Making an Anglo-Saxon Lyre

George von Gerolstein
(MKA: George A Cavender)
Submitted in consideration for The Bryn Madoc Baronial Champion of Arts and Sciences, AS XLII
Distributed under the the Creative Commons Attribution-ShareAlike 2.5 License *

Fig. 1: King David composing the Psalms. From Folio 30V of The Vespasian Psalter, English circa 750 (WEB: click for a larger version)
Public Domain image courtesy of Wikipedia


Overview and History: The Anglo-Saxon lyre is a stringed musical instrument dating (conservatively) approximately from 500-900 AD. Though I shall refer to it as an Anglo-Saxon lyre, a more correct designation would most likely be "Germanic Lyre", as similar artifacts have been discovered in places other than England, primarily in Germany. (**) This of Course makes sense given the Germanic origins of the Anglo-Saxon people. I chose to work primarily from the Sutton Hoo burial finds, as they are the most widely known. The Sutton Hoo burial mounds contain several graves, but the one most commonly thought of is the "Ship Burial", which as netted us the famous helmet (figure 2) , as well as the Lyre (figure 3)

Figure 2: The Sutton Hoo Helmet
Image Courtesy of Wikimedia Commons

Figure 3: Remains of the Sutton Hoo Lyre (WEB: Click for Larger)
Housed in The British Museum, Providence of picture unknown

As can be readily seen, very little of the original Lyre remained. So little in fact that the original reconstruction (Figure 4) looks nothing at all like the Psalter image above. Some people believe that this was because the Germanic tongues call both what we call a Lyre and what we call a (frame) Harp (e.g. the Irish harp) Haerpa (or some variation). Thus the 1948 reconstruction looks more like a frame harp than a lyre.



Figure 4: 1948 Reproduction.
Source of image and location of original object unknown


Eventually, someone realized that this was most likely an incorrect reconstruction and it was changed to a form more supported by other finds and manuscripts.

In The Treasure of Sutton Hoo , Bernice Grohskopf explains the impetus for the change:

In 1948 the musical instrument was reconstructed as a small, quadrangular harp, but additional fragments, originally thought to belong to the roof (of a building), proved to be part of the instrument. Newly reconstructed in 1969, it is a round lyre, 29-inches long, 8-inches wide, with six gut strings each 20-inches long, of varied thickness and tension. It has been tuned to a pentatonic scale, with a pitch between alto and tenor registers. The new reconstruction is supported by comparisons with manuscript illustrations as well as with fragments of lyres found on the continent. Beaver hairs were identified on the outside of the frame of the instrument and it is believed that it was kept and buried in a beaver skin bag with the fur inward. The maple wood fragments that led to the reappraisal had originally been boxed and cataloged as roof remains, and at that time it was thought that all that survived of the instrument had been contained in the bronze bowl. The old reconstruction was based on these fragments found in the bowl; many photographs were reproduced, and the music was recorded. (LC: GT3380 .G75 1970)

This of course brings us to the currently displayed reconstruction, upon which I primarily based this project. Figure five shows the Dolmetsch reconstruction.


Figure 5:Currently displayed reconstruction (WEB: Click for larger)
Lyre reconstruction by Messrs. Dolmetsch. Photograph by Steven J. Plunkett (cc)


My Reconstruction: My reconstruction is an attempt to make a period plausible, playable instrument similar to the Sutton Hoo lyre, while maintaining a reasonable budget (including labor). The following sections will outline the inspiration, materials and procedures used to produce the artifact.


