by Cynthia Cathcart
First published in the Folk Harp Journal, Summer 2009 Issue no. 143
One morning, while waiting at the school bus stop, I turned to one of my neighbors and asked him, “So, what do you do?” He replied,“I’m a goldsmith.” What a surprising answer! Lately I’d been thinking about how harpmakers were putting twisted or wound strings in the bass of the wire-strung clàrsach, or specifying unavailable thick gauges for the bass. Talk was that a gold string might be a good substitute, but I had no idea where to get one. How does one just walk into a jewelry store and ask for a piece of gold wire? But now, here in front of me, was a goldsmith!
“Can you help me get a gold string for my harp?” I asked, and his answer was a litany of questions. “Sure. What alloy do you need? What gauge? What hardness? How hardened?” to which I didn’t have even one sure answer. More perplexing, I didn’t know where to look for the answers. “So, there’s no data?” he asked. “Not that I know of. Can you help me find those answers?”
And so the journey began…
Following a long decline, which was well underway in the late 18th century, the wire-strung harp of Ireland and the Highlands and Islands of Scotland disappeared almost entirely for nearly three-quarters of a century. This small harp, called the Clàrsach (Scots Gaelic) or Cláirseach (Irish Gaelic), has been undergoing a revival over the past few decades. As a result, harpers have demanded and received instruments of increasing authenticity and quality.
There are some who perceive a weakness in the tone quality of the bass strings on the early, low-headed, clàrsach. Some experts suggest this may be a function of modern expectations, that what we perceive as unsatisfactory tone quality was perfectly satisfactory to medieval and renaissance listeners. Other experts suggest that something is lacking in our modern recreations of this early instrument. Still others propose that, authentic or not, we should play on instruments that satisfy audiences today.
In part as a response to this conundrum, the concept of putting precious metal strings on the metal-strung harp has drawn some attention. [1] Indeed, there are several references in literature to harps strung with both silver and gold.
From Galeran de Bretagne, a romance written by Jean Renaut, which is from a manuscript written in about the year 1230:
“Doulx est li chans et doulx li diz.
Et cil li chante tant et note,
Qu’elle scet le dit et la note;
A sa harpe l’a accordee
Qui estoit d’argent encordee
“The words and the music were sweet,
and he sang and played the lai until she knew both the words and the tune;
then she tuned her silver-stringed harp to the lai.”[2]
From a poem in the Irish Ulster Cycle, we have the story of a special harp:
The noble bed is prepared;
Fionn is the first to approach it;
Scathach asked before retiring,
The loan of the musician's harp.
The household harp was one of three strings,
Methinks it was a pleasant jewel:
A string of iron, a string of noble bronze,
And a string of entire silver.”[3]
Inspired by this and other information that emerged from the authors’ review of the literature, it is the purpose of this paper to share the results of experiments undertaken using silver as a stringing material on the clàrsach.
As the revival of the wire-strung harp takes hold, musicians and craftsmen are working to rediscover the techniques of making and playing this instrument.
Part of this rediscovery is the desire to find the “perfect” string material. People who play and study wire-strung harps recognize the common use of the base metals brass, bronze and, more recently, steel.
A review of the available string options for the bass reveals a particular deficiency. The experiences of players using phosphor bronze on the wire-strung harp indicate that bronze breaks too often to be entirely satisfactory.[4] Brass wire is extremely popular for many harpmakers, and the literature confirms it was available from Roman times, and so meets the historical requirement that some desire.[5] It is considered by many to be the metal of choice.
A high-headed harp is necessary in order to achieve the anticipated bass sound with monofilament[6] brass wire. The high-headed harp began to appear in the late 17th century. Consider, for example, the Mullagh Mast harp, dated to about 1700, or the Sirr harp dating from the late 17th to early 18th century.[7] It may well be that the high head was developed as a response to changes occurring in newer music.
A high head on the harp is clearly unsatisfactory for players who desire the earlier low-headed instrument. However, on a low-headed harp, the quality of the bass seems weak or may even buzz as the strings must be relatively loose, as compared to the higher strings, to obtain the pitch desired. If a thicker wire is used to achieve higher tension for the bass strings, it becomes too stiff to be easily manipulated. For this reason, some players of low-headed harps opt for wound or twisted brass strings in the bass. This option gives adequate thickness without compromising flexibility, but at the detriment of authenticity.
