How Did Medieval Shield Core Construction Develop?

The Short Answer

The surviving shields do not share one core recipe. Most are single layers of boards set side by side, but two are cross-laminated, and wood, board count and thickness vary throughout.

Linden is the wood most often recorded, because the largest group of surviving shields, from Marburg, is mostly linden. Outside that group the record includes alder, willow, poplar and a disputed fir or spruce. Thickness does not rise or fall steadily with date, and the present curvature of a surviving shield is not a safe record of the curve its maker gave it.

A Note on the Evidence

Shields warp, decay, are repaired, re-covered and sometimes substantially altered. A present measurement is not automatically an original workshop specification, and a detail documented on one shield is not transferred to another. Five kinds of statement are kept apart below.

Many of the figures below come from Jan Kohlmorgen's 2002 survey of the rider shield, which describes the Marburg shields one by one. His board counts and thicknesses are careful descriptions of what can be seen, not modern conservation measurements, and they are labelled accordingly.

Lenzen, Second Half of the 11th Century: Two Crossed Layers

The fragmentary shield from the Lenzen hillfort in Brandenburg is the earliest construction in this comparison. Its date comes from the archaeological context, not from the wood itself.

Lenzen matters because it puts a cross-laminated core at the beginning of the sequence. Medieval shield construction did not climb a ladder from simple planks to laminated wood: the more complex arrangement is already present in the 11th century.

It is still not plywood. Modern plywood is made from thin peeled veneers, an odd number of plies and industrial adhesives. "Cross-laminated core" describes what Lenzen is without importing those features.

Szczecin, Late 12th Century: Thin Alder Boards

Two shields from Szczecin, both of alder, are separate objects and their figures are not merged: the painted shield, catalogued as MNS/A/22250/1 and dated by the museum to 1176–1200, and a much more fragmentary shield conventionally called "unpainted," dated more broadly to the late 12th century. The museum catalogue gives the painted shield six boards.

The two shields show that a large late-12th-century shield did not need a thick slab: a core of thin alder boards could carry it. They also show, from the start, that multi-board construction did not mean a standard board count or thickness. Their square board edges and the open question of their glue are discussed in the planks-joined article.

Seedorf, c.1200–1220: Four Alder Boards, Once Thought Linden

The Seedorf shield, traditionally associated with Arnold von Brienz and dated 1200–1220 by the Swiss National Museum, shows how a wood identification can change.

Boeheim's "probably linden" was a reasonable guess for a shield of this kind; the later examination replaced it. Old species names in catalogues and descriptions deserve the same caution wherever no examination has followed them.

Marburg, 13th Century: Linden Cores with Different Board Counts

The shields that hung in St Elisabeth's Church in Marburg, now in the museum of the Philipps-Universität, are the largest related group of surviving shields. Because several survive together, they show variation within one place instead of letting one shield stand for a period.

Within one related group, then, linden recurs, but board count runs from five to seven or eight (Konrad's, an estimate) and thickness from about 8 to 12 mm. A "Marburg shield" is not a single specification.

Around 1300: From Two Boards to Six, 7 mm to 15 mm

The Marburg shields of around 1300 widen the range further. All are linden where the wood is recorded, and all the figures below are Kohlmorgen's unless stated otherwise.

• The lion shield that Kohlmorgen identifies as Heinrich I of Hesse's, 1292–1308 if that attribution is right, is described as five or six boards about 15 mm thick. A conflicting identification as Heinrich "der Junker" remains unresolved.
• The Nordeck zu Rabenau shield, conventionally late 13th century or c.1300, has a core of three linden boards, about 10 mm in Kohlmorgen and about 8 mm in Warzecha's later measurement. The sources word the orientation of the three pieces differently, so a definite board diagram would go beyond the evidence.
• The Stromberg shield, conventionally c.1300, is described as five boards about 12 mm thick, thinning toward the edges.
• The Steinau / Steinrück shield, conventionally c.1300–1320, has a multi-board core of about 12 mm; its board count has not been published.
• The Welfenberg / Helfenberg shield, c.1300, is about 9 mm thick, and Kohlmorgen notes that its grain runs diagonally across the shield instead of vertically. Its board count is not recorded.
• The older Schenk von Schweinsberg shield, conventionally 1300–1320, has only two boards, about 7 mm thick.

