Early heraldry in the late Romanesque wall paintings of the church of Artins (France)
Forgotten in the depot of the ‘Kunstsammlungen’ in Coburg Castle for decades, a ‘medieval tapestry’ was rediscovered in 2014. Soon after, it was determined, that the ‘tapestry’ was not in fact a tapestry, but a medieval wall painting that had been removed using the so called ‘strappo’-method. Besides the monumental battle scene, three smaller painted fragments – showing early modern coats of arms – were also found in Coburg Castle. Although it was obvious that the heraldic fragments had been painted in another period, their common discovery suggested that they might originate from the same context. The journey of the paintings, which were transferred onto pieces of fabric, can be traced back from the moment they were found to the end of World War II, when they were transported – without any indication regarding their provenance – to the Central Collecting Point (CCP) in Munich in 1945. After being rediscovered they were restored by Christoph Merzenich and Nico Wörner at the University of Applied Sciences in Erfurt and became part of the permanent exhibition of the ‘Deutsches Burgenmuseum’.
However, in 2018 many questions regarding the dating of the fresco, its original context as well as a possible function remained unanswered. In order to understand these and many other ambiguous aspects, an interdisciplinary conference was held in Bamberg in 2018, where scientists reviewed and discussed the painting. Especially due to the knights’ armors and helmets, dating the fresco in the last decades of the 12thcentury seemed the most likely. Furthermore, the examination of the early modern coats of arms pointed tendentiously to the noble des Loges family from western France. Consequently a French origin of the frescos became a strong possibility. This theory was proven shortly after the conference ended, owing to Fabian Brenker’s coincidental unearthing of a scientific article from 1909 about a church in the French village of Artins and its former medieval fresco-decoration. Thanks to a printed aquarelle showing the battle scene in this article, it is now certain that the fresco in question comes from the church Saint-Pierre, where it had been for more than 700 years until it was removed at the beginning of the 20thcentury.
Besides the dating of the fresco, the exploration of its early-heraldic elements also provide further information regarding the development of heraldry itself and its use by artists at the end of the 12thcentury. Despite the painting’s fragmentary condition, eight shields carried by the knights can still be seen. While half of them are round, the others illustrate the more prevalent triangular form seen in medieval Europe at that time.
The visual differentiation of two fighting groups in a battle scene evoked by equipping the knights with two contrasting types of shields, namely round and triangular, was not uncommon in medieval art. Corresponding examples can be found in many genres. The glass paintings of the Notre-Dame Cathedral in Chartres (France) and a sculptured capital of the ‘Palacio de los Reyes’ in Estella (Spain), both depicting Charlemagne’s nephew, Roland, fighting against the Saracen Ferragut, are proof. In both cases, the Christian heroes carry triangular escutcheons, while the enemy is characterized as ‘the other’ by the much older proto-heraldic round shield. The same observation can be made with regard to wall paintings depicting the battle of Antioch in 1098 in the church Saint-Julien in Poncé-sur-le-Loir – a village just six miles from Artins. The Christian crusades and the adversarial Saracens, are armed in the same way as seen in the previous examples. Particularly due to the latter being so similar to our topic in so many ways, it seems rather likely that the battle scene of Artins can be prudently set in a crusade context, as previously suggested by Clément and Hallopeau.
Concerning the stylistic development of the escutcheon, the fresco’s shields reflect a status between the ‘almond shield’ of the late 11thcentury and the explicitly triangular form of the 13thcentury heater shield. The top of the escutcheons are no longer curved, yet the corners remain rounded, and although they still tend to be elongated they no longer cover the knight’s entire bodies. Contemporaneous wall paintings portraying armed knights on horseback can be found in many regions of western and southern Europe. Two supportive examples for dating the shields in Artins are, in my opinion, the fresco decorations of the castle chapel in Hocheppan as well as the so-called ‘Iwein-Chamber’ in Rodenegg Castle. These prominent examples of high medieval wall paintings in South Tyrol (Italy) stage pairs of combatting knights. A cavalier riding on a griffin in Hocheppan and Iwein in Rodenegg are both equipped with very similarly shaped escutcheons, as seen in Artins. Since the ‘Iwein-Cycle’ is dated around 1205–25 and the lower part of the chapel decoration should be seen as having been executed around 1180, dating the escutcheons of the knights in Artins to the 1180/90s is probable.
