How Were the Edges of Medieval Shields Treated?

A Note on the Evidence

The surviving high-medieval evidence does not support treating a thick leather, rawhide, or metal rim as a universal feature of the medieval shield.

The perimeter is one of the most easily over-reconstructed parts of a medieval shield, so this article separates several kinds of evidence with particular care. A surviving turnover documented by conservation, a historical catalogue description, a plausible reconstruction and an open unknown are not the same kind of claim, even when they describe the same word: rim.

The labels describe the evidence behind a publication, not my access to it: where I know an examination only through a later summary, the text names the summary.

Lenzen: An Organic Perimeter in the Eleventh Century

One of the most unusual pieces of evidence comes from the remains of an oval West Slavic shield discovered at Lenzen, generally dated to the second half of the eleventh century. The shield was constructed from alder and survives only incompletely, but its perimeter is particularly important.

Around the edge was a woven organic structure approximately 10 cm wide. The exact botanical identification remains uncertain: different interpretations of the material have included bast, grasses, and possibly bark-derived material. The important fact is therefore not the precise plant species, which remains disputed, but the construction itself.

The Lenzen shield demonstrates directly that an eleventh-century shield perimeter could be treated with a substantial woven organic reinforcement. This is very different from the familiar modern reconstruction of a strip of leather folded neatly over the edge.

There was more than one medieval solution to protecting a wooden shield perimeter.

The Lenzen material should not be transformed into a universal eleventh-century construction method. It is one surviving object. But it proves that medieval shieldmakers could employ complex organic perimeter structures that would rarely survive archaeologically. That makes Lenzen one of the most important shields in this evidence trail.

Szczecin: What Remains Unknown Matters Too

The late-twelfth-century shield finds from Szczecin provide an equally useful lesson, although for almost the opposite reason. The unpainted kite shield from Szczecin, dating broadly to the late twelfth century, survives as two adjoining alder planks.

It is tempting to use the surviving object to answer the edge question. My review of the available evidence, however, found no secure object-specific evidence establishing a leather binding, rawhide rim, metal edge, nailed strip, or other perimeter reinforcement for this particular shield.

This distinction matters because leather and rawhide occur in discussions of the Szczecin material more generally. That does not permit evidence associated with another object, or with the wider publication as a whole, to be transferred automatically to this shield.

Historical reconstruction becomes unreliable when "not preserved" or "not reported" is silently changed into "probably rawhide."

The Twelfth Century Offers No Standard Rim

Taken together, Lenzen and Szczecin already undermine the idea of a single medieval edge treatment. Lenzen provides positive evidence for an organic perimeter structure. Szczecin does not currently provide sufficient object-specific evidence to reconstruct one confidently.

The responsible conclusion is therefore not that eleventh- and twelfth-century shields had a particular standard rim. It is that multiple construction traditions probably existed, while preservation heavily favours wood and metal over organic edge treatments.

This preservation problem follows into the thirteenth century. Fortunately, several later shields preserve their coverings well enough to show something extremely important: the "rim" was not necessarily a separate component at all.

Seedorf: The Covering Becomes the Edge

The shield traditionally associated with Arnold von Brienz at Seedorf, dated 1200–1220 in the Swiss federal inventory and between 1180 and the first quarter of the thirteenth century in the bibliography of its conservation report, provides some of the most valuable evidence in the entire group. Its front is covered with parchment.

Kohlmorgen, summarising the conservation investigation by Senn and Moser (1991/92), describes a single piece of parchment that covers the front and is carried tightly but irregularly around both side edges and around the upper edge.

For the rear, the sources differ. Rahn, examining the shield in 1878, recorded a heavily torn leather covering there. The later conservation-derived account describes a second piece of parchment filling the centre of the rear, beneath the rear preparation. The rear also carried leather straps and padding, which are a separate matter. How the parchment and the leather covering relate, and which is original, cannot be settled without the unpublished conservation records. The edge evidence concerns the front parchment, which turns around the perimeter.

