Category: Music

Reconstructing a ‘lost’ medieval instrument.

Music has been one of the great loves of my life since early childhood. Another passion is history and archaeology, especially as applied to the early British Isles. Sometimes the two combine, as when I discovered a type of lyre called a chrotta had been played across much of Europe for about a thousand years, from around the 9^th century BCE through to the early Middle Ages. A few years ago, I finally got my hands on a reconstruction of one, made by Koth na Fiach of Dark Age Crafts, and could hear what it sounded like and begin to work out how it might have been played. It is one thing to read about these instruments, quite another to actually handle one, play it and hear the sounds it produces.

In medieval texts, I came across another ‘lost’ instrument, the tiompán. One of the earliest writers to reference it is Giraldus Cambrensis (‘Gerald of Wales’), who, in his Topographia Hibernica (‘Topography of Ireland’), chapter XI, circa 1087, says that, “Scotland and Wales, the latter by propagation, the former by interchange and a pleasant affinity, strive to emulate Ireland in its musical modulation and to imitate its discipline. Ireland uses and delights in two instruments; namely the cithara, and the tympanum. Scotland uses three; the cithara, tympanum and chorus. Wales, in truth, the cithara, tibia and chorus. They use strings of brass, not gut. Many believe that Scotland today not only surpasses its teacher, Ireland, but, in musical expertise, far exceeds and outstrips it. So those seeking the source of the art now look to it.”

A Welsh Triad, recorded in the early 14th century manuscript known as Peniarth 20, also refers to the instrument:

Teir prifgerd tant ysyd, nyt amgen: kerd grwth, kerd delyn, a cherd timpan.

‘There are three chief crafts of the string, namely: the craft of the crwth, the craft of the harp, and the craft of the timpán.’

In Ireland, a poem from ‘The Siege of Dromdamhghaire,’ recorded in the 15^th century Book of Lismore, describes the appearance in the Brugh na Boinne (i.e. the Newgrange tomb-shrine) of the god, Aengus Mac ind Oc, to Cormac Mac Airt, King of Leinster, as follows:

“A silver tiompán in his hand, of red gold the strings of that tiompán;
Sweeter than every music under heaven
Was the sound of the strings of that tiompán.”

References to the tiompán occur in manuscripts from the 8^th century through to the 15^th, despite which there is considerable disagreement as to what it was. Several manuscript sources refer to it having a wooden body, possibly of Willow, and three strings, made of bronze, brass, gold or silver. This clearly rules out the suggestion of Irish harpist, Derek Bell, that it was a hammered dulcimer, since they have many more strings. Others assume it to have been similar to the Finnish Jouhikko, a two or three-stringed bowed lyre once common in Northern Scandinavia. A similar instrument, the Gue, was formerly played in the Shetland Isles, presumably having been introduced by Viking settlers. While the tiompán may have been a bowed lyre of this type, there are reasons to believe otherwise. For one thing, an instrument with no frets and only three strings obviously has a fairly limited range, although the melody string is shortened to produce different notes by ‘fretting’ it with the backs of the fingers. Manuscript references suggest the tiompán capable of considerable tonal range and subtlety of expression. The Irish cruitt, a word that covers both the early, 9-stringed lyre and the later, 25-or-more-stringed harp, and the tiompán were the only instruments deemed capable of playing the 'Three Noble Strains,' or modes, that were the crowning attainment of the musician’s art; goltraighe or weeping mode; geantraighe or laughing mode; and suantraighe or sleeping mode. The 'traighe' element derives from trai, meaning ‘a foot, or measure.’ In modern musical terminology, a 'measure' means everything that appears on a musical stave between two bar lines, including indications of the key, rhythm, tempo and notes to be played. Perhaps trai had a similar meaning. Whether we regard these strains/modes as keys, tunings, melodic structures or playing styles, however, achieving them on a small bowed instrument with only three unfretted strings seems like a tall order.

Then there is the name, tiompán. The letter ‘p’ being unknown in Old Irish suggests that it is a Latin loan word. Its nearest Latin equivalent is tympanum, though this applied in the Graeco-Roman world to a small, circular, hand-held frame drum, like a tambourine. This is the name given by Gerald of Wales, who wrote in Latin, although he plainly states that he is referring to a stringed instrument, not a drum. The similarity of names does, however, suggest that the tiompán may have had a round body, or soundbox, topped by a soundboard made of animal hide, presumably with a neck projecting from the body. This would put the tiompán in the category of long-necked lutes, a class of instrument still found in many Eurasian cultures, from Eastern Europe to Japan. Many have three strings. Examples include the Tuvan doshpuluur, the Central Asian rawap, the Chinese sanxian, the Japanese sanshin, the Siberian topshuur and the tungana of Nepal. Long-necked lutes similar to these have existed since at least 3100 BCE, when pictorial representations of them appear in Sumeria. It is hardly a stretch of the imagination to conceive of the idea reaching Britain and Ireland by the time the tiompán is first recorded around four thousand years later.