General opinion is that wrapped, wound or twisted strings give an inferior sound when compared to monofilament strings. This is supported by the experimental results given later in this paper. We suspect there is a loss of energy caused by the duo-twisted strings themselves as they twist and un-twist when vibrating. Notwithstanding the quality of their sound, they are unpleasant to play, especially when using the fingernail technique that many wire-strung harpers employ.
Harpists are concerned with the musicality of their harp strings, not merely perfection of pitch. The goal is to obtain a pleasing sound at a specific pitch, rather than merely being an exercise in manipulating length, tension and density. For example, one may be able to lower tension and tune a string to a pitch much lower than designated for that particular string. However, even if the pitch is absolutely correct, a string that feels “floppy” may not be satisfactory for a musical performance. Likewise, using a very thick, dense brass string will produce a lower pitch, but the thicker the string, the harder it is to bend, making it impractical to use on the harp.
The attraction of precious metals as stringing material is both practical and romantic in nature. As a practical matter, both silver and gold have a higher density than traditional stringing metals, with gold being the denser of the two. Does this increased density give a better tone in the bass? Using a denser metal such as gold or silver, can a thinner string be used to produce a sound superior to brass or bronze on the same harp?
Existing sources of information concerning historic use of wire strings are not published or are hard to find. Published experimental analysis and empirical data is currently lacking in the field.
Our goal is to approach the issue with an open mind in order to discover, through experimentation and research, which alloys and treatments contribute to the creation of a musical wire string.
A blind experimental method was used to pinpoint which mechanical properties of silver contribute to the most musical or best sounding harp string to our ears. Our hope is that our results will further the revival of an instrument that has caught the minds and hearts of many musicians.
The physical principals that best apply to the issue of harp strings are represented by Mersenne’s laws.[8] These indicate that the fundamental pitch of a plucked string is determined by three things: length, tension and mass per unit length (linear density).
The fundamental frequency of a vibrating string is proportional to the square root of the tension and inversely proportional both to the length and the square root of the mass per unit length.
| Period | Frequency | Constant | |
|---|---|---|---|
| Law 1 | T ∞ L | f ∞ 1/L | W , F |
| Law 2 | T ∞ 1/√F | f ∞ √F | W , L |
| Law 3 | T ∞ √W | f ∞ 1/√W | L , F |
|
W = mass per unit length of string L = length of string F = tension applied to string |
|||
Specifically, Mercene’s Law states, where ‘frequency’ is the pitch of the plucked string:
Stated in a slightly different way:
These formulas show that three factors affect the pitch of a vibrating string: length, tension and mass per unit length (MpUL). The length of any given harp string is fixed by the instrument design. The tension is variable within certain boundaries, but increasing the tension too much risks string breakage or damage to the harp. If the tension is too low the string becomes too loose to produce sound.
Changing either the diameter of the string, or changing the density of the string material can manipulate the MpUL. A fixed length for the string at a suitable tension, the pitch can only be lowered if the string’s diameter or density is increased.
During our experiments, the crystal grain structure and temper of the metal were also revealed as important elements in the equation. The type and percentage of the wire’s hardening resulted in noticeable variations of sound in strings of identical length, gauge and tension where temper was the only differentiation.
A string gauge can be changed on a harp and still attain the necessary pitch, as any harpist who has needed to substitute one gauge wire for another can attest. In our experiments, the same 1 millimeter gauge strings were used on two different harps with a disparity of about four inches in length at the same pitch. Therefore, the tension must have been different between the two instruments, but this difference was not discernible to any significant degree by those involved in these experiments.
Early references to the specific alloy of the strings in question were not to be found. Our research uncovered very little on the composition of historical silver or gold. Most alloys used were a natural result of the metals present in a particular ore. (Gold, silver and copper often appear together in the same mine.) While it was known that brass was an alloy of copper and zinc, and that bronze was an alloy of copper and tin, the specific proportions are not recorded in historical writing from this period. Nor is there much detail regarding the manufacture of the far more common stringing materials, such as brass and bronze, referenced in the historical writings.
