How Were Medieval Shields Painted? Paints, Pigments and the Medieval Shield Surface

A Note on Evidence

I separate direct object evidence, institutional or conservation evidence, specialist observation, historical description, reconstruction and unknown information. That distinction matters particularly for colour.

A shield that appears black today was not necessarily black when it was made. Pigments can darken or fade. Silver can tarnish almost to black. Organic layers disappear. Surfaces can be cleaned, repaired or repainted. Nineteenth-century descriptions may record details that have since been lost, but they are not chemical analyses.

A colour name is not, by itself, a pigment identification.

For that reason, names such as red, blue or yellow in this article do not automatically identify a pigment. A blue field is evidence for a blue surface. It is not, by itself, proof of azurite, ultramarine, indigo or another blue colourant.

The same caution applies to binders. Egg tempera is historically plausible in medieval painting, but it should not be assigned to a particular shield unless the evidence supports it.

Function matters too. Not every object here was made to be carried in a fight. The shields of Konrad von Thüringen and the Black Prince are funerary or memorial shields; the Karl von Trier shield is described as a Grand Master's processional shield; the large Sion Chapter object MV 1 is a pavise, a big standing shield; and the Oña shields were probably votive. Display objects were more likely to be finished lavishly, and more likely to be kept, than shields worn out in use. Much of the most colourful evidence in this article therefore comes from them, and it should not be read as a description of an ordinary fighting shield.

Much of what is known about the Marburg shields comes from one specialist, Jan Kohlmorgen, whose detailed descriptions were published in 2002. Where a Marburg claim below rests on him, I say so. His observations are valuable, but they have not been checked against modern conservation reports, and I have found no published pigment or binder analysis for any of those shields.

11th Century: Lenzen and the Earliest Colour in the Corpus

The earliest shield in my surviving-object corpus is the fragmentary shield from Lenzen in Brandenburg, associated with the second half of the 11th century.

Its importance to shield construction is considerable because its alder core is cross-laminated. Its importance to the history of painted shields is quieter, but still significant. Traces of red paint have been reported on the outside of the original.

That shows that colour formed part of the shield's visible surface. What it does not show is equally important.

Lenzen therefore provides an appropriate starting point. By the 11th century, a shield could carry a deliberately coloured exterior, but the surviving evidence is not sufficient to reconstruct a complete painting recipe.

The first lesson is restraint: colour can survive even when the original painting system does not.

Late 12th Century: The Painted Szczecin Shield

The painted shield from Szczecin, usually dated to the late 12th century, is one of the most important objects for understanding medieval shield painting because its surface has been examined as a sequence of layers.

Irena Jagielska's conservation study examined four paint samples under the microscope and by XRF. In the clearest cross-section the sequence runs, from the wood upwards: a black organic layer that penetrates the wood structure, a lime-white layer, and then vermilion. A second sample showed the same sequence with an iron-oxide red and orpiment as well. The binder was reported as oily or fatty.

This is unusually valuable evidence. Vermilion is a brilliant mercury sulphide red. The iron-oxide red, sometimes given as hematite, is an earth red. Orpiment is a strongly coloured arsenic sulphide yellow. XRF detects elements such as mercury, iron and arsenic, so the identifications rest on elemental analysis combined with microscopy. Even so, these are not hypothetical colours selected for a modern reconstruction. They were found on the surviving shield.

The museum describes red, yellow and black decoration painted on the planks. I have not found a published reconstruction of the complete design.

The function of the black organic layer is not established. It belongs to the paint sequence: it penetrates the wood and lies beneath the lime-white layer. No skin, parchment or textile layer was found between the wood and the paint in the sampled areas, and the museum describes the decoration as painted directly on the wooden planks.

What the samples do establish is the sequence:

• wood → black organic layer → lime-white layer → coloured paint

The colour was not simply brushed onto bare timber. The painter worked over a black layer and a pale preparation.

The second Szczecin shield is conventionally described as unpainted, but that label must be treated carefully. Its present condition does not prove that it never possessed a lost facing, preparation or painted surface. My dossier therefore treats "unpainted" as a descriptive modern name rather than evidence for an intentional medieval bare-wood finish.

