’t write it down.
Played as 5 single strings, no matter how I think about this tuning I cannot make sense of it. On the surviving manuscript, a commentator has written in the margin, “I do not see how note b is sounded”, and neither can I. In the first tuning, I can see how strings 4 and 5 both tuned to d’ can create effective music, with, perhaps, the melody played on string 4 while string 5 continues the d’ drone (or perhaps vice versa); but I cannot see how two single strings tuned to c’ can work while other strings are tuned to G and d.
Christopher Page’s tuning solution
Christopher Page is one of our most important, meticulous and creative early music scholars, who also directs the early music group, Gothic Voices. In his Voices & Instruments of the Middle Ages he makes suggestions for vielle tuning which go beyond Jerome’s statements, and which he writes are on “uncertain ground”, but which do help make sense of Jerome’s intent more readily and logically: adjacent strings in octave leaps or in unison are two strings in a course. On this basis, he proposes:
Tuning 1 in 3 courses: d (bourdon), G-g, d’-d’
This makes an octave between the plucked/bowed bourdon and the highest-pitched bowed course and an octave integral to the inner course. Played as a block of strings, this would produce parallel moving octaves (octave organum) on the middle course which come and go, depending on whether this course is playing the tune or an open drone.
Tuning 2 in 4 courses: d, G-g, d’, g’
With the first string d freed from being a bourdon drone and the highest string now up to g’, this tuning has (almost) the whole gamut of two octaves and a sixth available (it has two octaves and a fifth) while, assuming a flat bridge, the non-stopped strings continue to play drones. All strings are single except for the octaves on the G-g course, so it retains an element of octave organum. On this octave course, the ear would automatically pick out the octave intended for the melody and hear the other octave as parallel organum (as is also true on the octave course of the renaissance cittern, in my experience).
Tuning 3 in 3 courses: G-G, d, c’-c’; or in 4 courses: G (bourdon), G, d, c’-c’
Christopher Page’s third suggestion does not mention, as Ephraim Segerman (2001) does, that Jerome’s contemporary Pierre de Limoges identified the first G as a bourdon.
It may appear that, with the bridge completely flat, with a constant “humming block” (as Christopher Page evocatively calls it), it doesn’t matter if the strings are arranged singly or in courses. For the bowing hand, this is true; but for the stopping or fretting hand, it is most important, as two strings placed together in unison or in octave courses need to be placed such that they can be stopped or fretted together.
Christopher Page’s suggestion of octave or unison courses makes much more sense than 5 single strings. It is not possible that a single instrument could accommodate all 3 tunings, even with a series of alternative slots in the bridge and grooves in the nut to account for the 3 different arrangements of single strings and courses, as no gut string can be tuned successfully to the tensions required for both G and d, or both g’ and d’. In this sense, vielle tunings are like baroque guitar tunings, in that there were three different ways of tuning that could only be effected by changing strings or having three different instruments.
Tuning 3 remains problematic. Christopher Page suggests that because some notes are missing in tuning 1, playable only at the octave above, then it is conceivable to have many more missing notes in tuning 3, and not available at all, with the melody played only on course c’-c’, and with G-G and d playing only drones. Realistically, this only gives the player a range of a fifth. He cites the only surviving example of a chanson de geste melody, Audigier, dit Raimberge, which has a range of only a fourth and could therefore be played in this tuning. He argues that if we cease to think of the vielle as a melodic instrument, but as an accompanying instrument, then the tuning could work in the suggested way. He may be right, but the circumstantial evidence of a single song doesn’t seem strong grounds on which to base a whole repertoire for a tuning with such a small range.
I’m going to suggest that there is a more viable and historically-attested solution for tuning 3, arrived at through comparison with crwth tuning and technique.
The implications of crwth technique and tuning
The European bowed crwth (or crowd) was a development that emerged in AD 900–1000 from the plucked lyre, which dates back to 2000 BC. The crwth flourished in western Britain from the 11th century, and in the rest of Europe was known as the chrotta, crotta, or rotta. Pockets of crwth players continued to play in Wales right through to the 19th century.
