The bray harp is not a sound modern ears are used to, and even most early music groups with harps don’t use the period instrument, yet it was the standard European harp of the late medieval, renaissance and early baroque periods, from the 15th century to the 1630s, and still used until the late 18th century in some places. The bray harp gets its name from the L shaped wooden pins at the base of the strings, positioned so the strings buzz against them as they vibrate: the effect was said to sound like a donkey’s bray. It’s an older idea than the bray harp, one shared by the oldest surviving stringed instruments, made four and a half thousand years ago.
This article traces the earliest evidence for the European harp, leading to the origins and popular rise of the bray harp specifically. We ask why there is so little surviving early harp music and try to come to a feasible historical answer; and along the way take in Pictish stones, illuminated Psalters, Geoffrey Chaucer, harp-playing angels and an ape playing a rota, with a video of the harp being brayed.
Harp and bray origins
The harp is descended from a family of musical instruments with unfretted strings stretched across a frame, including the lyre of classical Greece, evidenced from the 14th century BC. The earliest known surviving stringed instruments were even earlier than the Greek lyres. In 1929, British archaeologist Leonard Woolley excavated instruments from royal tombs in Ur, southern Iraq, which have a direct relationship to the bray harp in their concept and execution of sound. Amongst the artefacts made of gold, silver and alabaster, Woolley found three lyres. They look like harps but are classed as lyres on the distinction that harp strings rise directly from the soundboard, whereas lyre strings pass over a bridge, as do the strings of the Ur lyres. They are dated 2600–2400 BC, decorated with golden bull’s heads and not coincidentally described in contemporaneous texts as sounding like a “softly lowing bull”. This sound is created by the strings vibrating against the top edge of the bridge. This is the same principle as bray pins and essentially the same sound as the bray harp. A lyre or a harp with some kind of buzzing device, be it a bridge or a wooden bray pin for the strings to vibrate against, resembles a “softly lowing bull” with a large instrument at a low pitch, or a braying donkey with a smaller instrument at a higher pitch.
The same principle applies on the begena or bèguèna (seen in three photographs below), an Ethiopian 10 string lyre used exclusively for sacred functions in the Ethiopian Orthodox Church. The begena has pieces of leather lodged between each string and the bridge to create buzzing, which audibly imitates the sound of Ur’s “softly lowing bull”. According to Ethiopian tradition, this was the Biblical King David’s instrument, which he played to calm Saul: “So it came about whenever the evil spirit from God came to Saul, David would take the harp [some translations have lyre] and play it with his hand; and Saul would be refreshed and be well, and the evil spirit would depart from him” (1 Samuel 16:23). Tradition has it that Menelik I, son of King Solomon of Israel and of Makeda, Queen of Sheba (in modern Ethiopia), the first Emperor of Ethiopia ruling from 950 BC, brought the begena from Israel to Ethiopia. King David’s life is usually dated from 1040 to 970 BC so, if this story is true, it takes the origin of the buzzing string principle back considerably less far than the lyres of Ur. There is, however, no firm evidence of the begena in Ethiopia before the 15th century.
The rise of the European harp and the addition of bray pins
The first unambiguous illustration of a European triangular harp comes from the culture of the Picts, the Scots of the late iron age through to early middle ages. Twelve decorated Pictish stones have survived, dating from AD 8th or 9th century, all with Christian symbols, all depicting harps, some harps with players, some freely floating. These stone carvings are not detailed enough to give us any clues about construction, but their shape certainly indicates they are harps rather than lyres.
We have a more detailed view in the 10th century, in the Caedmon or Junius manuscript of 930, and this harp is without brays. The harp illustrations in the Hunterian or York Psalter of 1170 clearly show round string holders rather than rectangular brays. We see something open to interpretation on the harp in the Peterborough Psalter of 1310–20. The shapes on the soundboard look flat, more like the brass ‘shoes’ of the string band on the soundboard of a clarsach, the wire-strung harp of the Irish and of the Scottish Highland Gaels. Brass ‘shoes’, however, are semi-circular, so the shapes are probably round string holders flattened by the artist. Bray pins on European harps are not mentioned in writing nor clearly seen in iconography until the early 15th century, at which point they become ubiquitous. Both bray pins and round string holders perform the function of holding the string in place; in addition, the L shape of the bray pin is turned so that, when played, the string vibrates against it.
Iconography and written sources in the 15th century soon show the bray harp to be the standard harp (except in the above mentioned regions of the clarsach), remaining so until the 17th century. Harps of the late medieval period came in a range of sizes, from 60 centimetre lap harps with only 10 or 11 strings to 120 centimetre harps with 25 or 26 strings. Their distinctive rising, sleek shape, reminiscent of contemporaneous Gothic architecture, has earned them the retrospective name ‘Gothic harps’, though at the time, of course, they were just called harps. As can be seen above, not all Gothic harps were bray harps until we reach the 15th century.