This is enormously important for reconstructing medieval shield edges. There is no need here to imagine a separate thick rim protecting the wooden core. The front covering itself wraps around the perimeter. The edge therefore forms part of the overall composite construction:

• Wooden core → covering → turnover around edge → rear termination.
• Rather than: wooden core → front covering → separate protective rim.

The Seedorf evidence is even more valuable because the upper edge has been involved in a long-running debate about the shield's original shape. Older scholarship proposed that the shield had once been a longer "Norman" or kite shield and was subsequently cut down. Later conservation-derived evidence challenges this interpretation.

The present upper edge reportedly displays the same manner of parchment working and attachment as the side edges. Neither the parchment nor the associated painted surface reportedly shows the disruption that would be expected if a substantial upper section had been sawn away and the covering subsequently rearranged.

Whatever conclusion is ultimately drawn about the shield's complete history, the perimeter evidence itself is significant. Seedorf gives strong support for covering turnover as an edge-treatment system. It does not provide evidence for a separate rawhide rim.

Konrad von Thüringen: Covering Turnover Without a Separate Rim

The shield associated with Landgrave Konrad von Thüringen, from the Elisabethkirche in Marburg and dated around 1230–1240, is the earliest of the Marburg shields discussed here, and it strengthens the evidence considerably. Kohlmorgen describes its parchment as turned around the left edge and narrowly over the top onto the rear, and a passage from Warnecke (1884), quoted by a later writer, says the same of its parchment-leather covering.

The precise construction still has uncertainties, but one conclusion is particularly important: no separate nailed leather or rawhide edge binding has been established.

Again, the shield does not require a separate rim to be invented. The surviving covering itself provides the described perimeter treatment.

Nordeck zu Rabenau: Another Turned Edge

The shield associated with Nordeck zu Rabenau, dated 1250–1300 by Kohlmorgen and "1300?" in the museum's own record, provides another useful piece of evidence. Its covering has been described as parchment or leather, and surviving material at the left edge appears to turn around the perimeter.

Unfortunately, the preservation does not allow the entire edge system to be reconstructed confidently. There is evidence for turnover at one part of the perimeter, but not enough evidence to claim that the exact same arrangement continued around every edge.

Kohlmorgen also reports a narrow strip on the rear along the upper edge, about 5 mm thick and 7 mm wide and half let into the wood, which he reads as a reinforcement. Among the shields I have researched, it is the only separate edge element reported on a Marburg shield. No conservation report has established whether it is original or exactly what it did.

This is an important distinction. It would be easy to reconstruct the missing perimeter by symmetry. Archaeologically, however, that would convert a plausible reconstruction into supposed fact.

Even this limited evidence reinforces what Seedorf suggests. The shield facing itself could serve as part of the edge treatment.

W. von Liederbach: Turnover Without Evidence for a Separate Rim

The shield associated with W. von Liederbach, dated 1250–1300 by Kohlmorgen (the church inventory places the original in the fourteenth century), continues the same pattern. Its front covering is parchment, while the rear is also reported to carry parchment. At the perimeter, the front parchment turns around the edge. The turnover onto the rear is somewhat irregular.

Warnecke (1884) recorded a brass strip let in near a board joint at the upper left edge, and an iron strip at the left corner. Kohlmorgen reports that neither is present now, and interprets both as probably later repairs made to stabilise a break.

Once again, research found no secure evidence for a separate rim placed over this construction. The perimeter therefore appears to have been created primarily through the covering system itself.

This repeated evidence is significant. Seedorf is not standing alone. Across several surviving high-medieval shields, the covering and the edge are structurally connected.