In 1801, William Bingley interviewed a crwth-player in Caernarfon, who gave a 3 course tuning of fifths doubled with octaves. In relative terms (the actual pitches may vary), this would be G-g, d-d’, a-a’. That tuning is clearly very late for our purposes. Jerome’s vielle tunings only had one interval of a fifth: this tuning, with two fifth intervals, is the same as rebec tuning but, given the date, it may have been influenced by violin tuning.
There is a tuning recorded slightly earlier, which is not possible on the same strings. British Library Add. MS 15020:91, dated 1768-9, with a copy in Wales now known as Aberystwyth MS 168:6, gives relative pitch instructions which are very clear. In relative terms, the strings are tuned: G-g, c’-c, d-d’. This tuning is accepted by researchers and players as traditional for the instrument.
I’m going to need to convince you. We are comparing two different instruments, and there is a gap of 500 years between Jerome of Moravia’s description of vielle tuning in c. 1280 and the British Library ms. description of crwth tuning in 1768-9.
In the 12th century we have the first evidence of a crwth with drone bourdon strings, 2 strings positioned across the bridge and then off the fingerboard, either plucked by the player’s left thumb or bowed. Bowing the bourdon is possible as the crwth has no pegbox and all strings and tuning pegs are on the same plane. Bourdon bowing and plucking was still used in 1768-9 and until the instrument’s demise in the 19th century. Since this practice began in the 12th century and became commonplace in the mid to late 14th century, we can speculate that either the 5 string vielle with a bourdon was influenced by the crwth, or that the crwth took on a feature of the vielle.
The available notes on the crwth in the tuning G-g (bourdon), c’-c, d-d’ are G, c, d, e, f, g, a, c’, d’, e’, f’, g’, a’. The anonymous commentator wrote in the margin of Jerome’s manuscript for his third vielle tuning, “I do not see how note b is sounded”. In crwth tuning, the same is true. That’s the first point of commonality.
The available notes in Jerome’s third vielle tuning G (bourdon) G, d, c’-c’ are G, A, B, c, d, e, f, g, a, c’, d’, e’, f’, g’. That is remarkably similar: only the addition of A and B on the vielle and the addition of a’ on the crwth are different.
To make my point clear, there follows two almost identical paragraphs on actual crwth technique and my imputed vielle technique in the third tuning, with nothing significant changed. The only changes, for ease of comparison, are in bold.
With the melody played on the crwth on the d-d’ course, we have a G-g c’-c drone. Thinking in modern terms, we could say this is nominally a C chord (major or minor, since it lacks a third). Playing the melody on the c’-c course gives a G-g d-d’ drone, so nominally a G chord (major or minor, since again it lacks a third). This means, in modern terms, we have chords I and V in C or chords I and IV in G readily available as drones. This is established crwth technique.
With the melody played on the vielle in the third tuning on string d, we have a G c’-c’ drone. Thinking in modern terms, we could say this is nominally a C chord (major or minor, since it lacks a third). Playing the melody on the c’-c’ course gives a G d drone, so nominally a G chord (major or minor, since again it lacks a third). This means, in modern terms, we have chords I and V in C or chords I and IV in G constantly available as drones.
The string spacing on the crwth is such that gaps between strings of a course are smaller than between courses, as one would expect, enough to identify each pair of strings. At the same time, the gaps between the two strings of a course are enough that a player can easily chose to stop one or the other octave of a course, or stop both together. Much of the iconography of the medieval fiddle appears to show equal string spacing: I suggest this is because double-strung courses had string spacing that looked equal to a non-player, or from only a short distance, probably so as to be able to split a course when the player required it, as on the crwth.
So both the 5 string vielle and the crwth employ a bourdon off the fingerboard; a flat bridge to effect a block of strings played as a drone; both use strings in courses tuned in octaves but with a wide string spacing within the course; and Jerome’s third vielle tuning is, for all practical purposes, identical to traditional crwth tuning. I suggest that there no significant differences in playing technique between the two instruments.
So what looked like a very unpromising vielle tuning has a playable range of two octaves, with only the note b missing, which is available at the lower octave if we work on the same principle as tuning 1, that missing notes in one octave can be played in another. In practice, this can clearly only be achieved if the whole phrase with b in it is played at the lower octave. In any case, the missing note b could simply give rise to a moment’s heterophony. If, for example, the fiddle is tracking the vocal melody for a tune that goes from c’ to b (for example), the player could simply have a moment’s harmony and play c’ to d’.