Part 1: Body:

Overview: The body of the lyre is made from 7/8" thick walnut. Walnut was chosen because I happened to find a suitable piece in the scrap pile of an associate. This allowed me to maintain my "reasonable budget", as well as providing a beautiful contrast with the other woods. It also limited the size of the reproduction, as the off cut was approximately 22 inches long. This means that the final product will be shorter than the lyre shown in figure 5. That said, based upon the remnants found (figure 3), I am uncertain as to how Messrs. Dolmetsch arrived at an overall height of 29 inches. Walnut wood (Juglans regia) was known in Europe (and England) at the time, but this sample is an New-world species, most likely Juglans nigra. The shape of the

Cutting/Shaping: The body was sketched onto the board with a normal graphite pencil. The width of the body was increased slightly (one inch) from the original to avoid a naturally occurring knot. Figure 6 is the general pattern used. The shape was formed using a table saw, a jig saw and a 2" forstner bit for the internal curves. A hole was drilled into the bottom for the tailpeg.

Figure 5: The Pattern (WEB: Click for larger)
Image (c)2008 by Author (cc)

Figure 6: The body piece
Image (c)2007 by Author (cc)


Hollowing: The body was hollowed using a combination of a modified (point removed) forstner bit in a drill press and a straight cut router bit. After using the aforementioned tools, the inside was sanded smooth. Figure 7 shows the hollowing partway through the sanding.

Figure 7: Partway through sanding the inside. Notice some of the telltale forstner bit "rings" are still apparent.
Image (c)2007 by Author (cc)


Part 2: Headstock:

Overview: The headstock on the original had its grain oriented 90 degrees from the body, I chose to do similarly. Initially, the plan was for the headstock to be of the same walnut, however, the stock that was to be used (the part removed from between the arms) cracked during cutting. Thus I was forced to use another wood, namely Oak. Again, as with the walnut, while oak was known to the English and Europeans (Quercus robur, among others), this oak was an American species (most likely Quercus falcata or Quercus georgiana)

Cutting/Shaping: The headstock was sketched onto the board with a normal graphite pencil. The shape was formed using a table saw, a jig saw and a 2" forstner bit for the internal curves. Figure 7 shows the plans used.

Figure 8: The Pattern (WEB: Click for larger)
Image (c)2008 by Author (cc)


Part 3: Fitting the headstock/ body together:

Overview: The headstock and body are joined by a mortise/tennon joint 1.5" long. this joint is glued (though we cannot tell if the original was) and then riveted with brass.

Cutting/Shaping: As is evident in Figure 9, the tennon is cut into the body, while the headstock receives the mortise. The mortise was cut using a table saw, while a handheld coping saw, chisel and wood rasp formed the tennon.

Figure 9: The Mortise/ Tennon Joint
Image (c)2008 by Author (cc)


Part 4: The Soundboard:

Overview: The soundboard is made from thin maple, as was the original (see Grohskopf, above). As with the original, tacks were used to affix it to the body. The original's tacks (or what was left of them) were crude wedges of copper alloy, which led Messrs. Dolmetsch to use cut copper wedges. I used commercially available solid copper cut tacks. The primary difference in these and the tacks used by Messrs. Dolmetsch is the obvious heads. However, headed nails were known in period, so it is not that much of a stretch.

Cutting/Shaping: A rough outline of the body was marked onto the maple (after aggressive sanding to reduce some thickness) and cut out with a jig saw. Then pilot holes were drilled in the soundboard and body to prevent splitting. Thin layer of glue was applied, and then the nails were driven in. After the glue dried, a veneer trimming bit was used to trim the soundboard to final shape. The offcut maple from between the arms was then applied to the headstock in a similar manner. In order to prevent any difficulty/ damage to the strings, the three tacks at the bottom of the headstock were countersunk Also at this point, the peg holes were drilled in the headstock, and tapered using a rattail file which had been chucked in a drill and run in reverse. In order to prevent the file from going too deep, tape was used as a "bit stop".

Figure 10: The soundboard attached
Image (c)2007 by Author (cc)

Part 5: The bridge, Tailpiece and tailpeg:

Overview: While technically different pieces, their construction is similar enough to cover in one section. The original most certainly had all 3 parts, though none apparently survived. However, we have found existent bridges all over Europe, and many stringed instruments to this day still use tailpegs and tailpieces. I chose walnut for the bridge and tailpiece due to the aesthetics of the contrast.