The sound and playing style of the tiompán are described in a probably 12^th century Irish manuscript that refers to the Battle of Magh Rath (637 CE). On the eve of the battle, music is played to bring sleep to the Ulster prince, Congal Claen: “And after that Congal slept to the quiet sound of the musical bagpipes and the prophetic ominous truly-sad shadows of the strings and tiompans being touched by the fronts, sides, tips and nails of the performers who played so well on them.”

This description of the sound of the tiompán beautifully evokes the emotional power of the instrument. A playing style that uses the “fronts, sides, tips and nails” of the fingers closely parallels the technique used in North Africa on the lute-like instrument variously known as the guembri, lotar or sintir, where the strings, of which there are normally three, are plucked or strummed with the right hand, the fingers of which are also used to beat out a rhythm on the animal skin soundboard. The body, or soundbox, of the guembri is usually roughly rectangular, being carved from a single block of wood. A similar instrument found in West Africa, the akonting, has a circular soundbox, also covered with skin, although some modern versions use timber. Both guembri and akonting have strings of animal gut. Other long-necked lutes, such as the Persian setar or Turkish saz, are wire-strung as was the tiompán according to manuscript sources. My suggestion, then, is that the tiompán was one of this extremely widespread and long-lived family of long-necked lutes, having a circular wooden body, or soundbox, covered with animal skin, and three bronze or brass strings, unless the player or a patron could afford silver or gold.

The addition of frets makes the location of notes far easier for the player and, since frets have been added to lutes since at least the Sumerian era, it seems not unreasonable to suggest they may have been present on the tiompán. On most traditional long-necked lutes, frets are created by winding animal gut around the neck. They have the twin advantages of being movable and fairly easy to replace. For the tiompán, the positioning of frets must be conjectural, though we may take our lead from the lutes that survive in other cultures.

In many cultures, long-necked lutes are played to accompany singing, with the instrument tuned to whatever the vocal range of the singer happens to be. Rendered into Western musical terms, two common tunings for three-stringed lutes are C-G-C and D-A-D, the latter reminiscent of the D-A-D-G-A-D tuning favoured by folk guitarists, originally devised by Davey Graham to facilitate playing along with traditional Moroccan musicians. In most cases, one of the two repeated notes is pitched an octave apart from the other. Held in the playing position, the two upper strings, including the uppermost ‘bass’ string, commonly act as drones, while the melody is played primarily on the bottom ‘treble’ string. Giraldus says that the favoured key in Irish music was B flat (Bb), or A sharp (A#), although we have no way of knowing to what extent what he thought of as Bb resembles its modern concert pitch equivalent. Taking Gerald at his word, however, we might perhaps tune our reconstructed tiompán down a tone from the commonly used C-G-C to give us Bb-F-Bb. We may assume that the tiompán was tuned in ‘just intonation,’ as used in ancient Greek music and many indigenous musical traditions today. Modern ‘equal temperament’ tuning was only developed in the 16^th century, by which time the tiompán had fallen out of use, or at least was no longer mentioned in manuscripts.

It took about 30 years to find someone to make me a chrotta and I probably don’t have enough years left in me to wait that long for a tiompán, so decided to try making one myself. Obtaining the soundbox was easy enough, just kept an eye out in charity shops for a turned wooden bowl of the right size and weight. My friend, Garth Reynolds, is a fine cabinet-maker, and his partner, Elaine Gregory, owns 80 acres of woodland. Her woods provided a beautiful Ash tree, some of whose timber I’ve used to make frame drum hoops. Garth took one of the remaining pieces, by now well seasoned, and made a blank to my specifications for the neck. I have spare guitar strings lying around, and some pieces of Red Deer rawhide left over from drum-making. Some cheap violin tuning pegs were bought online. With these pieces assembled, work could begin.

I wanted to give the neck a pleasing shape, which meant learning how to handle a draw-knife and a spokeshave, both of which I own but had rarely used. Several hours of careful labour and a lot of sawdust and shavings in the carpet later, a tapered shape I was happy with was achieved. I then sawed a shallow cut across where the headstock meets the neck and inserted a piece of horn with three small v-shaped notches cut in it as a ‘nut’ for the strings to pass over. Three holes were then drilled through the headstock and reamed using a hand tool designed for shaping holes for violin pegs.