Silver wire with draw plate and tongs.
Silver wire being pulled through the drawplate.
Sterling silver is an alloy[10] of pure silver (92.5%) and copper which adds hardness (copper represents 7.5% of the alloy).[11] We know silver was being used from early times on the island home of the clàrsach. Henry II (1133 – 1189) established sterling silver as a standard alloy. The silver penny was the first coin struck in Scotland, during the reign of Alexander I (1078-1124).[12] Silver pennies were also being struck in Ireland in the 12 th century.[13] Sterling silver is harder and more durable than fine (pure) silver and also exhibits good working characteristics, for which reasons it has been the standard silver alloy for hundreds of years.
When choosing an alloy to experiment with, there are several factors to take into consideration.
Sterling silver is a precisely defined alloy. However, today’s gold is not. Modern gold wire is destined primarily for jewelry production, and manufacturers often use trade-secret formulas to create different alloys.
By using a standard silver alloy the experimental results from our study may be replicated with confidence, as the exact same alloy can being used.Once we have established a better understanding of wire strings through the use of silver, we may then apply what we have learned to other metals.
Demonstrating that a metal of higher density produces a better sounding bass string may be done using either silver or gold. By choosing to experiment with the less expensive of these two metals, we could afford to be more comprehensive in our experimental plan.
Thus the reasons for the choice of sterling silver for the experimental metal were:
The two harps used in the experiments were made by different luthiers, to different standards, while still of the highest craftsmanship. Additionally, both harps are replicable models made by these luthiers, and so someone attempting to duplicate these experiments should be able to obtain a nearly identical harp to what was used here.
The Witcher-built harp has for its lowest pitch the second G below middle C. The length of this string is 29.5 inches (74.93cm), as measured from the tuning pin to the string shoe. The harp is built of American birch and maple, has a sound box constructed of pieced and glued sections of wood, withfour partially obscured sound-holes on the soundboard of 0.875 inch (2.22cm) in diameter. There are two large access holes in the back of the soundbox. This harp is the “Balinderry Reconstruction” model of Mr. Jay Witcher of Maine, which is in essence an enlarged Lamont reproduction designed by him.
The Scottish Ardival harp has for its lowest pitch also the second G below middle C. The length of this string is shorter than it is on the Witcher harp, at 25.375 inches long (64.45cm). It is built of Scottish sycamore with a hollowed body, glueless construction for its soundbox. It has 4 small sound-holes on the soundboard, two that are 1 inch in diameter (2.54 cm) and two that are 0.75 inch in diameter (1.905 cm). It has a solid back. It is the “Rose” model, which is an interpreted version of the historical clàrsach of the 14th and 15th centuries and not a precise reproduction of any one historical harp.
Both these harps are low-headed in design, as shown by comparisin to early specimens. The well-known three smaller, low-headed harps, the Queen Mary, Trinity College and Lamont harps, each have a lowest string of either 24 inches or 24.4 inches (61 or 62 centimeters) long. These compare favorably with the Ardival. The Witcher compares with the later, large low-headed designs as can be found in the Castle Otway harp, with a lowest string at 30.31 inches (77 centimeters).
However, we cannot be absolutely certain of the historic harps’ exact range or pitch, nor know if they were always tuned in the same way. The earliest known tuning instructions for the wire-strung harp were published in 1840, from notes taken no earlier than 1792. [15] Both the Ardival and the Witcher were designed by their makers to have Great G (the second G below middle C) at the bottom of their range, and we performed our experiments using that pitch.
One interesting discovery of the experiments was that equal results were found on either harp with the same strings. In other words, a string which sounded good on the Witcher harp sounded as good on the Ardival harp, despite the difference of slightly more than 4 inches in length of the same pitched string on the two instruments.
In order to compare the sound of several different strings, the ideal situation would be to place those strings on precisely identical harps and test them against each other side-by-side. This is not only impractical, but well nigh impossible as no two harps, even of the same maker and materials, ever sound exactly the same even with identical strings installed. The diversity of the organic component of the wood precludes such a precise comparison.