Bridge: A rough outline of the shape was marked onto the walnut and it was cut with a hand held coping saw. Files, chisels and emery boards were then used to for the requisite shape. Finally, a fine triangle file was used to apply the string notches.

Tailpiece: A rough outline was cut in walnut stock and then formed with rasps and a drill mounted sanding wheel. Holes for the tailgut and strings were drilled on a drill press.

Tailpeg: This was a simple modification of a commercially available shaker peg. I removed the knob, and extended the cylindrical section through carving with a knife. It was then glued in place.

Figures 11-12 show an early step in forming both the bridge and tailpiece. Finished versions can be seen in Figure 10 (above)

Figure 11: Rough forming the tailpiece
Image (c)2007 by Author (cc)

Figure 12: Rough forming the bridge
Image (c)2007 by Author (cc)


Part 6: The Pegs:

Overview: We have no idea what the pegs in the original Sutton Hoo find looked like. It is likely that they were wood (as they didn't survive well). Another similar find, the Trossingen lyre (WEB: Click for image (cc)) had surviving pegs with square heads and tapered shafts. I chose to use this style of peg, although mine turn from the front, rather than the back. For materials, I used commercial square "hardwood" dowels. I have no idea which tree species these come from.

Cutting/Shaping: After cutting 3-4 inch "blanks", the pegs were shaped. The shaping process has 3 steps: Rough carving, Shaving and sanding. Rough carving is done with a knife, and renders a roughly round shape to the shaft. Shaving is done with a peg shaver, the only specialty tool I purchased for the project. Sanding then smoothes the transition from the square prism of the peg top to the conical shape of the peg shaft. Figures 13-18 show the process. After making the pegs, a small hole was drilled to accommodate the string.

Figure 13: Peg blank
Image (c)2008 by Author (cc)
Figure 14: Rough carving
Image (c)2008 by Author (cc)
Figure 15: Peg shaping (note thepeg shaper)
Image (c)2008 by Author (cc)
Figure 16: Peg after shaping but before sanding
Image (c)2008 by Author (cc)
Figure 17: Sanding Peg
Image (c)2008 by Author (cc)
Figure 18: Finished peg (sans string hole)
Image (c)2008 by Author (cc)


Intermission: Sanding/ Finish

Overview: At this point, everything (except the pegs) had to be sanded. Using a variety of sand papers starting with 60-80 grit and working down to 240 grit. This brings out the beauty of the wood and manages to smooth out any slight imperfections in the cutting/ shaping process. As for finish, there are plenty of options, ranging from the perfectly period through the plausibly period to the totally modern

Un-finished: using no finish is an option. Unfinished wood, However is prone to humidity related changes, warping, staining from being handled, etc. Thus, in my opinion, it is not a good finish for musical instruments.

Boiled Linseed Oil: Historically in Europe boiled linseed oil was a popular finish. This is a very rich, durable finish, However it poses a wide range of concerns. The traditional application follows the following mnemonic: Once a day for a week, once a week for a month, once a month for a year and yearly thereafter. In order to get even a reasonable finish, one needs several months. Additionally, the rags are notorious for bursting into flame of their own accord. A careless rag tossed in the trash can could burn down one's home. For these reasons, I chose another finish.

Other European Oils: Walnut, poppy, and safflower oils will all dry and harden, and thus could be used as a finish. Of the three, only walnut appears to have been used regularly, and in fact is still used today, mainly as a finish for wooden salad bowls. That said, they all have a very long drying time, and some (albeit minor) chance of spontaneous combustion.

Beeswax: Somewhat waterproof, but expensive (in period) and can be sticky.