Attaching the neck to the wooden bowl gave me cause for concern as I’ve never been much good at wood-working joints. Going extremely carefully though, to my delight, I managed to produce two passable slotted joints, one on either side of the bowl, into which the neck was inserted, glued and left overnight to dry.

I had cut down the neck where it passes across the interior of the bowl so that it wouldn’t impede the vibration of the rawhide I was going to use as a soundboard. However, not trusting the strength and stability of the hide, I had left a small pillar to support the bridge when it was put in place. Having glued neck and bowl together, I went through a pile of pieces of Red Deer hide and found one from the neck end so thick it had dried rock hard. Obviously it wasn’t going to need the pillar, being more than capable of supporting the bridge on its own. Rather belatedly, I decided to look online to see what I could find about the construction of acoustic instruments, particularly bridge and soundboard. I learned that the job of the bridge is to transfer as much of the vibration of the plucked or bowed string through to the soundboard as possible. Had I left the pillar, it would have reduced the vibration in the hide soundboard, reducing the volume and affecting the tone. So I cut the pillar out with a small hand-saw and sanded it flat.

The next job was to apply varnish to the woodwork to protect it from weather, insects and injury. I used a modern, shop-bought clear varnish, but there is evidence that our ancestors made and used natural varnishes a few thousand years ago. Varnish not only protects the wood, it also brings out the colour nicely, rendering variations in the grain more visible. Between coats, the wood was sanded using fine grains of sandpaper (240, 600 and 800). Again, this is a substitute for natural abrasives our ancestors would have used.

Having completed the varnishing, the next step was to fit the soundboard. The chosen piece of hide was put to soak in a tub of lukewarm water, adding some rawhide cord (also left over from drum-making) after an hour or so. It took a long time for such a thick, hard piece of hide to soften. Eventually, simply because time was getting on and I was getting impatient, I decided to give it a try. Using the experience of making drums, I punched small holes around the edge of the hide, adding two concentric circles of holes in the middle to act as soundholes. The rawhide cord was then threaded through the holes around the edge, criss-crossing the back of the bowl from one side to the other until I ran out of holes. This was then left to dry. As cord and soundboard dry, they shrink and, therefore, tighten. They needed to tighten a lot as the hide was still so stiff when I lashed it on that its surface resembled a contour map of a range of hills. Checking it next morning, it had flattened somewhat, giving me hope that it might flatten more during the course of the day. Fortunately, it did. Even when fully dry, the surface is still a bit rough, but it’s flat enough, tight and very hard, and produced a pleasing sound when struck as a drum.

Next I made a couple of bridges. The first attempt was based on the Iron Age High Pasture Cave bridge piece. However, on fitting it, it was apparent that such a design is useless on this instrument, giving an action that is far too high. I therefore took another small piece of seasoned Yew and made a much lower bridge. The violin pegs were then fitted, after having holes pierced through them with an awl through which to thread the strings. None of the spare guitar strings I had in the house exactly matched the gauges I had calculated would produce the best results, but they were fairly close. Now to put the thing together and see if it played.

The strings were looped over the three pegs at the far end of the neck, which protruded from holes pierced through the hide soundboard. They were then passed over the bridge and the nut and threaded through the holes in the tuning pegs. The pegs turned out to be a very tight fit and took some turning, occasionally resorting to a pair of pliers! New strings never stay in tune for long, and these were no exception. It took about three days for them to more-or-less settle. Given the difficulty of turning the pegs, I settled for being nearly in tune rather than spot on. Given the gauge of the strings available to me, I opted for a compromise C-F-C tuning as the most easily achievable.

Somewhat to my surprise, I found my conjecturally reconstructed tiompán has quite a pleasant tone and a decent amount of sustain, i.e. about 10 seconds. Having plucked, strummed, tapped and slapped it a little, I couldn’t resist making a recording to share. This was done prior to adding frets or removing the rawhide lashings that held on the soundboard. Even so, I was quite pleased with the result. After all, I’d never made a stringed instrument before and was just amazed that it made any kind of sound at all, let alone a relatively pleasant one! Tuning C-F-C.

To allow for the removal of the cords, the edges of the soundboard were pinned to the soundbox bowl with drawing pins. Fortunately, the wood of the bowl (species so far unidentified) is soft enough to be able to push the pins into, finishing off with a couple of light hammer taps. I did, however, have to use an awl to make holes through the rawhide which was, in most places, too hard to push a pin through. For the same reason, it took some hours to remove the excess hide and release the cords, gradually slicing through the hide with a Stanley knife.