Two shields of about the same date, the older Schenk shield at about 7 mm and the Heinrich shield at about 15 mm, differ in thickness by a factor of two. Thickness alone cannot date a shield.

The eagle shield MV 172 at Sion, also around 1300, was described by Laking as built of wood, leather and canvas; I have found no published wood species, board count or thickness for it.

One of these shields also has a reported wooden fitting at its seams: Kohlmorgen records the older Schenk shield's glue edges as dowelled. That evidence, and its limits, belong to the planks-joined article.

Karl von Trier, c.1320: Fir or Spruce, Four Boards, Bent-Back Edges

The shield associated with Karl von Trier, Grand Master of the Teutonic Order from 1311 to 1324, is not one of the Marburg shields. It came from Burg Reifenstein in South Tyrol and is now in the Tiroler Landesmuseen in Innsbruck. Its date of around 1320 rests on its painted heraldry, and it is described in the heraldic literature as a processional or display shield, which matters when comparing it with shields made for the field.

Either way, the shield is a softwood core among linden, alder, willow and poplar ones, and it shows that the Marburg linden should not be taken for the medieval default. Engel's tenons are a second reported case of wooden fittings at a shield seam, discussed in the planks-joined article. The central seam has since opened; that is later damage, not evidence of how the joint was made.

The Mid-14th Century: Thin and Thick Linden, and Willow at Sion

The later Marburg shields continue to differ from one another. Kohlmorgen describes the Muschenheim shield, conventionally c.1350, as three linden boards about 8 mm thick. The younger Schenk von Schweinsberg shield, also c.1350, has a linden core about 15 mm thick whose board arrangement has not been described; the older shield's two boards and dowels are not transferred to it.

Two further Marburg originals, the unidentified lion shield (Iv-Nr 16) and the shield traditionally labelled von Dernbach (Iv-Nr 9), have no published description of their wood, board count or thickness that I have found. The neighbouring shields' figures are not a reason to fill those gaps.

Bartolomei is the second cross-laminated core in this comparison, three centuries after Lenzen, in a different region and a different wood. Nothing connects the two in a documented tradition, but together they show that crossed layers were not a single isolated experiment.

The Later 14th and 15th Centuries: Canterbury and the Sion Pavois

The Canterbury shield associated with Edward of Woodstock, the Black Prince, belongs to his funerary achievements of 1376; that association does not prove it was made in that year.

The large Sion Cathedral Chapter pavois, MV 1, is stylistically assigned to the first half of the 15th century. The museum records it as poplar covered with canvas, with a Chapter-painted textile over an earlier one. Its board count, board direction and thickness are not established, and its layered coverings do not make it a layered core.

In both shields the wood is only one part of a multi-layer body, and in both that body hides the wooden core from view. What can be said about their cores is correspondingly brief.

The Oña Shields: Early Shapes, Possibly Later Objects

Two shields from the monastery of San Salvador de Oña, now in the Real Armería in Madrid as D 59 and D 60, have shapes of around 1200 and were long dated accordingly. Álvaro Soler del Campo's 1996 technical study, made while both were under restoration, dates the objects much later. They close this comparison because their date is uncertain, not because they are the latest securely dated shields.

The date of around 1190–1200 sometimes still given for D 59 belongs to its shape. The "cedar-like" wood comes from the Real Armería catalogue of 1898, not from analysis. The unequal boards are real construction evidence whichever date is right, but they cannot stand for a shield made around 1200.

The Surviving Cores Side by Side

Figures are approximate values from the sources named in the text, not standardised measurements. "Not published" means not established in the published evidence, not necessarily absent from the object.