The Counts of Arnsberg Shield: A Narrow Turn onto the Rear

The shield associated with the Counts of Arnsberg, dated 1250–1300 by Kohlmorgen though the church inventory places the original in the fourteenth century, provides another particularly useful example. Kohlmorgen describes it as covered on both faces with a thin parchment-like leather, and the front covering as turning narrowly around the edge onto the rear face. Again, no independent separate rim has been securely established.

The width of the turnover varied. At Arnsberg the front covering is reported to turn only narrowly onto the rear; at Welfenberg-Helfenberg, discussed below, about a hand-width. Neither description reports a separate band over the turnover.

The edge may have been treated as a junction between surfaces, rather than as a separate armoured component.

Heinrich's Lion Shield at Marburg: An Important Unknown

The so-called Heinrich lion shield at Marburg, dated 1292–1308 by its attribution to Landgrave Heinrich I, preserves another valuable high-medieval composite construction. Its wooden core consists of several linden boards, and the shield carries parchment on both faces and painted heraldic decoration.

But despite the survival of the covering, research has not yet established the exact geometry of its termination at the perimeter. Did the parchment stop at the edge? Did it turn onto the rear? How far? Was it secured mechanically? Was another material added over it? At present these questions remain unresolved.

Kohlmorgen does describe its edges as strongly rounded on both faces, so the wood itself was shaped at the perimeter.

That makes the Marburg shield useful precisely because it prevents generalising too quickly from Seedorf. Parchment-covered shields existed. Parchment turnover around the perimeter existed. But it cannot therefore be assumed that every parchment-covered shield used exactly the same edge construction.

Welfenberg-Helfenberg: The Clearest Construction Evidence

The Welfenberg-Helfenberg shield, dated around 1300, provides perhaps the most technically useful evidence of all. It is covered with parchment on both faces, and here the perimeter treatment is unusually well described.

This gives something much closer to an actual construction sequence. The shieldmaker covered the face. At the perimeter, the covering was folded around the edge. Where necessary, the material was cut or arranged so that it could negotiate the shield's geometry. The resulting turnover was then mechanically secured with nails.

This is very different from attaching a separate leather rim after completing the shield surface. The edge is an integral part of the facing operation.

For reconstruction purposes, this is extraordinarily valuable because it shows not merely that parchment could wrap around a shield. It records a practical medieval method for doing so.

By this point in the chronological evidence trail, a pattern has emerged. Several high-medieval shields show or suggest front covering → around wooden edge → onto rear, rather than front covering → wooden edge → separate leather/rawhide rim. That difference has major consequences for reconstruction.

Other Thirteenth- and Fourteenth-Century Shields: Evidence Ceilings

Several other shields I have researched reinforce the need for caution.

• The Vögte von Keseberg shield, dated around 1250, preserves parchment on both front and rear, but the exact perimeter system remains insufficiently resolved.
• The Stromberg shield, of around 1300, is covered on each face with one piece of parchment in Kohlmorgen's description; some later shield lists call it white-tanned leather. Its core thins toward the edges, but its exact edge treatment remains unknown.
• The Valère eagle shield, of around 1300, likewise does not currently provide enough evidence to reconstruct its edge treatment confidently.
• The Steinau-Steinrück shield, of around 1300–1320, is described in later shield lists as covered with white-tanned leather, yet its perimeter remains effectively unknown.
• The older and younger Schenk von Schweinsberg shields, of around 1300–1320 and around 1350, preserve important evidence for their coverings and painted surfaces, but their perimeter arrangements remain incomplete in the available documentation.

The Sion Shields

The surviving shields associated with Sion and Valère add another complex group to the evidence. Some preserve multiple textile layers and evidence of later repainting or reuse. For the Bartolomei/Sion shield, whose core dates to the mid-fourteenth century, an original linen covering survives beneath a later painted textile layer, added when the shield was reused in the later fifteenth century. On the related Sion Cathedral Chapter shield, painted in the first half of the fifteenth century, the canvas bearing the Chapter's arms overlies earlier textile material.