Tung Oil: Oil pressed from the nut of the Tung tree. Not period for Europe (though period for Eastern Asia). Produces a finish in-between that of walnut and Linseed. The primary benefits are that it dries much faster than Linseed (2-3 days will net a glossy, well penetrated finish, as compared to months for linseed) It also is available pre-thinned and catalyzed at most home improvement stores. This is the finish eventually chosen for the project.

Varnish: A term for a combination of oil, resin and drying compound. Is likely period, but I have no evidence of it.

Shellac: The resin-like secretion of an Asian Insect. Wasn't known to the Europeans until after Marco Polo's voyages, which post-dates the Anglo-Saxon's by several centuries.

Lacquer: a term for 2 types of finish: one is very period for the far east (being derived from The Lacquer tree -Toxicodendron vernicifluum- a botanical relative of (with similar effects to) Poison Oak!), the other is a wide class modern synthetic analogs. Neither is appropriate for a 7th century reproduction.

Polyurethane: A product of the late 20th century.

Part 7: Brass fittings

Overview: The original had two riveted gilt brass fittings either holding together or re-enforcing the mortise joint. (see figures 3 & 5). they were inset with semi-precious stones and most likely had been mercury-gilt. They may have been stamped, physically etched or cast. For this reproduction, un-gilt brass fittings with etching and enamel work were selected due to the danger/ illegality of mercury gilding, as well as the costs of gemstones and the time involved in chemically etching versus casting, repousse, stamping or physical etching. Additionally, the original design is what scholars sometimes call "elephantine birds" and knotwork. In this reproduction, I decided to forgo the terribly stylistic for a more realistic elephantine animal- an actual pachyderm (see below). Brass was secured from a local discount store in the form of a solid brass "mail-slot".


Cutting/ Prep: To size the fittings, an image manipulation program was used to clip an image of the fitting from the picture of the original find (Figures 3 and 19 ) and re-size and print the image to the needed size (~1.3 wide) This was then used as a template to cut the brass. Cutting was done with shears and the edges were cleaned up with a file (Figure 20). Next, a similar printout of the authors original design (Fig. 21) was made using a laser printer on transparency film. This was then ironed onto the brass to transfer the toner, providing a resist for the chemical etchant.

Figure 19: The original fitting, cropped from Figure 3.

Figure 20: In progress cutting- the center object is a print out of Fig. 19, the other two are brass in the process of cutting
Image (c)2007 by Author (cc)

Figure 21: Design to be etched
Image (c)2007 by Author (cc)


Etching: The toner transfer had inadequate coverage, therefore a more traditional approach had to be be used. Paint was applied to both sides. The toner application did help, as it provided guidelines for the application of the front. After applying the resist (Figure 22), Etchant was prepared. Typically, one would use a commercial PCB etchant containing Ferric chloride. However, the author was unable to find said reagent through typical channels. Thus something else was needed. The final etchant used was a mixture of 2 parts (Secret ingredient 1) and one part (Secret ingredient 2).*** This is a very effective etchant, and in a short amount of time it produced a deep cut (Figure 23). After removing the pieces from the bath and neutralizing the etchant, the resist was removed with acetone (Figure 24). Then a contrast of thinned black in was applied to bring out the etching and enamel was applied. Finally, a clear coat was applied to prevent tarnishing (Figure 25).

Figure 22: Resist applied
Image (c)2007 by Author (cc)

Figure 23: Etching fittings. The feather is used to remove bubbles that form during the process
Image (c)2007 by Author (cc)

Figure 24: Etched fitting
Image (c)2007 by Author (cc)

Figure 25: After etching and contrast have been applied
Image (c)2007 by Author (cc)



Part 8: Strings and Tailgut

Overview: The original was either strung with gut or horse hair, with gut being most likely. The tailgut may have been any number of things, though it would be a safe assumption to conclude that the whatever was used wasn't very elastic. Based upon those criteria, Rawhide was selected for the tailgut, and modern nylon strings for the strings [originally D'addairio. Strings of the following gauges: 1.04mm, 1.02mm, 0.99mm (X2)and 0.97mm (X2), but now strings of gauges 1.024mm (X2), 0.818 mm(X2) and 0.711 mm (X2)]