It is tempting to suggest that these two instruments are so closely related that it cannot be a coincidence. It may well be that the vielle was a development of the crwth, just as the crwth was a development of the lyre.
Certainty is impossible, but it is tempting to link the 3 styles of bridges to the 3 tunings. Could it be that tunings 1 and 3 required a flat bridge, while tuning 2 was played with a curved bridge? Or that all tunings could be played with a flat bridge, but only tuning 2 could he played with a curved bridge?
4 and 3 string fiddle tunings: some suggestions
Now let’s turn to another gap in the historical record, and see if we can reach viable and justifiable answers. Though evidence is lacking, it seems sensible to suggest, on the basis that 4 string fiddles did not have a bourdon, that they were tuned as 5 string fiddles but without the potential bourdon string, what Jerome calls the first string. This, I suggest, is the reason Jerome gives no tunings for 4 or 3 string fiddles, since they “should have” 5 strings, i.e. they “should have” the possibility of a bourdon.
In this case, then Jerome’s tuning 1 – d, G, g, d’, d’ – becomes G, g, d’, d’. It now becomes obvious that this tuning has the potential to be a 2 course rather than a 4 single string instrument, an octave course and a unison course, in line with both Christopher Page’s suggestions and crwth practice of doubling bowed strings in octaves. So it is most intriguing that there is a carving of a 4 string fiddle in Beverley Minster, Yorkshire, England, dated to c. 1335, suggesting exactly this arrangement, with a wider gap between the middle strings than between the outer strings, and with a flat bridge. It would also be possible to have this as a 3 string fiddle tuning in 2 courses, simply by not doubling the d’ string, i.e. G-g, d’.
Jerome’s tuning 2 – d, G, g, d’, g’ – lacks a bourdon, giving more potential options for which string to miss. If we work on the principle that the re-entrant first string is always the addition when moving from a 4 string to a 5 string vielle, then tuning 2 becomes a 3 course arrangement of an octave pair and 2 single strings: G-g, d’, g’. A 3 string fiddle tuning then suggests itself without the lower octave G, 3 single strings, g, d’, g’. This would work as a drone block with a flat bridge or as a single melodic line played in first position with a curved bridge.
Jerome’s tuning 3 – G, G, d, c’, c’ – also leads us to possibilities for both the 4 string and 3 string vielles: 4 strings would become G, d, c’-c’ in 3 courses, doubling the top c’; with the 3 string fiddle having a single c’.
If these suggestions are correct, then the removal of the first string removes re-entrant tuning on the 4 and 3 string vielle; and all 5 string tunings are available on 4 strings without losing any of the range of pitch, and most on 3 strings.
That is one solution. Another is to miss out the second string. It could be argued that this is a more logical step for the 4 string fiddle. When the lute, over the course of its evolution, gained a wider range of pitch, it was generally the bottom end that was added. The vielle appeared in the 11th century and Jerome was writing in the 13th, so if the same is true of the vielle as of the lute, then it could be that the 5 string instrument evolved from the 4 string when the low G was added.
Removing the low G would make 4 string tuning 1 into d, g, d’-d’ in 4 courses, with an option for the 3 string fiddle of a single d’. Tuning 2 would become d, g, d’, g’, of necessity 4 single strings. This is perfectly possible because, as we have seen, there was a range of practices for fiddle stringing.
On this method, tuning 3 would be the odd one out, as both low Gs would have to be removed, making a 4 string tuning impossible, and leaving only the 3 string tuning in 2 courses, d, c’-c’. However, this scenario would undermine the close relationship between the crwth and the vielle I have already argued for, which necessitates the low G. I therefore have to conclude that the first solution to 4 and 3 string fiddle tuning, the removal of the first (bourdon) string, is the preferred historically-implied option.
Of course, this is all guesswork. But the first solution for 4 and 3 string vielle tunings does retain the integrity of Jerome’s 3 tunings; the integrity of what the evidence suggests about vielle drone playing styles; and the implications of crwth playing style for the fiddle.