The popularity of the harp
Harps were clearly very popular, being mentioned, for example, several times in Geoffrey Chaucer’s Canterbury Tales, written between 1387 and 1400, just before the advent of brays, as an instrument to sing with, associated with the gittern, lute and psaltery, and with taverns. It was often painted on its own, with other instruments, or duetting with a lute in the 14th and 15th centuries, often in a religious setting and played by an angel.
The sound of the bray harp from c. 1400 evidently had a wider influence.
The beautifully decorated Capirola lute book, written in Venice 1515-1520, states that a player should “make it so that the first fret almost touches the strings, and so on to the end, because as the frets are nearer to the string, the strings sound like a harp, and the lute appears better.” In other words, lute strings should buzz against the frets. The lute began to be fretted in around 1400, the same time that brays began to appear on harps, so I wonder if this was more than coincidental, if the sound of frets on lutes was the same as brays on harps, that they shared the same braying frisson for at least the first century or so of the fretted lute, perhaps for much longer.
In The Garden of Earthly Delights, painted between 1490 and 1510 by Dutch artist Hieronymus Bosch, we see the first visual evidence of the evolution of the simfony into the vielle à rue or wheel fiddle, known now as the hurdy gurdy. The key change was simple but revolutionary. A drone string, called the trompette (trumpet), was given its own little bridge, the chien (dog), which produces a rhythmic percussive buzz, the persistence and rhythm of which depends on the speed and pressure with which the player turns the crank. The vielle à rue was later to develop further but, at this point in time, the rhythmic buzz was the feature which distinguished it from its simfony forebear, putting it in the same soundworld as the bray harp.
German composer and music theorist Sebastian Virdung, in his Musica getutscht, 1511, describes a new keyboard, “just like the virginals, except that it has different strings [made with gut rather than the usual wire] and nails which make it harp” or buzz. This is likely to be the arpichordo mentioned by Italian writers of the 16th and early 17th century, a cross between a harpsichord and a bray harp. The Flemish and northern European renaissance keyboard, the muselar, was often equipped with a moveable baton, called a harpichordium or arpichordium, which had the same effect. The baton was fitted with brass hooks which, when the baton was ‘switched on’, made contact with the bass strings. You can hear the effect in this video (click on this blue text), and judge for yourself whether Dutch music theorist Quirinus van Blankenburg was right when he commented in 1739 that muselars “grunt in the bass like young pigs”.
The bray harp continued to be played through the renaissance and, when Michael Praetorius’ second and third volumes of Syntagma Musicum were published in 1618-1619, in the early baroque period, he wrote that “the strings of the harp rattle and crackle” and still described the bray harp as “the ordinary harp”.
The puzzle of harp music
We have a great deal of iconography testifying to the place of the harp in European culture, and yet we have little surviving harp music. In the medieval period, this is to be expected, as music was written either for the voice or for an instrument generally rather than one instrument specifically. This indicates the freedom of musicians to arrange their own instrumental parts and, if they could read music, they used the notation only as a starting point and improvised the rest. From relatively near the beginning of the renaissance onwards, different systems of tablature notation were developed for different specific instruments. Staff notation only shows the notes and the rhythm, and the instrumentalist works out the mechanics of producing those pitches; whereas tablature is tied directly to the instrument, showing the player the rhythm and the exact placement of both hands for all notes.
From the renaissance to the baroque period there is a wealth of tablature for the lute, cittern, and viola da gamba, and enough to keep us fairly happy for the bandora and 4 course guitar (for the latter there was ample but much of it is lost). For the harp, we have very little and it is all from the late renaissance onwards. We know there was a harp tablature system in Spain, for example, in Hernando de Cabezón’s Obras para tecla, arpa y vihuela (Works for keyboard, harp and vihuela), 1578, well within the heyday of the bray harp; then 60 years later in France, for example, in Antoine Parran’s Traité de la musique théorique et pratique contenant les préceptes de la composition (Treaty of theoretical and practical music containing the precepts of composition), 1639.
In Britain, only three sources of early harp repertoire survive, one English and two Welsh.
The first source is the George Cely Papers. In the early 1470s, George Cely joined his father’s trade to become a merchant dealing in raw wool. In the large volume of documents he left behind, dated 1475-1488, we discover that he had both bray harp lessons and dance lessons from Thomas Rede in Calais, where he spent a great deal of time on business. The bray harp is specifically mentioned, as is help in a lesson with setting the brays. There is no harp music in his papers; but his references to the music he learned is a valuable insight into the English harp repertoire back to the 1440s and ‘50s, since neither Rede nor Cely felt the need to keep up with the very latest music. George was learning unspecified dance tunes and, amongst his songs, Myn hertis lust and O Rosa bella by John Bedyngham, whose work also appears in continental manuscripts, Tos juirs and Go heart hurt with adversity, both anonymous, and O freshest flower, now lost. The work of English composers predominates: this was their heyday at home and on the continent, as John Dunstaple had effectively created the new sound of the renaissance, known as the English countenance, and thus paved the way for other English composers to be heard. (You’ll find a little more about this in Music of the renaissance: a whistle-stop tour.)