Procedure: The tailgut was made from bovine rawhide recovered from a "Chew toy" for dogs. A 1/2" section was cut from the toy with a handsaw and allowed to soak in water for several days. Once it had become soft and pliable, scissors were used to remove any jagged edges. It was then fitted through the tailpiece, around the tailpiece and secured with a square knot. Upholstery thread laced through the tailpiece and around 2 pegs was then used to provide tension during drying (Figure 26). After the tailgut had dried, the lyre was strung, using Figure 8 knots as stoppers.

Figure 26: Stretching the tailgut. Brass fittings have not been attached at this point.
Image (c)2007 by Author (cc)


Part 9: Tuning

Overview: We have no idea how the original was tuned. Several theories exist, one of the more likely being a type of pentatonic tuning. Both Benjamin Bagby (WEB: Click for article) and Master Orrick of Romney (unknown mundane name) have suggested scales. Master Orrick suggests D-E-G-A-B-D (or moving up a note or 2 while keeping the same progression), while Bagby posits three possibilities: F-G-A-C-D-F (his "open" method), D-E-G-A-B-D (his "centered" method, the same as Master Orrick) or the Method of Hucbald of St. Amand: D-E-F#-G-A-B.

Procedure: The process is actually very simple- a microphone was attached to the soundboard and the given string plucked. The resultant note was analyzed by a software package (In-tune Multi-Instrument Tuner, Aspire Software) The pegs were manipulated to get the ideal tone. The first stringing was tuned to the "centered" method in E (thus E3-G3-A4-B4-D4-E4), while the second was tuned in the "open method" in F (F3-G3-A4-C4-D4-F4) . Most recently, I have tuned it to C4-D4-F4-G4-A5-C5

Note that the tinny sound of the high E is an artifact of the recording/encoding methodology
Audio: (c)2007 by Author (cc)

Audio: (c)2007 by Author (cc)

Audio: (c)2007 by Author (cc)

Audio: (c)2007 by Author (cc)



Part 10: Final Artifact

Figure 27: Front view
Image (c)2008 by Author (cc)
Figure 28: Rear View
Image (c)2008 by Author (cc)
Figure 29: Side View
Image (c)2008 by Author (cc)


Part 11 :Plectrum

Overview: There is some debate over whether the Lyre was plucked, like a frame harp, or strummed and blocked (analogous to guitar playing, but without fretting, a limited # of "chords"). The image of King David seem to suggest strumming and blocking, so the Author figured it would be a good idea to have a plectrum (colloquially, a "pick") so one was cut from horn and polished.

Procedure: A thin piece of horn was cut to shape with utility shears, formed with manicure tools and polished with successively fine emery boards. Final polishing used a commercially available polishing block for fingernails.

Figure 30: Horn Plectrum. Bandage covers stitches incurred during construction. (WEB: click for grisly image of stitches)
Image (c)2008 by Author (cc)


APPENDIX: Here are 2 videos of someone more skilled than I playing a similar Lyre

Youtube Video: Michael J King plays a Saxon Lyre via plucking
Video (c)?? 2006 by Michael J. King, Shared via YouTube

Youtube Video: Michael J King plays a Saxon Lyre via strumming and blocking
Video (c)?? 2006 by Michael J. King, Shared via YouTube




* Images marked (cc) Are Creative commons Licensed images. Other images covered by their individual licenses/ copyrights

** For a good overview of the finds: note that most of the links to the finds have moved, and thus will be unavailable for viewing.

*** While I am a broke graduate student, I still don't want to be sued. I'm sure if you wanted to etch cupric metals, you could Google "PCB etchant" and find what you need, as well as a potential target for your lawsuit.