For the modern player who wishes to stay with 4 single strings rather than strings doubled into courses, a combination of Jerome’s essential open strings d and g is immediately user-friendly and retains the historical drone characteristic. On the basis that it is the relationship between pitches that is important rather than absolute pitch, and that players will have voice pitches and other accompanying instruments to match, fiddlers will make their own adaptations. Helen Wilding, fiddler with medieval trio Fleurs de Lys, utilises just such an adaptation for her melody-in-drone playing to suit bagpipes and hurdy gurdy: G, d, g, d’.
The Berkeley Theory manuscript tuning
Up to 80 years or so after Jerome of Moravia’s treatise, we have a quite different tuning in the Berkeley Theory Manuscript. The author is uncertain, but Christopher Page has presented strong circumstantial evidence for it being Johannes Vaillant, a 14th century Parisian music teacher, who died in 1361. This puzzlingly neglected source shows a tuning of c d g c’, a tone followed by two fourths, without a bourdon and without Jerome’s re-entrant tuning.
There is some doubt about the instrument’s identity. Christopher Page works on the assumption that it is a vielle and, as such, comments that the illustration is curious: a frontal string-holder replaces the usual tail-piece; and peg boxes, such as those on gitterns and citterns, are often decorated with carved heads or other ornamentation, but to have a carving on a peg board appears to be unique to this source (except the erroneously restored Beverley Minster vielle – see photo above).
It may be that the author was drawing a highly unusual individual instrument, but this seems unlikely since the purpose of his text was to describe common practice. More likely is that this isn’t a fiddle, but a citole. While the instrument is in one of the shapes used for the vielle, it is also in one of the shapes used for the citole and, if it is such, the decorated peg box is not unusual. The frontal string-holder is still problematic, appearing to be a tail-piece morphed into the shape of a fixed bridge. The artist may have been confused but, as we see below, though vielles and citoles often shared the same tail-piece string-holder arrangement, citoles did sometimes have other ways of affixing strings not seen on fiddles.
The Berkeley instrument is fretless. While fiddles were often fretless and citoles were always fretted, the lack of frets does not preclude this being a citole, since all of the Berkeley illustrations are incomplete, including an apparently fretless gittern on the same page and a harp with only 4 strings. The author drew just enough to make his point. Still, if this is a vielle, it is curious that the author did not draw a bow or mention one in the text.
The name given to the instrument in the text is cithara, which is where the problem arises. Music theorists of the medieval and renaissance periods loved to give instruments antique names, to link them to ancient practice. Thus the cithara, from the Greek kithára (κιθάρα), the Assyrian chetarah, and the Latin cithara – a lyre for which there is evidence dating back to 1700 B. C. – was the same name used indiscriminately by many writers for lyres, harps, psalteries and, indeed, it sometimes seems, for almost any instrument, open-stringed or fretted. The cithara in Berkeley is a case in point, a word by then more associated with the citole than the vielle. What muddies the waters is that ancient Greek comic playwright, Aristophanes, gave the first recorded version of the word for fiddle – but not to describe a fiddle. The Latin fidiculae (fidēs = chord + -culae = diminutive suffix) in his play of 405 B. C., Ranae – The Frogs, described a kind of cithara or lyre. The Latin fidiculae gave rise to associated words for both bowed fiddles and plucked vihuelas: the German fidula and videle, the French vielle and viole, the Italian viola and violino, the Spanish viguela and vihuela, the English fithele and fidel, and so on.
Certainty is impossible, but my conclusion is that the Berkeley cithara is a citole. If so, this means we don’t have an additional tuning for the vielle, but we do have the only surviving tuning for the citole.
Were there other vielle tunings?: a lesson from the wheel fiddle
Evidence is lacking about fiddle tunings from its early life in the 11th century. Did vielle tunings change over time? Since there were 3 known tunings only in France in c. 1280, we can legitimately ask if they tuned the same contemporaneously in France, Spain, Italy, Germany, and England.
I suggest a general answer can be reached with a degree of certainty though, without corroborating evidence, I cannot be definitive. The structure and tuning of the flat-bridge medieval fiddle suggests that it should not be thought of in the same musical category or soundworld as later bowed curved bridge instruments such as the viola da gamba and violin, which function as melody alone or melody with simple harmony. Rather, the vielle should be associated with the family of instruments which function as a melody played over or within a continuous drone, as with the crwth.