Together, the two Szczecin objects warn against assuming either universal painting or universal bare wood.

Around 1200: Seedorf, Blue and Silver

The Seedorf / Arnold von Brienz shield, dated 1200–1220 in the Swiss federal inventory and from about 1180 to the first quarter of the 13th century in the conservation literature, moves the evidence from coloured paint into a more elaborate heraldic surface.

The Swiss National Museum reports that analysis found the shield's field was originally a bright blue, despite its much duller grey-green appearance today, and that the lion was covered with silver leaf. The analytical methods behind that result are not published in the sources available to me: the conservation report by Matthias Senn and Franz Moser cites laboratory reports that remain unpublished.

This distinction matters. Medieval heraldic "white" did not always have to be made with white paint. A metallic silver surface could perform the heraldic role of argent while producing a much more brilliant visual effect.

The shield also preserves evidence for a prepared surface associated with its parchment covering. An observer in 1878 recorded a chalk preparation on the front parchment; later conservation-derived accounts describe a gypsum or chalk and glue preparation applied in stages. The exact recipe of the ground and the precise blue pigment have not been recovered in the source chain accessible to me.

The important point is therefore not to guess the blue. It is to recognise the system. A shield around the beginning of the 13th century could combine:

• wooden core → parchment facing → preparation layer → coloured field → metallic heraldic decoration

Seedorf also demonstrates why modern appearance can be deceptive. The field visible today is not a reliable colour chart for the object as it left the medieval workshop.

The shield was once considerably more brilliant.

c. 1230–1240: Konrad von Thüringen and the Shield as a Painted Object

The shield associated with Konrad von Thüringen is among the richest surviving examples. Historical literature describes it as a funerary or memorial shield associated with his grave, so it shows what a high-status display shield could be, not necessarily what a fighting shield carried.

Its linden core was covered and prepared for decoration. The museum record gives linden, leather, chalk ground and polychromy. In the 19th century Jakob Heinrich von Hefner-Alteneck described the front as parchment-covered and painted blue, with a large lion of raised, pressed leather in alternating red and silver bands and the remains of a formerly gilded crown.

The reverse is perhaps even more remarkable. Hefner-Alteneck described it as parchment-covered, originally gilded all over, and painted with a figural composition. He recorded an armoured knight, a woman in red, green tendrils and red flowers.

This is far beyond the idea of a shield as a board carrying a simple two-colour coat of arms. It was a complex decorated object.

The museum record documents the chalk ground, and the lion is a separate raised element. Hefner-Alteneck also recorded a later repair: when the lion came loose it was crudely nailed back around its outline, and the nails driven through the shield damaged the gilding and paint on the reverse.

Konrad therefore shows several things at once: a pale ground prepared the surface, strong heraldic colour was combined with metallic decoration, relief could be incorporated into the design, and the reverse could receive expensive and elaborate decoration as well as the front.

c. 1250: Keseberg and a Recurring Covering and Ground

The Vögte von Keseberg shield, dated around 1250 by Kohlmorgen, is reported by him with one piece of parchment on each face, followed by chalk ground and paint. Its heraldry is a black field with two silver leopards, armed red.

With Seedorf and Konrad, Keseberg shows the same arrangement on three shields from the first half and middle of the 13th century: a skin or parchment covering, a pale mineral ground, then paint. Most of the later Marburg shields below are described in the same way. Three dated examples are not a rule, but they show that a pale preparation of the kind seen at Szczecin and Seedorf was not an isolated curiosity.

This had obvious advantages. The irregular grain and joints of a wooden core could be visually separated from the painted surface. A pale ground could create a more uniform base. Translucent or relatively weak colours could appear brighter against it. Fine heraldic edges and decorative details could be painted on a smoother surface than raw timber would provide.

Those advantages help explain the technique, but they are not evidence that every shield used it. The surviving objects show a recurring workshop solution, not a universal law.

c. 1250–1300: Arnsberg, Liederbach and Silver Under the Paint

The shield attributed since 1986 to the Counts of Arnsberg, dated 1250–1300 by Kohlmorgen, has thin parchment-like leather on both faces and a chalk ground. Its red field is built up in dotted, modelled relief about one and a half millimetres high. The eagle is in shallower relief and silvered, but Kohlmorgen notes that its finest parts, the wing feathers, legs and tail, were not silvered: they were painted in opaque white over the red.