These shields demonstrate how profoundly a shield could be modified during its lifetime. Their exact edge systems remain only partially understood. That uncertainty itself is useful because it is a reminder that the perimeter visible today may belong to more than one historical phase. An edge could be recovered. A damaged covering could be replaced. A shield could be repainted. New textile could be applied over an older surface.

The surviving perimeter of a medieval shield must therefore sometimes be read as a sequence of layers from different dates, not simply as a construction detail.

Herren von Muschenheim and the Fourteenth Century

The Herren von Muschenheim shield, dated to the fourteenth century in the church inventory and around 1350 in the specialist literature, preserves a linden core and parchment covering. Its edge treatment, however, remains only partially resolved.

Once again, this is significant. By the fourteenth century the evidence still does not force a universal heavy perimeter binding. Instead, the shields I have researched continue to show a mixture of described turnovers, incompletely preserved edges and unresolved constructions. The historical picture therefore remains diverse.

The Oña Shields: An Early Shape, a Later Surface

Two shields in the Real Armería in Madrid, D 59 and D 60, come from the monastery of San Salvador de Oña. Their shapes belong to an early type. Soler's 1996 technical study reads the surviving decoration and its techniques as late medieval. He considers that older cores might have been re-covered, but finds it more likely that the shields were newly made, perhaps after an older model, for the rebuilding of the cloister in 1503–1508, probably for display. That remains a hypothesis, and when each part of the shields was made has not been established.

Each shield is covered with two pieces of parchment. Soler does not describe how the parchment is finished at the perimeter, so these shields are not evidence for thirteenth-century edges, and their own edge treatment remains unknown. This is especially important because a parchment-covered shield can tempt the reconstructor to assume that the parchment must have wrapped around the perimeter in exactly the manner described at Welfenberg-Helfenberg.

It may have done so. But "may have" is not the same as "did." The evidence from one surviving shield should guide interpretation of another, but it should not replace the missing evidence.

What Does the Chronological Evidence Actually Show?

Once the surviving shields are placed beside one another, several conclusions become much clearer.

The first is that organic edge treatment has deep roots. Lenzen demonstrates a substantial woven organic perimeter in the eleventh century. But that construction should not automatically be transformed into a leather or rawhide strip, nor projected unchanged into later centuries.

The second major development visible in the surviving high-medieval shields is the integration of the edge into the shield's overall covering system. Seedorf provides evidence for parchment carried around the sides and upper edge. Konrad von Thüringen provides further evidence for perimeter turnover without an established separate rim. Nordeck zu Rabenau preserves turnover at least along the left edge. W. von Liederbach again preserves front parchment carried around the perimeter. The Counts of Arnsberg shield preserves a narrow turnover of its parchment-like leather onto the rear. Welfenberg-Helfenberg provides especially strong construction evidence: parchment turned around the edge, mitred and nailed.

This is no longer an isolated observation. It is a recurring pattern in the published descriptions.

One caution applies to all of it. Apart from Seedorf, where a modern conservation investigation lies behind the description, these turnovers are known mainly from Jan Kohlmorgen's published descriptions of the Marburg shields, with Warnecke and Hefner-Alteneck behind some of them. I have not found a modern conservation report that checks them on any Marburg shield.

The Medieval "Rim" May Be the Wrong Mental Model

This evidence suggests that the word rim can sometimes distort the understanding of high-medieval shields. A rim sounds like a separate component. It suggests something added around an already completed shield. But several surviving shields instead point toward an integrated construction:

• Wooden core → front covering → covering carried across the perimeter → turnover onto rear → glued, nailed, otherwise secured, or combined with rear covering.

In this system there may be no separate "rim" at all. The edge is simply the place where the shield's composite skin changes direction.

As a maker, I think this construction could also make mechanical sense. A covering bonded to the front could help hold the wooden boards together. Carrying that covering around the perimeter may have protected the vulnerable ends of the boards. Turning it onto the rear could make it harder to peel away from the face, and nails could reinforce the turnover where necessary. These are my engineering interpretations, not functions demonstrated on the originals.