Characterised in this way, there is a group of instruments which, like the vielle, was used in church music: the family which evolved from organistrum to simfony to vielle à roue (hurdy gurdy). From the 10th century we have evidence of the organistrum, named after the church vocal style of organum. The organistrum was a large drone instrument played by two people, one to turn the crank for the rosined wheel which acted like a continuous bow against the two drone strings and single melody string, and one to play the tangents (keys) for the melody string over the drones. By the 13th century, a smaller, one person version had been developed, the simfony. By c. 1500 the simfony had evolved further into the vielle à roue, literally the wheel fiddle, with its rhythmically buzzing bridge (which became known in later times as the hurdy gurdy – from 1749, according to the Oxford English Dictionary).
I suggest that this naming is significant. Indeed, it seems sensible to assume that the wheel fiddle, the vielle à roue, gained its name, not just from the wheel acting like a continuous fiddle bow, but also that, and even primarily that, the drone playing effected the same soundworld as the vielle. Certainly Jerome’s 13th century drone tunings are in a style of music that was fundamental in 12th and 13th century sacred music, and the emergence of the vielle à roue in c. 1500 suggests that this style was still played on the vielle then. So, while we can’t say for certain that the exact tunings Jerome gives were used since its emergence (from the crwth?), I suggest at the very least that the drone principles upon which his fiddle tunings were based remained a constant feature.
The renaissance vielle
The tunings explored so far cover most of the life of the vielle, but it is not the complete story.
In the 15th century, at the beginning of the renaissance, musical norms were rapidly changing. The vielle was still being played, but was past its dominance of musical culture. The discarding of the quill by lute players, spreading gradually from the 1470s to the 1490s, meant that a single player could produce solo polyphony with individually moving fingers, and this was the dawn of the lute’s dominance of renaissance music. The musical shift that led to this advance led to divergence of practice among fiddlers and some structural changes to the instrument. While the vielle played by the angel in green from the Immaculate Conception Altarpiece (pictured above), painted 1490–99, shows that old practices continued, others innovated, increasing the height of the bridge and normalising the arc such that each string could be individually bowed, thus enabling these fiddlers to participate in polyphony with other instruments without having to play drones.
To answer this question we have to turn to the only renaissance witness in the matter, Johannes Tinctoris (c. 1435–1511), a Belgian composer and music theorist who lived most of his life in Italy from c. 1472. Tinctoris stated that the 3 string fiddle was tuned in fifths, which is the same tuning as the rebec. (We could conjecture that if there was continued interchange between the vielle and the crwth, then it may be that the crwth tuning in fifths given to William Bingley in 1801 was not the result of influence from the violin, but the result of these earlier 15th century changes.)
The 5 string fiddle, stated Tinctoris, had an arced bridge to facilitate the bowing of individual strings, and was tuned “unevenly in fifths and unisons”. What this means precisely he didn’t state, but Jerome’s first tuning can be seen to fit this description – d (bourdon), G-g, d’-d’ – as it has “fifths and unisons”, and the re-entrant tuning and the G-g octave course makes the pitch distribution “uneven”. So it may have been that this tuning continued to the 15th century, except now without the ability to play a drone block with all strings at once, as musical styles had changed and the bridge along with them.
What of the 4 string fiddle in the early renaissance? Ephraim Segerman states that the “only historical evidence on the tuning of 4 string fiddles is the occasional 15th century picture that shows a string thickness sequence of thickest, thinner, somewhat thicker and thinnest, seeming to imply two octave pairs a fifth apart.” If these depictions of string thicknesses are an accurate reflection of practice – and that’s a crucial and unanswerable question – then this 15th century arrangement doesn’t precisely fit any of my 4 string suggestions above, but it is very close to one, and we are back to my 4 string modification of Jerome’s first tuning: G-g, d’-d’. To comply with Ephraim Segerman’s observation, all we need do is make the top course octaves rather than unisons: G-g, d-d’, which reflects crwth practice of courses in octave pairs.