The shield associated with W. von Liederbach, also dated 1250–1300, has chalk ground on both parchment faces. Kohlmorgen considers the front likely to have been silvered all over before it was painted. The half eagle on one side is the exposed silver, given black outlines and inner detail, with the red painted around it. The reverse carries two red chevrons on white, painted in red and white only, with no silvering reported.

Together they show that heraldic silver could be made in more than one way within the same group of shields: metal leaf, white paint, or both side by side.

The sources accessible to me provide no modern chemical identification of the pigments, binders or silvering materials on either shield.

Technique can sometimes be established more securely than chemistry.

It may be known that parchment was grounded and painted while it remains unknown whether a particular red was vermilion, red lead, an iron oxide or a mixture. An area may be known to have been silvered without a complete analysis of its mordant or protective coating.

The historical surface can therefore be reconstructed at different levels of confidence.

Around 1300: Relief, Gilding and Modelled Surfaces

Several Marburg shields dated around 1300 show that decoration could become physically three-dimensional. Their dates are conventional, based on form and heraldry rather than scientific dating.

Kohlmorgen describes the Welfenberg / Helfenberg shield with parchment on both faces and a chalk ground with modelled decoration. Its gilded background is damasked in a grid, each square containing a four-petalled flower. The heraldic charge rises about two millimetres above the surrounding surface, is bordered by a dotted band, and was painted directly on the chalk ground, without gilding beneath it.

The Stromberg shield belongs to the same broad technological world. Kohlmorgen and other specialist descriptions record parchment or skin covering, chalk ground, modelled decoration, silvering and paint. Some details of the exact covering material remain contradictory in the source tradition, but the existence of a deliberately built decorative surface is clear.

The Nordeck zu Rabenau shield, dated "1300?" in the museum record, is simpler: three black sea- or wood-lily leaves painted on a white chalk ground, which Kohlmorgen reports on the front only.

This changes how the medieval shield painter should be imagined. Painting was not always the final application of flat colour. The surface could first be built, modelled or textured. Metal leaf could be incorporated. Paint then interacted with a relief surface that caught light physically.

A shield could therefore occupy a space somewhere between heraldic painting, panel painting and low-relief decorative work.

c. 1300: The Valère Eagle Shield

The Valère eagle shield in Sion, conventionally dated around 1300, preserves an important heraldic surface, but its pigment evidence illustrates the opposite problem from Szczecin.

The object and its decoration survive, yet the accessible research did not recover a modern object-specific analysis identifying the pigments, binder, gilding materials or exact ground composition.

The correct response is not to borrow a recipe from Seedorf, Konrad or Szczecin. It is to leave the chemistry unknown.

c. 1292–1308: Heinrich I of Hesse and a Gilded, Modelled Surface

The shield associated with Heinrich I of Hesse, conventionally dated 1292–1308, preserves one of the most technically elaborate surfaces among the shields studied here.

Kohlmorgen reports that its front parchment carries a gilded chalk ground. Above this is a modelled heraldic animal formed from two layers of linen impregnated with chalk ground, pierced and worked into relief. Helmut Nickel described the same textile relief, stiffened with gesso, over a gilded background.

This is a striking piece of evidence. Textile was not merely a hidden reinforcement. In this case it participated directly in the decorative construction of the heraldic figure.

The surface was therefore assembled in stages:

• wood → covering → chalk ground → gilded surface → textile-and-ground relief → heraldic finishing

Again, this is closer to a complex painted and gilded artwork than to a simple piece of military equipment receiving a quick coat of colour. The exact pigment and binder chemistry remains unknown, but the physical construction of the decorative surface is itself highly informative.

c. 1300–1320: Schenk von Schweinsberg and Steinau

Kohlmorgen dates the older Schenk von Schweinsberg shield 1300–1320 and reports parchment on both faces, chalk ground and paint. Its arms are a blue upper field with a silver lion, above a lower field of red and silver lozenges. The silver lozenges are painted white, further patterned with white foliage on grey; they are not metal leaf.