While the form of the shields and their stylistic contextualization allowed a rather decided assignment of the paintings to the last decades of the 12thcentury, the examination of their decoration remains less clear. As far as the round shields are concerned, their pre-heraldic status is obvious, not only because of their form, but also their decoration with concentric circles. The triangular escutcheons in contrast reflect an intermediate status. In heraldic terms they show pales, bars, hurts and bordures as wells as – in one case – a fragmentary inscription. Only the characters ‘E’ and ‘R’ are evident, while the rest of the word remains unclear. A ‘C’ or a ‘G’ could be identified as the first letter and it might be followed by an “I” or an “L”. Since the battle scene almost certainly depicts a crusade scene, it could be argued, that the Old French word ‘ESCLER’ (‘Saracen’) was originally written there. Although the shield with pales respects the correct use of tinctures, the others still show prohibited color combinations as well as the entirely non-heraldic characters.
However, most of the triangular escutcheons show future heraldic ordinaries and other charges that all derive from iron elements, originally placed on the shields to strengthen them. In doing so, they build a strong contrast to the round shields of the enemies showing exclusively circular elements. Although the escutcheons of the supposedly Christian knights already use heraldic elements, they do not show ‘true’ coats of arms, as documented, in my opinion, by the visual resemblance of the bars in the center shield and the frame borders of the frescos. Both elements reflect a widespread color combination in secular wall paintings in this region of Western France. In this respect, the early-heraldic bars seem to have been integrated into the coats of arms for artistic reasons rather than being a true reflection of actual heraldically “correct” coats of arms.
In summary: Portraying the two combatting groups – likely Christians and Saracens – with two different types of shields, that express a proto-heraldic and an early-heraldic status, respectively, not only in form, but also in decoration, prove the differential use of the very early phenomenon of heraldry by artists at the end of the 12thcentury.
 For the ‘strappo-method’, which makes it possible to remove wall paintings from their original context by transferring them onto pieces of fabric, see Paolo Mora/Maura Mora/Paul Philippot, Conservation of Wall Paintings, London 1984, p. 257–260.
 Paul Clément/Louis-Alfred Hallopeau, Peintures murales de l’église paroissiale d’Artins, in: Bulletin archéologique du Comité des travaux historiques et scientifiques 1909, p. 136–148, pl. XIII.
 See Fabian Brenker/Steffen Kremer/Christoph Merzenich, Wiederentdeckte mittelalterliche Wandmalereien aus Artins (Frankreich) im Deutschen Burgenmuseum auf der Veste Heldburg, in: Kunstchronik 72/4 (2019), p. 190–196 for an even more detailed report. Furthermore, the wall paintings will be discussed in the forthcoming article by Anja Grebe/G. Ulrich Großmann, Une peinture murale romane française dans le ‘Deutsches Burgenmuseum’ à Heldburg (Thuringe), in: Bulletin monumental 179/1 (2021).
 For a more extensive discussion and contextualization of the early heraldic elements see the article by Steffen Kremer, Zur Frühheraldik in den spätromanischen Wandmalereien der Pfarrkirche von Artins, in: Archiv für Diplomatik 67 (2021) (forthcoming).
 Corresponding examples of high and late medieval illustrations in manuscripts, that also reflect a differential use of round and triangular shields, are discussed by Fanny Caroff, Différencier, caractériser, avertir: les armoiries imaginaires attribuées au monde musulman, in: Médiévales 38 (2000), p. 137–147.
 As regards the capital and its decoration see the contribution of Joan T. Grimbert, Tristan and Iseult at the Cathedral of Santiago de Compostela, in: Elizabeth Archibald/David F. Johnson (eds.), Arthurian Literature 31, Woodbridge 2014, p. 131–164, here p. 147–148.
 With regard to the question, whether the church’s fresco cycle reflects a sort of visual crusade propaganda, the wall paintings of Poncé-sur-le-Loir were discussed by Elizabeth Lapina, La représentation de la bataille d’Antioche (1098) sur les peintures murales de Poncé-sur-le-Loir, in: Cahiers de civilisation médiévale 52 (2009), p. 137–157.
 Clément/Hallopeau, Peintures (as in no. 1), p.145.
 Torsten Hiltmann, L’héraldique dans l’espace domestique. Perspectives historiques sur les armoiries et le décor héraldique dans l’espace profane (espace germanique, XIIIe– XVIesiècle), in: Le Moyen Âge 123 (2017), p. 527–570.
 Franz-Heinz von Hye, Wappen in Tirol – Zeugen der Geschichte. Handbuch der Tiroler Heraldik (Schlern-Schriften 321), Innsbruck 2004, p. 25.
 Wolfram von Eschenbach, Willehalm (Bibliothek des Mittelalters in vierundzwanzig Bänden 9) hrsg. von Joachim Heinzle, Frankfurt am Main 2009, p. 1210. For this indication I am grateful to Birgit Zacke (Bonn).
 I am also very grateful to Hillary Hatch-Conrad (Cologne) for the linguistic correction of this paper.