The solution to this challenge was to set up a sound recording studio wherein each string was put on the self-same harp and recorded in a controlled situation for later comparison. The controls were as follows:
The strings were each prepared off-site for the sole purpose of these experiments. The strings were given to the harper identified only with a numeral. No relationship existed between the numeral and any characteristics of the wire. The gauge of every wire used was precisely 1 millimeter.
The strings were each tied onto the same metal toggle, using the same type of knot each time. The strings were wrapped 3 times around the toggle and 5 times around themselves, as shown in the following photograph:
Each wire was put on the harp in the same string/pitch position and tuned to the same pitch, as measured by an electronic tuner.
The number of wraps necessary around the pin to obtain the pitch were also carefully counted and recorded.
Benjamin Zastrow, gemologist and goldsmith, prepared the strings for these experiments. He designed the experiments, choosing the specific types of treatments (i.e. hardening) for the individual strings.
Eric Cathcart, recording producer, was responsible for the recording of the strings on the harp. He chose the specific pieces of recording equipment used, and engineered the recordings themselves.
Cynthia Cathcart, professional harper, installed each string on the harp herself, performed the recorded examples, is the lead author of this report, and is responsible for the subjective ratings of each strings' performance.
Our hypothesis was that a monofilament wire of a dense, precious metal (specifically sterling silver) would produce a better bass tone than either twisted or monofilament brass and bronze.
We expected that objective experimentation would also reveal how the hardness or softness of the treated precious metal would affect musicality. This would demonstrate the dependence of musicality on the crystal grain structure of the metal. Different techniques for creating hardness were used to discover which technique created the most musical sterling silver string.
The sterling silver alloy can be hardened either by exposure to heat or by working, or a combination of these two methods. Metal is "worked" when it is hammered, rolled, bent, or otherwise physically processed. Drawing wire through a drawplate to reduce its diameter would result in work-hardening of the metal.
The first set of results, for yellow brass and phosphor bronze, are included in the following table as representative of the types of strings commonly in use. Both were monofilament.
Compound wires (wrapped, twisted and wound) are represented by only one sample. Even had these wires proved to be superior in sound, many harpers avoid playing these types of strings because of the unpleasant feel of the strings and fear of damage to the fingernails. The twisted brass included in the table was the original standard bass string on the Ardival Rose clàrsach.[17]
Mr. Witcher of Maine, the builder of the other harp used in the experiments, disallows the use of any sort of wrapped, twisted of wound strings on his harps. Thus, the only non-monofilament example is from the Ardival harp.
All wires were cut a uniform 2.5 inches (approx. 6.35 cm) above the tuning pin prior to wrapping the end on the pin and tuning. For all tested wires, there were five wraps around the pin to gain the correct pitch, with the single exception of the full-soft silver string.
The full-soft silver string completed 8 wraps around the pin and still did not gain the correct pitch. The original gauge of the wire was 1 mm, as it was for all the test wires, but when the decision to abandon this string was made at the time of making the eighth wrap around the pin, the measured gauge was .89 mm. The string had stretched thinner, the sound was dull and produced very little resonance, and it was never able to hold a pitch. For these reasons, this one string was not recorded.