The Steinau / Steinrück shield, conventionally dated c.1300–1320, is described in historical material as having a magnificently modelled surface with a golden ground, although gold-versus-silver descriptions require care because old metallic surfaces and historical terminology can be misleading.

These shields show a mature decorative practice involving pale grounds, strong heraldic colour, metallic effects and, on some objects, modelled surfaces.

But there was still no reason to expect a single standardized pigment box.

c. 1320: Karl von Trier

The shield associated with Karl von Trier, Grand Master of the Teutonic Order from 1311 to 1324, is in the Tiroler Landesmuseum Ferdinandeum in Innsbruck and is conventionally dated around 1320. It is described in the heraldic literature as a Grand Master's processional or display shield, and it provides a particularly vivid surviving colour scheme.

Kohlmorgen describes its front as having a yellow ground. Against it is a white Gothic or Manesse-form shield carrying the arms of the Grand Master: a black Order cross, a gold or yellow fleur-de-lis-ended cross, and a small gold or yellow central shield with a black eagle.

The paint surface is badly damaged. The central core joint has opened, tearing the covering, and large areas of paint have flaked away, especially toward the lower part of the shield.

This damage is instructive. A medieval shield surface was a layered structure attached to a moving organic support. Wood changes with humidity. Board joints move. Skin, parchment, textile, ground and paint respond differently. When the support deforms or a seam opens, brittle preparation and paint layers can crack and detach.

The surviving condition of Karl von Trier therefore helps explain why so little original medieval shield colour remains intact.

The disappearance of paint is not evidence that shields were visually plain.

14th Century: Muschenheim, Dernbach and the Bartolomei Shield

Kohlmorgen dates the Herren von Muschenheim shield around 1350; the current church register says 14th century. It belongs to the same parchment, chalk-ground and painted tradition, with a red field and a silver chevron whose technique has not been confirmed on the object.

The von Dernbach shield is dated only broadly, to the 14th century, and possibly into the 15th. Its arms are a blue field with three silver or white hearts, but whether the hearts are metal or white paint, and how much of the surface is original, has not been established.

The Bartolomei shield in Sion, MV 4, is a mid-14th-century knightly shield that was later re-covered and reused as a targe in the second half of the 15th century. The museum reports that a 1981 examination by Hugo Schneider found its original linen layer painted in a red-and-gold checkered pattern, with a central medallion of Saint Maurice on horseback, beneath the later Cathedral Chapter scheme. The two campaigns have to be kept apart.

c. 1376: The Black Prince's Funerary Shield

The shield among the funerary achievements of Edward of Woodstock, the Black Prince (died 1376), at Canterbury Cathedral is a funerary shield, not a fighting one.

Laking described its front as built up from a light wooden core, white canvas, gesso, paper and finally leather. Its heraldic surface includes lions and evidence of gilding, but the exact construction of the relief decoration and the complete pigment, binder and gilding system have not been published in the sources I have read.

Together with Konrad von Thüringen and Karl von Trier, it shows a shield that could be read at once as a heraldic identifier, a status object, a memorial and a work of decorative craft.

Early 15th Century: The Sion Chapter Pavise

The large pavise of the Sion Cathedral Chapter, MV 1, is conventionally dated to the first half of the 15th century on stylistic grounds. Its visible programme is particularly complex. Its central arms are surrounded by black foliage or damask on a yellow-green ground. An outer border uses a yellow quatrefoil or network pattern against black, enclosed by a white line. The object therefore preserves not merely heraldic tinctures but ornamental patterning around the arms.

Visual complexity does not bring chemical certainty.

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 cloister of the monastery of San Salvador de Oña near Burgos. Their shapes belong to types of the 12th century and around 1200, and older literature dated them accordingly. I place them here, after the 15th-century material, because the most detailed study of them, by Álvaro Soler del Campo (1996), dates their surfaces much later.

The Oña shields therefore show prepared, skin-covered surfaces outside the German and Swiss material, but as late-medieval, probably votive objects. They are not evidence for how shields were painted around 1200.

Function matters. A highly elaborate memorial, ceremonial or display shield should not automatically be treated as a direct specification for a working battlefield shield.

Metallic Heraldry and the Problem of Modern Appearance

Several shields in this article show that metallic heraldry can become visually misleading with age.