The end of the vielle and the birth of a successor
The challenges to the vielle from the changing musical styles of the renaissance spelled its gradual decline, but not before it had inspired the birth of another instrument towards the end of the 15th century, whose lifespan overlapped the demise of the vielle in the middle of the 16th century. The lira (or lyra) da braccio (meaning bowed instrument on the arm) was to be short lived, appearing in paintings, mostly Italian, from c. 1490 to the early 17th century, leaving behind only a short section of music added in 1540–45 to the Italian Pesaro lute manuscript, c. 1490–1511.
The lira da braccio music in the Pesaro manuscript is for 6 strings in 4 courses – D-d (bourdon), G-g, d, a – but other witnesses give 7 strings in 5 courses: Giovanni Maria Lanfranco, Scintille di musica, Italy, 1533, has the same tuning but with an additional e’; and Michael Praetorius, Syntagma Musicum III: Termini musici, Germany, 1619, reproduces this but with a top string of d’.
Since the lira da braccio was derived from the vielle, this arrangement in courses which includes unison and octave stringing is another justification of reading Jerome’s tunings this way. The lira da braccio shared many features of the vielle, but was different enough to accommodate renaissance music: its neck was usually unfretted but occasionally fretted; it had 6 or 7 strings, arranged in a combination of octave courses and single strings; with 2 strings as a bourdon in an octave course; it had a wide fingerboard and a bridge with a gentle enough arc to play triple and quadruple stops (3 or 4 strings at once) chordally, rather than as a drone, due its tuning.
So we see how the lira da braccio succeeded, for a brief time, in overcoming the melody-in-a-drone strength of the vielle in the medieval period which had become its outmoded weakness in the renaissance, with the typical flat-bridged humming block replaced by the gently curved bridge tuned in fifths, enabling melodic and a degree of harmonic playing.
This difference is illustrated in the final two videos: Igor Pomykalo demonstrating first the vielle, then the lira da braccio.
References (with links to websites):
Bevil, J. Marshall (1973) The Welsh Crwth, Its History, and Its Genealogy.
Brown, Howard Mayer (1989) ‘The trecento fiddle and its bridges’, in Early Music, vol. 17, no. 3, August 1989.
Green, Corey (2012) A Player’s Introductory Guide to the Medieval Vielle, a thesis in musicology submitted to Texas Tech University.
Greenhill, Peter (2000) The Robert ap Huw Manuscript: An Exploration of its Possible Solutions – 3 tuning.
Jones, Sterling Scott (1995) The Lira Da Braccio (Indiana: Indiana University Press).
Meurig, Cass (2009) Talharpa 2009 DVD, Walesi Crwthi tutvustus, on the ViljandiFolk YouTube channel.
Page, Christopher (1980) Fourteenth-Century Instruments and Tunings: A Treatise by Jean Vaillant? (Berkeley, MS 744) in The Galpin Society Journal 33, March 1980.
Page, Christopher (1987) Voices & Instruments of the Middle Ages. Instrumental practice and songs in France 1100-1300 (London: J. M. Dent).
Palmer, Frances (1983) ‘Musical instruments from the Mary Rose: A report on work in progress’, in Early Music, vol. 11, no. 1, January 1983. (With thanks to Peter Forrester for showing me this article.)
Pearce, Barry (2016) Bowed Strings Iconography Project (online)
Poplawska, Dorota & Czechak, Tadeuz (2002) ‘The Tuning and Playing of a Medieval Gittern and Fiddle from Elblag, Poland’, in The Consort, vol. 58, Summer 2002. The Dolmetsch Foundation.
Segerman, Ephraim (2001) ‘Tuning and Stringing Medieval Fiddles’, in FoMRHI Quarterly, No. 105, October 2001. Fellowship of Makers and Researchers of Historical Instruments.
Skeaping, Joe (2001) ‘Jerome of Moravia and the Tractatus de musica’, in FoMRHI Quarterly, No. 105, October 2001. Fellowship of Makers and Researchers of Historical Instruments.
Smith, Douglas Alton (2002) A History of the Lute from Antiquity to the Renaissance (The Lute Society of America).
Springfels, Mary (2000) ‘The Vielle After 1300’, in Ross W. Duffin (editor), A Performer’s Guide to Medieval Music (Bloomington: Indiana University Press).
van der Werf, Hendrik (1988) ‘The “Not-so-precisely Measured” Music of the Middle Ages’, in Performance Practice Review, Vol. 1, No. 1, Article 5.