| String Identifier | Alloy of the string | Specific treatment of the string | Ratings | ||
|---|---|---|---|---|---|
| As first heard | 1 month later | 7 years later | |||
| Common Strings (brass and bronze) | |||||
| YB | Yellow brass | Spring hard. As supplied by The Instrument Workshop[18] (70% copper, 30% zinc) | 7 | 7 | 7 |
| PB | Phosphor Bronze | Spring hard. As supplied by The Instrument Workshop (95% copper, 5% zinc, trace phosphorus) | 6 | - | - |
| TB | Twisted brass | Two pieces of spring brass twisted together | 6 | - | - |
| Initial test series, to determine direction: | |||||
| 1-1 | Sterling Silver | Three-quarters work hardened | 7 | 7 | 8 |
| 1-1 | Sterling Silver | String 1-1, afer one full day on the harp | 8 | 8 | 8 |
| 1-2 | Sterling Silver | Heat hardened then drawn to spring hard | 4 | 3 | 3 |
| 1-3 | Sterling Silver | Heat hardened then drawn to full hard | 4 | 4 | 4 |
| 1-4 | Sterling Silver | Heat hardened to three-quarters | 5 | 5 | 6 |
| 1-5 | Sterling Silver | Full hard, work hardened | 6 | 7 | 7 |
| 1-6 | Sterling Silver | Dead soft (unsatisfactory. See discussion above | 1 | 1 | 1 |
| 1-7 | Sterling Silver | Semi-spring, work hardened | 9 | 9 | 9 |
| Second Test Series, all work-hardened (no heat hardening) | |||||
| 2-1 | Sterling Silver | 1/2 hard | 7 | 7 | 6 |
| 2-2 | Sterling Silver | 3/4 hard | 7 | 6 | 7 |
| 2-3 | Sterling Silver | Full hard | 9 | 8 | 9 |
| 2-4 | Sterling Silver | Semi-spring | 8 | 9 | 9 |
The ratings were made on each string three separate times, and are given in the far right columns in the table. The ratings are on a scale of 1 to 10, where 1 is poor and 10 is excellent. A poor sound is typified by weak tone, limited resonance, a short sustain or ring, and an inability to play dynamics (loud and soft) on the string. The words dull, dead or thuddy would describe a poor string. A good sound is typified by ample resonance, full sound, strong harmonics, capable of a dynamic range, and a long sustain. A good string would be described by words such as rich, alive, or warm.
The first rating was made at the time of the initial placement of each string on the harp, upon hearing the string live for the first time. The "1 month" rating was made three weeks later, from listening to the recordings. A final rating was made seven years after the experiments, again from the recordings. The ratings, made at different times, under different circumstances, are remarkably similar and thus we feel confidence may be placed in them.
Treatments for the wires were specifically as follows:
The hardness matrix used was as in the following table:
| Temper: | Full soft | 1/4 hard | 1/2 hard | 3/4 hard | Full hard | Semi-spring | Spring |
|---|---|---|---|---|---|---|---|
| Hardness: | 0% | 20% | 35% | 50% | 60% | 75% | 85% |
The final result shows that the ideal sound to be gained from a sterling silver string is from either a full hard or semi-spring, work hardened string, which is brought to this stage in degrees. In other words, multiple "drawings" are required to harden the string; it is not achieved in one pass but gradually over several passes. This represents a string with somewhere between 60% and 75% reduction in cross-sectional area.
We were pleased to find that excellent sound could be obtained from full or semi hard wires, and that we did not need to use a spring-hard wire. Full spring hard would create a more brittle string. Working at a very brittle level has the potential to produce greater breakage at the points of stress along the string, specifically at the tuning pins and sound board, where the wire is bent. This we know from the lamentable breakage history of full spring phosphor bronze wires. A string of a softer metal will be more durable on a harp, as it will be more ductile and thus more elastic.
In anecdotal proof of the durability of these sterling silver strings, at the time of publication of this paper a full seven years after these experiments were done, the sterling silver strings on one of the test harps is the very same string that was rated a '9' on the Table of String Comparisons (above.) The other harp has a silver string that was installed at the same time.
The experiments in sterling silver were relatively affordable and offer a suggestion for future experimentation. After careful review of all findings, the authors believe that it may be beneficial to experiment with electrum, "Silver-rich gold" also called green gold. Likewise, further controlled experiments with 14, 18 and 22 carat gold may prove useful to discover if our sterling silver experimental results hold true for gold.
The silver experiments lead the authors to hypothesize that full hard silver-rich gold, at 18 carats (75% gold and 25% silver) may produce a superior sound. Cursory research shows that there is evidence of this alloy being used historically, in large part due to its presence in nature.
Other suggestions for further objective experimentation are to explore widely different pitches and lengths of strings.
The results reported here have been applied to strings of other length and pitch with success. Sterling silver strings on a wire-strung clàrsach were recorded for the first time by one of the authors in a professional recording just months after the completion of these experiments, and these lovely strings continue to be played upon, listened to, and enjoyed.
For some in the wire-strung harp community, sterling silver has been the answer to the need for affordable, monofilament, and beautiful-sounding solution to gaining an agreeable bass tone on the low-headed harps of the tradition. It is our hope that our results will benefit the revival of this striking instrument.