One of the Welsh sources is too corrupt to be of value. BM Add MS 14970 is an inaccurate early 19th century copy of an earlier manuscript, probably from the turn of the 17th century and now lost.
The second Welsh source is the Robert ap Huw manuscript, named after its writer, a harper from Anglesey, who calls for the bray harp to be played with fingernails. Though it was compiled in c. 1613, it is very much a retrospective: all the compositions noted date from 1340–1500, a remarkable stretch backwards in time. The reason for the book’s anachronism may also be why so little harp music appears to have been written down: harping was an expressly aural tradition, certainly in Wales, and probably as a feature of harping generally. The ordinances for the annual Welsh festival of literature, music and performance, the eisteddfodau, were drawn up in 1523 in the Statute of Gruffudd ap Cynan. This sets out the repertoire that harpists and crwth (bowed lyre) players were required to memorise for their bardic degrees, examinations in the Welsh traditional aural arts. Memory and improvisation were key: there is no reference to written music. It appears, then, that the medieval aural and improvising musical tradition, which became a written tradition during the renaissance, continued to be predominantly aural for the harp.
This aural and improvising tradition would explain three things: the ap Huw manuscript contents, its unique system of tablature, and the musical forms it contains.
The contents of the Robert ap Huw manuscript, stretching back 270 years or more, seem to be born of a fear that the old traditions would die out if they were not written down. Its harp tablature combines elements of lute, harp and keyboard tablatures of the 15th and 16th centuries from Germany, Spain and Italy, arriving at a unique system. The most plausible explanation is that there simply wasn’t a Welsh system of writing harp music, nor was there one over the borders in England, Scotland or Ireland to borrow, nor one overseas that could easily be adopted for the Welsh style of playing, so one had to be invented. As such, the tablature remained a cryptic puzzle to harpists for many years until it was finally decoded. The contents are based around 24 musical patterns, chord progressions or grounds known as “the 24 measures of string music”. The ground or ground bass, a repeating pattern upon which a melody is fashioned, is the basis for this improvising tradition. An intriguing feature of the names given these measures is that many of them don’t make any sense in Welsh, but appear to be transliterations of Irish names, suggesting a cultural link between the bardic traditions of Wales and Ireland.
Tuning: scordatura and fretting
The late medieval and renaissance bray harp was tuned and strung diatonically, that is, with only natural notes, not sharps or flats (like having only the white notes on a piano), though there are records of harps having both a B natural and a B flat string, as both were often needed for medieval and renaissance modes. Being diatonic, this meant sometimes employing scordatura, alternative ways of tuning the harp to suit the pitch of a voice or to employ musica ficta, adding sharps or flats to modify a mode.
If an individual note was needed both natural and sharpened or both flattened and natural in the same piece, the player would ‘fret’ the string. This involves reaching up with one hand to the part of the string that lies vertically across the wood below the tuning pin, and pressing the string against the wood, thus effectively shortening the vibrating string length and raising the note by a semi-tone. This is a harp technique documented independently by Spanish vihuela composers Alonso Mudarra and Juan Bermudo in the mid 16th century. Bermudo also documents a different solution to needing the same note natural and sharp or natural and flat in the same piece: tune it sharp or flat in one octave but not the other. The effect of this technique can be heard in practice in Mudarra’s wonderful imitative piece for vihuela, Fantasía X que contrahaze la harpa en la manera de Ludovico, ‘Fantasía 10 that imitates the harp in the style of Ludovico’ (Ludovico being the royal harper to King Ferdinand II) where, in the final passage, we have lower f played against higher f#’. (This piece can be heard on vihuela by clicking on this text).
The slow eclipse of the bray harp
Developments in European music in the 15th century and beyond, just on the cusp of the renaissance, were to mark the end of the diatonic simfony and psaltery: the increased use of notes sharpened or flattened in some parts of the music but natural in others, together with the rise of chord sequences and chordal playing, made those instruments obsolete. The diatonic bray harp, though, remained popular, perhaps because it had the option of both retuning and fretting. Still, being diatonic was increasingly seen as a problem for which a solution had to be found. Chromaticism was achieved by adding an extra row or two rows of strings. This was not a new idea, but it was applied in a new way.