Wood Species: Linden Is Common, Not Universal

Linden dominates the table because the Marburg group dominates the surviving shields, and the Marburg shields are mostly linden. Outside Marburg the record includes alder at Lenzen, Szczecin and Seedorf, fir or spruce at Karl von Trier, willow in the Bartolomei shield and poplar in the Sion pavois. The Black Prince's "poplar" and Oña D 59's "cedar-like" wood are later or older descriptions, not identifications.

Seedorf is a reminder that species names can be wrong: described as probably linden in 1897, it proved to be alder when it was examined. Where a species rests only on an old catalogue, it is best treated as a description.

Board Count and Thickness: No Steady Development

Where single-layer cores have a published board count, it runs from two (the older Schenk shield) through three (Nordeck, whose three pieces are described differently in their arrangement; Muschenheim; Oña D 60), four (Seedorf, Karl von Trier), five (Stromberg, Arnsberg) and six (Keseberg, Liederbach, Oña D 59, and the painted Szczecin shield in two accounts) to seven (the painted Szczecin shield in the third) and seven or eight (Konrad, an estimate). Both Oña shields have boards of strongly unequal width.

Thickness runs from about 4 mm at the thinnest point of the painted Szczecin shield to about 15 mm in the Heinrich shield, Karl von Trier, the younger Schenk shield and the side boards of Oña D 59. Seedorf is left out of this range: Boeheim's 15 mm was measured on the covered shield, and the thickness of its bare boards is not known. Thin and thick cores coexist in every period with enough evidence: around 1300, the older Schenk shield at about 7 mm and the Heinrich shield at about 15 mm. Thickness is an object-by-object decision, not a dating formula.

Grain Direction and Cross-Lamination

In most single-layer cores the boards, and so the grain, run from top to bottom. Two kinds of exception are recorded. Lenzen and Bartolomei are cross-laminated: one layer's grain runs across the other's, an arrangement that, in woodworking generally, resists splitting along the grain better than a single layer. And Welfenberg, a single-layer core, has grain running diagonally across the shield.

The cross-laminated cores are the clearest evidence that grain direction could be a deliberate choice. Calling them plywood would mislead; denying medieval laminated construction would mislead as much.

What About Curvature?

The present shape of a surviving shield is the shape it has after six or eight centuries. Wood moves, coverings shrink, repairs change stresses, and display and storage leave their mark. A few cases show how differently present shape can be read.

• Engel described Karl von Trier as flat both lengthwise and across, with only the long edges bent back by about 45 degrees, which he judged intentional. A flat core with bent edges is not the evenly curved shield often assumed.
• Kohlmorgen gives present curvature depths of about 6.5 cm for the Heinrich shield and about 3 cm for Stromberg and Steinau, without saying how they were measured.
• The Arnsberg shield is now only slightly curved and strongly warped into an S-shape: its present profile is clearly not its original one.
• For Seedorf I have found no published curvature measurement at all.

What This History Means for Reconstruction

The useful question is not how a medieval shield core was made, but which shield is being reconstructed, and what its core is known to have been.

Where the original is not known, the choice made should be named as a choice. That is what separates a reconstruction of a particular shield from a generic medieval-style one.

Conclusion: A History of Solutions, Not a Single Recipe

From Lenzen in the 11th century to the Sion pavois in the 15th, the surviving cores differ in almost every respect that can be measured. Lenzen and Bartolomei are cross-laminated; the rest that are documented are single layers of two to seven boards, or eight if the higher estimate for Konrad is right. Linden is common because Marburg is, but alder, willow, poplar and fir or spruce are all recorded. Thickness runs from about 4 to about 15 mm without a chronological trend, and present curvature is a weak guide to the original shape. The Oña shields, long dated to around 1200 by their shape, are dated by Soler, tentatively, to the late Middle Ages, possibly made or reworked in the early 16th century.

The most historically accurate shield core is not the one that follows a universal medieval recipe. It is the one whose construction follows the evidence for the particular shield being reconstructed.