Silver tarnishes. Gilding wears away. Varnishes darken. Ground becomes exposed. Later retouching can alter the balance of colours.

The clearest case is the Oña shield D 59. Its front field looks black today. Soler del Campo's analysis shows that it is silver, blackened by sulphuration. Judged by its present appearance, it would be described as a black shield; it was made as a silver one. Its companion D 60 makes the same point with paint: its almost-black charges are azurite blue beneath translucent layers.

Seedorf is the other clear example: a field that now looks grey-green was originally blue, and its lion was silver. The same caution applies to the silver and gold described on the Marburg shields. Equally, not every heraldic silver was metal: on the older Schenk von Schweinsberg shield the silver lozenges are white paint, and on the Arnsberg shield white paint completes a silvered eagle.

A modern viewer looking only at the present object can therefore underestimate the brightness of the medieval shield. A freshly completed high-status shield may have combined saturated red, blue, yellow, black and white with large areas of reflective silver or gold. In sunlight, torchlight or candlelight, such a surface would have behaved very differently from the aged, matte museum object seen today.

This is not an invitation to make reconstructions excessively bright without evidence. It is a warning not to use centuries of ageing as the colour standard for a newly made medieval object.

Which Pigments Can Actually Be Identified?

The answer is narrower than a general list of pigments available to medieval painters.

For the medieval shields I have studied, the strongest object-specific pigment evidence presently comes from the painted Szczecin shield.

Soler del Campo's study of the Oña shield D 60 names azurite, lead white, minium and vermilion, but on a late-medieval surface, not a 12th- or 13th-century one.

Other shields securely preserve or historically record red, blue, yellow, green, black and white surfaces, together with silver and gold effects. But a colour name is not the same thing as a chemical identification.

That distinction should remain visible in any reconstruction. For example, medieval painters had access to several possible blue materials. Knowing that Seedorf was blue does not establish which one was used unless the analytical evidence identifies it. Likewise, "red" can describe several entirely different pigments.

My approach is to reconstruct from the identified pigment where one is known, and from the historically supported colour family where chemistry remains unknown.

What Was the Paint Bound With?

This is one of the most tempting places to overstate the evidence.

Medieval painters knew several binding systems. Egg-based tempera, animal glue, oils and other organic binders all belong to the wider medieval painting tradition. But a general medieval recipe is not automatically a shield recipe.

For the painted Szczecin shield, Jagielska reports an oily or fatty binder; the published summary does not make clear how an organic binder was identified. On the Oña shield D 60, Soler del Campo reports oil only in the stucco layers on the back, on a late-medieval surface. For the other shields in this article, the exact binder has not been securely identified in the accessible object-specific evidence.

That means resisting sentences such as "medieval shields were painted in egg tempera." Some may have been. The shields I have studied do not presently justify turning that possibility into a universal rule.

Medieval shield painters used coloured pigments within layered organic and mineral surface systems, but the precise paint binder is often one of the least securely identified parts of the surviving object.

Was Gesso Used on Medieval Shields?

Yes, but with an important qualification.

Several surviving shields provide strong evidence for pale mineral preparation layers commonly described in the literature as chalk ground, lime preparation, gypsum or gesso.

The painted Szczecin shield has a lime-white layer over a black organic layer. Seedorf preserves evidence for a prepared layer associated with its covered surface. Konrad von Thüringen's museum record names a chalk ground. Kohlmorgen describes chalk grounds on the Keseberg, Nordeck, Welfenberg/Helfenberg, Stromberg, Arnsberg, Heinrich I and several other Marburg shields. The late-medieval Oña shields have successive layers of gypsum stucco.

So the broad answer is clear. Prepared mineral grounds were an established part of medieval shield decoration.

"Gesso" should not be treated as one chemically identical substance in every century and region. Chalk, gypsum, lime-rich material and their organic binders must be distinguished where the evidence allows.

For reconstruction, the object being reconstructed should determine the recipe, not the modern umbrella word.

How Was a Medieval Shield Painted?

The surviving shields do not produce one recipe, but they do make it possible to reconstruct a recurring sequence of steps.

First came the wooden core. On many high-medieval examples, the wood was then covered with parchment, skin, leather or, in some decorative systems, textile.