Where Does Rawhide Fit?

Rawhide remains a perfectly plausible shieldmaking material. When soaked, it becomes flexible and can be formed around complex shapes. As it dries it contracts and becomes hard. These properties make it an attractive edge material for a modern maker.

But the distinction between plausibility and evidence is essential. The fact that rawhide works well does not prove that every medieval shield had a rawhide rim. Indeed, the strongest surviving high-medieval evidence among the shields I have researched points instead toward turnover of the shield's existing parchment or leather covering.

What About Leather Rims?

The same principle applies to tanned leather. Leather is historically appropriate, workable and mechanically useful. But a thick separate leather strip stitched or nailed around every heater shield is not what the shields I have researched show.

Some shields certainly employ leather or parchment-like leather as covering material. The Counts of Arnsberg shield is particularly useful here. But again, its described perimeter treatment is the turnover of that covering, not a clearly separate heavy leather rim.

The distinction is small visually but enormous archaeologically.

What About Metal Rims?

Metal requires even greater caution. Among the shields I have researched, the only metal recorded at an edge is on the Liederbach shield: small, local strips that Kohlmorgen interprets as probable later repairs. Local metal reinforcement is plausible. But small edge fittings, clips or repairs should not automatically be interpreted as fragments of a continuous iron border.

A complete metal channel around a large kite or heater shield is a very specific construction claim. It adds weight. It requires substantial metalworking. It changes the balance of the shield. It also produces an obvious structural and visual element.

Absence of Evidence Must Remain Absence of Evidence

Perhaps the most important lesson from this research is methodological. For several shields something concrete can be said about the perimeter. For others it cannot. Those unknown cases must remain unknown.

• If the evidence for a shield leaves the edge treatment unresolved, that does not mean "probably rawhide."
• If no nails survive, it does not mean "probably stitched."
• If no organic material survives, it does not mean "the edge was bare."
• If parchment covers the front, it does not automatically mean "the parchment wrapped around the edge."

Each of those may be reasonable hypotheses. None is automatically evidence.

A Chronological Evidence Trail

The surviving shields therefore give a useful, if incomplete, sequence.

The gaps in that sequence are as important as the positive evidence.

What Should a Historically Reconstructed Shield Use?

There is therefore no single correct answer. The question should never begin "What did medieval shields use for rims?" It should begin "Which shield, from which place and period, is being reconstructed?"

A reconstruction can still be made in such circumstances. But the reconstructed element should be identified as interpretation rather than archaeological fact.

Conclusion

The edge of a medieval shield was not necessarily a separate rim. That is perhaps the most important conclusion to emerge from the shields I have researched.

The eleventh-century Lenzen remains demonstrate that substantial organic perimeter reinforcement existed. But the high-medieval shields reveal another important construction tradition. At Seedorf, parchment turns around the perimeter. On the Konrad von Thüringen shield, parchment again turns around the edge. At Nordeck zu Rabenau, turnover survives at least along one edge. W. von Liederbach repeats the turnover principle. The Counts of Arnsberg shield shows its thin parchment-like leather covering carried narrowly onto the rear. At Welfenberg-Helfenberg, the construction becomes exceptionally clear: the front parchment is carried around the perimeter, mitred and nailed.

Alongside them stand numerous shields for which the edge remains uncertain. Together, they produce a picture far more interesting than the generic "leather-rimmed medieval shield."

Medieval shieldmakers could treat the perimeter as part of the complete composite structure of the shield. The wooden core, covering, edge and rear were not necessarily separate systems. They could form one continuous construction. And sometimes the archaeological record simply does not show how the edge was finished. That uncertainty should not be hidden.

Reconstruct what the individual object demonstrates, infer only where necessary, and never turn a mechanically convenient modern solution into a universal medieval practice.