[1] See for example: Heymann, Ann and Heymann, Charlie, “Strings of Gold,” Journal of the Historical Harp Society, Summer 2003. pp. 9-15.
[2] Lines 2317-2321 from the manuscript of Jean Renaut. Source of the translation: Renart, Jean, and Lucien Foulet. Galeran de Bretagne. Paris: 1925.
The poem may be found in the original French in the following book: Jean Renart, Anatole Boucherie, and Camille Chabaneau. Le roman de Galerent, comte de Bretagne. Montpellier: Au Bureau des publications de la Société pour l'étude des langues romanes, 1888.
[3] Source: O’Curry, Eugene. On the Manners and Customs of the Ancient Irish – a Series of Lectures by Eugene O’Curry, M.R.I.A. ed. By W.K. Sullivan, Ph.D. (John F. Fowler, Dublin. ) Vol. III p. 223 (From The Legend of Fionn MacCumhaill, part of a poem composed in the late 12th century). O’Curry’s source is M.S. No 23/E.22. R.I.A. p. 420
[4] Certain bronze-strung instruments have been known to break multiple strings in a single event.
[5] Young, Suzanne M. M. Metals in Antiquity. BAR international series, 792. Oxford: Archaeopress, 1999. (This reference contains a useful article on Roman and Medieval Brass production)
[6] Monofilament wire is a single strand of wire. In other words, it is wire that is not two or more wires wrapped, twisted, or wound about each other.
[7] Rensch, Roslyn. Harps and Harpists. Bloomington: Indiana University Press, 1989. pp 104-105.
[8] Mersenne, Marin (1588-1648). Harmonie Universelle: the Books on Instruments. Paris, 1636-37
[9] Berg, R. E., & Stork, D. G. The Physics of Sound. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. 1982
[10] An alloy is a combination of two or more metals that, when mixed together, have different properties than the original metals. For example copper is easily worked into various shapes but it is not very hard. Tin is extremely soft. When mixed together copper and tin form the alloy bronze, which is hard and durable while retaining malleability.
[11] Linick, Leslie Lloyd.Jewelers' Workshop Practices; Processes, Shop-Kinks, Short-Cuts and Trade-Secrets of Jewelry Manufacture and Precious Metal Finishing in Non-Technical Language. Chicago: H. Paulson, 1948. p 99 (among numerous others.)
[12] Excerpt from Confirmed Charter (Hadinton Collection) by Thomas Earl of Mar, Chamberlain of Scotland, to John of Mar, a canon of Aberdeen, confirmed by K. David II, 22 Nov. 1358: " Reddeudo indo, &c. unum denarium argenti [silver penny], qui " vocatur Sterlingus, nomine alba3 firma:." As quoted on page 215 of Rodger, William. The Feudal Forms of Scotland Viewed Historically. 1857.
[13] Edwards, Nancy. The Archaeology of Early Medieval Ireland. Middle Ages series. Philadelphia: University of Pennsylvania Press, 1990. p 176
[14] On April 28, 2009 the price per ounce of gold was $891.00 (689.58 EUR). The price for an ounce of silver was $12.56. (9.65 EUR). Source: Kitco, Inc. “All metal quotes: London Fix Price” <http://www.kitco.com/market/> accessed on 28 April, 2009
[15] Bunting, Edward. The Ancient Music of Ireland, Arranged for the Piano Forte. Dublin: Hodges and Smith, 1840. p. 23 Bunting began his collection at the Belfast Harp Festival, 1792.
[16] These arrangements have been published by Cynthia Cathcart in From My Music Stand. Silver Spring, Md: Highland Circle Pub, 2001. pp 32, 33
[17] Ardival Harps has since switched to a monofilament sterling silver string for this position.
[18] The Instrument Workshop, PO Box 1060. Ashland, Oregon, 97520. USA +15415520989
Submitted by Cynthia Cathcart, 28 September, 2010
Except where otherwise noted, content on this site is licensed under a Creative Commons License.
Permissions beyond the scope of this license may be available by contacting us at editor@wirestrungharp.com.