I had bought some 1 mm thick nylon line to make frets with, but decided instead to try repurposing the removed rawhide cord. Rawhide has the advantage that, having been soaked before use, it dries and shrinks into place, tightening itself. I had no idea if it would work as frets, but decided to chance it. After soaking for some hours, the cord was both flexible and stretchable. Using a ‘just intonation’ calculator found online, I measured out the fret positions, marking them in pencil across the neck under the strings. I then began to tie on the frets, starting from the soundbox end and working back towards the nut. Each fret was tied about five fret positions down from where it was going to end up, then slid up into place. Since the neck gets thicker as you move towards the soundbox, this has the effect of tightening the fret. Sometimes I had to trim the width of the cord down with a pair of sharp scissors. The total string length from bridge to nut is 27.5 inches (70 cm), and from the front edge of the soundbox to the nut is 21.5 inches (55 cm). This meant tying 24 frets. A lot of work, especially since I had no idea if the rawhide cord would actually do the job. The overnight wait for it to dry was quite anxious…

Next morning, I picked the thing up and, yes, the frets worked! Well, all except a couple in the middle that were a little lower than the next one up. Inserting a small file under the strings, I was able to file down the too high fret so that the ones below it could sound properly. Other than that, all my recycled rawhide frets performed as they should. Yay!

By this time, the strings were beginning to settle, becoming better at holding their tuning. This encouraged me to try a couple more recordings. The experiment switched from making to playing. From watching videos online of similar long-necked lutes being played in other cultures, I had some ideas to try out and plenty of inspiration to draw on. I’d become especially enraptured by the music of the late Iranian setar virtuoso, Mohammad Reza Lotfi, a musical genius virtually unknown in the West but who ranks alongside Ravi Shankar in the expressiveness, purity and spirituality of his playing. While my humble efforts will never get anywhere near such giants, they offer a vision of the mountain-top to strive towards. The word setar, incidentally, means ‘three strings,’ although modern setars have four, usually arranged in three courses.

For my first recording with frets in place, I used a long plectrum called a risha, reshee, or mizrab, used to play Middle Eastern lutes such as the oud. I chose this because of references in Irish manuscripts to the tiompán being played with a ‘wand.’ Some have interpreted this as meaning a bow, but this type of long pick, originally made from cow horn, is equally worthy of being called a ‘wand.’ Tuning: C-F-C.

For the second recording with frets, in keeping with the playing style described in the 12^th century manuscript quoted above, the strings were plucked or strummed with the nails and fingers of the right hand, while the soundboard was played like a bongo or conga drum, initially with both hands, then with the sides of fingers and thumb, rocking the right hand to and fro. It became immediately apparent, as the recording shows, that playing the soundboard like this causes the strings to sound as a rhythmic drone. The strings vibrate to the drum-beat a lot more than I thought they would. Tuning: C-F-C.

All in all, having started out with fairly low expectations due to my shaky crafting skills in some areas, I’m quite pleased with the results and looking forward to improving my playing technique and maybe even trying singing with the tiompán, as the manuscript sources indicate was done. I’m not sure that I’ve brought an authentic tiompán back to life for the first time in more than half a millennium, but, given that no one seems quite sure what an authentic tiompán was, I’ll settle for what I’ve got until further evidence comes to light. At this point, that would probably be one being found in a peat bog. Now wouldn't that be something?

Making and playing my tiompán during the Covid-19 pandemic has certainly been beneficial to my mental health and general well-being. In Chapter XII of his Topography of Ireland, Giraldus Cambrensis, has this to say on the beneficial effects of music:

“The sweet harmony of music not only affords us pleasures, but renders us important services. It greatly cheers the drooping spirit, clears the face from clouds, smooths the wrinkled brow, checks moroseness, promotes hilarity; of all the most pleasant things in the world, nothing more delights and enlivens the human heart. … Moreover, music soothes disease and pain; the sounds which strike the ear operating within, and either healing our maladies, or enabling us to bear them with greater patience. It is a comfort to all, and an effectual remedy to many; for there are no sufferings which it will not mitigate, and there are some which it cures.”

Awen to that!

Many blessings, keep safe and be well,

Greywolf /|\

For two excellent articles on the Tiompan, see:

Polly Jones: https://owlcation.com/social-sciences/The-Irish-Timpan

Ann Buckley: https://www.academia.edu/19416124/What_was_the_Tiompán