In the late 14th century, the arpa doble or two row harp had been developed in Spain, as seen in the triptych in the Monasterio de Piedra, c. 1390 (right). The distance between doubled strings indicates that they are two distinct rows rather than doubled courses, but this does not represent an early attempt at chromaticism since there are an equal number of strings on each side, indicating unison stringing rather than a diatonic row and a row of accidentals. The artist has made a very odd mistake, with the tuners apparently sitting on top of the harp arm rather than fixed into the side. While a fundamental mistake like this may call into question the accuracy of the depiction of equal stringing on both sides, in c. 1390 the necessity of available sharps and flats was not the issue it was later to become, so it seems sensible to conclude that the arpa doble was unconnected with the later search for a chromatic harp.
The idea of double stringing a plucked, unfretted instrument was in any case not new in the late 14th century. The triangular instrument known as the rota, rote, rothe, or rotta is evidenced from the late 11th century. Strung with gut, it had two parallel rows of strings, as seen on the stone carving of an ape playing a rota on the outside of the 12th century church, Notre-Dame de l’Assomption, in Surgères, France (below). Petrus de Abano, in his Expositio problematum Aristotelis, written in Paris and Padua and completed in 1310, stated that the rota had 44 strings in all, 22 on each side, and that this arrangement of double strings was archaic. As with the arpa doble, the rota had two rows of unison strings. In this period, when polyphonic voices were close in pitch, sometimes crossing over each other or singing/playing unison notes, the double string band of the rota and the later arpa doble would have provided the perfect instrument on which one person could play polyphony with voices sometimes crossing or in unison.
The later use of double or triple rows of strings to create a chromatic instrument would both change the status of the diatonic bray harp and change the harp itself.
Juan Bermudo, in his Declaración de Instrumentos musicales, Spain, 1555, proposed an additional 8 strings on the same plane as the diatonic row for the most needed accidentals, differentiated by colour, further stating that some players were adding between 15 and 19 extra strings for accidentals. The idea was short-lived and appears not to have spread.
According to Tetrachordum musices by the German writer Johann Dobneck, known as Johannes Cochlaeus, some kind of triple harp, with 3 rows of strings, was in use in England when he wrote in 1511, presumably with brays, at this point. Chromaticism was achieved by having two outer rows of diatonic notes, with the player reaching inwards to the middle row of accidentals.
There are competing claims for the precise origin of the chromatic triple harp. Despite Johannes Cochlaeus’ writing in 1511, the invention of the (brayless) triple harp is often attributed to Italy just before 1600; and yet the triple harp was soon known as the Welsh harp, and is to this day. The truth is probably that musicians at different times and in several places were looking for a way to make the harp meet the new demands of evolving music, and therefore the triple harp was an idea whose time had come.
The change from diatonic bray to chromatic brayless harp came slowly and in patches. It took until the 1630s for the triple to become the new standard harp. The change was not universal and this was by no means the end for brays. In Harmonie universelle, 1636, Marin Mersenne claimed that brays were out of fashion in France, but in c. 1680-1700, when opera houses were gaining popularity in the German-speaking world, a two row chromatic Davidsharffe or David’s harp was developed, with brays on all strings. Brays have the effect of increasing volume, so the large Davidsharffe can be heard even as part of a full baroque orchestra. German maker Johann Volckmann Rabe (c. 1700–1750) made such instruments, six of which have survived. The Davidsharffe was probably the intended instrument for German-born English composer George Frideric Handel’s (1685–1759) harp solos in opera and oratorio, since early baroque German operas, cantatas and dance pieces were played on it; and it was the instrument played by Francesco Petrini (1744-1819), harpist to the Prussian court of Frederick the Great (ruled 1740-1786). This means that into the late 18th century a form of bray harp was played in Germany, was very likely played in England, and was part of Prussian life, which included parts of present-day Germany, Poland, Russia, Lithuania, Denmark, Belgium and the Czech Republic.
Buzzing brays yesterday and today
Today the buzzing sound of the bray harp is little heard. Now it is sometimes viewed as an unwelcome abberation even by some early music performers and players of early harps, but the musical aesthetic that eschews brays is clearly out of step with their historical use. Buzzing brays were no abberation: they were an integral part of late medieval, renaissance and baroque music. The audible frisson was present in west Asian lyres from the 26th century BC; in Ethiopian sacred music from at least the 15th century AD; in European harps from 1400, still described as “the ordinary harp” in 1618, and common until the 18th century in some quarters; on the vielle à roue (hurdy gurdy), with its rhythmically buzzing bridge from c. 1500; and in some keyboards from the early 16th century. If early musicians are to attempt to recreate the sounds of early instruments as faithfully as possible, then we need to include these good vibrations.
With thanks to Peter Wilson for information on the muselar, harpichordium and bray harps beyond 1660.