A pale mineral preparation could be applied over that covering. This created a more uniform painting surface and, on some shields, provided material for relief decoration.

The heraldic field and charges were then painted. Silver or gold leaf could be incorporated where metallic heraldic tinctures or luxurious decoration were required. On elaborate shields, relief, patterned grounds, gilding and additional painted ornament could be added.

The reverse might remain simpler, but it was not necessarily ignored. Konrad von Thüringen's funerary shield shows that the back of a shield could itself become an elaborate painted and gilded surface.

The result was not simply painted wood. It was a composite decorated skin built over a wooden structure.

What Colours Would a New Medieval Shield Have Looked Like?

Probably much stronger than many surviving originals suggest.

Fresh vermilion can be intensely red. Orpiment can be brilliantly yellow. A properly prepared pale ground increases luminosity. Fresh silver leaf reflects light rather than appearing dark grey or black, as the silver field of the Oña shield D 59 does today. Gold leaf remains visually powerful even in relatively small areas.

The medieval shield therefore needs to be imagined before centuries of oxidation, abrasion, dirt, flaking, restoration and fading.

This does not mean every shield was colourful or expensive. The unpainted Szczecin object warns against that assumption, as do the many objects for which the original surface is incomplete or unknown. Many of the most colourful survivors are also funerary, processional or votive objects, which were more likely to be finished lavishly and to be kept.

But where the evidence demonstrates paint, ground and metal leaf, the original visual effect could be striking.

What the Surviving Shields Show

Read in date order, the shields studied here show several recurring features, though not a steady progression.

Painted shield surfaces existed by at least the 11th century within the material studied here. In the late 12th century, Szczecin provides direct evidence for a layered paint system and analysed pigments. Around 1200, Seedorf shows the combination of prepared covering, strong colour and silver leaf.

From Seedorf, Konrad (c.1230–1240) and Keseberg (c.1250) onward, parchment or skin coverings with chalk grounds recur in the surviving material, most of it from Marburg and known largely through Kohlmorgen. Marburg shields of the later 13th century and around 1300 show heraldic decoration that is modelled, gilded, silvered and physically three-dimensional. Konrad von Thüringen shows that elaborate painting was not necessarily restricted to the front.

Later shields continue to combine prepared surfaces, heraldic colour, metallic decoration and ornamental patterning, and many of the richest are display, funerary or votive objects.

At the same time, many exact pigments and binders remain unidentified. That final point is as important as the discoveries.

The surviving shields provide enough evidence to reject the idea that medieval shields were simply rough boards painted with whatever colour was available. But they do not give permission to invent one universal medieval paint recipe.

A Historically Responsible Reconstruction

For my workshop, the practical conclusion is straightforward.

This is the difference between a medieval-looking shield and an evidence-led reconstruction.

Conclusion

The surviving shields studied here do not reveal one medieval method of painting shields. They reveal a craft tradition.

At its simplest, that tradition could leave only a trace of red on a wooden object. At its most elaborate, it could involve parchment, textile, chalk ground, relief work, saturated pigments, silver, gold, patterned backgrounds and painted decoration on both faces.

The late-12th-century Szczecin shield provides the clearest pigment evidence: vermilion, an iron-oxide red and orpiment over a lime-white layer and a black organic layer, with an oily or fatty binder reported. Seedorf and the Oña shield D 59 show how dramatically an aged surface can mislead: a grey-green field that was blue, and a black field that is tarnished silver. Konrad von Thüringen's funerary shield shows the shield transformed into a richly painted, gilded and modelled object. The wider 13th- and 14th-century material, much of it from Marburg, shows parchment or skin coverings and chalk grounds as recurring technologies, while metallic and relief decoration could turn heraldry into a surface that reflected and modelled light as well as colour.

But the most important conclusion may be methodological. A reconstruction should follow what the evidence preserves, not what a medieval shield is expected to have been.

Sometimes that evidence preserves the pigment. Sometimes it preserves only the colour. Sometimes it preserves the ground but not the binder. And sometimes the historically correct answer is simply that the original surface is no longer known.

That uncertainty is not a weakness in reconstruction. It is part of reconstructing the past responsibly.