Previous Section Contents Page Next Section



4) Case study; the Charlemagne window at Chartres Cathedral

Elements of the sometimes abstruse theory presented in the preceding chapters may become clearer when put into practice. With that in mind, this chapter is given over to a detailed case study - a ‘close narratological reading’ of the Charlemagne window at Chartres Cathedral. Naturally, a ‘narratological reading’ does not replace a traditional iconographical reading but simply aims to build another layer of analysis on top of it. With that in mind, I have tried to follow Panofsky’s excellent advice and begin with ‘pre-iconographic’ description in each case, before moving on to ‘deeper’ readings. 1

As a physical object, the Charlemagne window is a single lancet, measuring 9.03m high by 2.22m wide. Although the chronology of glazing at Chartres has at times  been controversial, Grodecki’s dating to around 1225 seems most likely.2   The window occupies bay 07 and along with its neighbour (the Life of St James the Greater in bay 05) separates the eastern axial chapel from the deep radiating chapel in the northern ambulatory. 3 Although the base of the window is around 7m above the ambulatory floor, a significant proportion of the detail in most panels is legible to any observer at ground-level with good eyesight.

This chapter begins with an outline of the story (at the fabula level) and an overview of the visual discourse presented in the window. The rest of the chapter is then given over to a panel-by-panel account of the details of that discourse.

In all the discussion that follows, the numbering of the panels is as per the diagram in Fig. 4.00. Currently, there is no standardisation in the way different authors number panels within stained glass windows. Some authors try to number the panels according to the supposed narrative sequence, which causes problems every time the identification of a particular scene is reconsidered. The system I have adopted throughout is to number panels according to their location, counting from left to right and bottom to top (where they overlap, I count from the geometric centre of the panel). All of the individual panels are reproduced in detail in the accompanying plates.


4.a) the story

Apart from the extra-narrative signature panel (Panel 01), the story elements represented in bay 07 at Chartres naturally split into four major segments; the Jerusalem Crusade, the two Spanish campaigns and the Mass of St Giles. The first three of these can also be broken into distinct fragments as shown in Figure 5 (the numbers in square brackets indicate the panels within the window which depict that part of the narrative).

Figure 5 - Breakdown of the story elements in the Charlemagne window [with panel numbers]

In terms of relative chronology, segment 1 (Jerusalem crusade) happens quite early in Charlemagne’s rule. There is then an unspecified ellipsis of several years before segment 2, which lasts around three years and is closely followed by segment 3. Segment 4 happens at an indeterminate time, as discussed below in the section on panel 22. The main events at the fabula level can be summarised as follows:


Segment 1: The Jerusalem Crusade

In Constantinople, the Emperor (Constantine VI - reg.780-797), longs to recover the Holy Land.4 He has a dream in which a Frankish king defeats the Muslim armies and recaptures Jerusalem. He therefore sends emissaries to Charlemagne’s court at Aachen, asking for his help. Charlemagne then travels to the Levant and does indeed conquer the Holy Land. On his triumphant arrival back at Constantinople, the Emperor offers him great riches, however Charlemagne asks for nothing but some holy relics. He then returns to Aachen and presents these relics to the Church.

Segment 2: The First Spanish Campaign

Every night, Charlemagne studies a band of stars (the Milky Way) stretching across the sky from east to west and wonders at its meaning. One night St James appears to him in a dream and tells him it is a sign that he must undertake a crusade to free Galicia from Muslim rule and restore the saint’s shrine at Santiago. Charlemagne and his armies cross the Pyrenees and besiege Pamplona, which only falls after divine intervention causes the walls to collapse. The king conquers and converts the rest of Spain and spends three years rebuilding the Church of St James. He then returns to France and builds many more churches using the riches he acquired in Spain.

Segment 3: The Second Spanish Campaign

Spain has fallen back into the hands of the Muslims and their allies. Charlemagne marches his armies westwards again and engages the Saracens in a series of battles. Before one great battle, at Sahagún, his knights plant their spears in the ground and in the morning many of the spears have sprouted new growth, as a sign that their owners will achieve martyrdom that day. Despite heavy losses, the Christians are victorious and press on with their campaign. Their progress is delayed by a Saracen giant named Ferracutus, who defeats all comers in single-combat. The king’s nephew, Roland, eventually overcomes Ferracutus and kills him with a blow to his only vulnerable spot - his navel. After many further victories, Charlemagne accepts promises and tributes from his remaining enemies and leads his weary army homewards again. Roland stays behind at the Roncesvalles Pass to protect the rearguard but owing to the treachery of one of Charlemagne’s noblemen (Ganelon), his small force is attacked and overwhelmed. Roland himself is mortally wounded. After trying to smash his famous sword, Durandel, he sounds his horn (the oliphant) and collapses. His friend Baldwin finds the dying hero and carries the sad [96] news to Charlemagne, who has arrived back too late to help. The heroic dead are given suitable burials and after trial by combat, Ganelon is executed by quartering.

Segment 4: The Mass of St Giles

At some unspecified point in his career, the King is troubled by an ‘unnameable sin’, for which he is truly repentant -  but which he cannot bring himself to confess to anybody. One day, whilst the priest Giles is celebrating mass, an angel appears and hands him a scroll on which the details of the King’s sin are written, along with confirmation of his absolution.

4.b) background to the stories

4.b.1) Sources

The events depicted in the window derive from three main textual traditions. Segment 1, with its entirely mythical account of Charlemagne’s liberation of Jerusalem,  is based on an anonymous text composed at the Abbey of St Denis early in the twelfth century (shortly after the actual First Crusade), known as the Descriptio.5 According to this fictional chronicle, the relics given  to Charlemagne by the grateful Constantine included the Crown of Thorns, a nail from the Crucifixion, a piece of wood from the True Cross, the Sancta Camisa (the tunic worn by the Virgin at Christ’s birth), Christ’s swaddling clothes and the arm of Simeon (the arm which had held the Holy Infant at the Presentation). All of these, claimed the Descriptio, were donated by Charlemagne to the Abbey of St Denis. As Emile Mâle recognised, the legend of the Jerusalem Crusade was most likely invented by the monks of St Denis in an effort to help validate the relics which they claimed to possess at that time. 6 In choosing to include this story in bay 07, the cannons of Chartres Cathedral were, in effect, ‘piggy-backing’ onto this legend in order to validate their own most sacred relic, the Sancta Camisa. Physically possessing a relic was only the beginning - real power meant being in control of its narratives, not just of its origins but also of its invention, its translation and the record of miracles it worked. [97]

Segments 2 and 3 (the Spanish campaigns)  are based on the Historia Karoli Magni et Rotholandi more commonly known as the Pseudo-Turpin Chronicles (hereafter Pseudo-Turpin) named, as Hamilton Smyser rather harshly put it,

...after that graceless Unknown who, writing in the middle of the twelfth century, had the temerity to pass his chronicle off as the mémoires of Archbishop Tylpin, or Turpin, of Rheims, a contemporary of Charlemagne.7

As an historical figure, ‘Tylpin, or Turpin’ was an Archbishop of Reims in the late eighth century, about whom nothing else is known. As a twelfth century narrative construction however, he becomes a very colourful character indeed;  friend and confidant of Charlemagne and a brave warrior-priest, who accompanies his King on foreign crusades, fighting valiantly alongside him and documenting their adventures in his eponymous chronicle. This chronicle, which was almost certainly concocted in northern France around 1140, was the well from which several of the later prose epics and chansons de geste about Charlemagne and his ‘nephew’ Roland sprang. A number of early Latin manuscript variants of the text survive, as well as several vernacular versions.8   Thomas Rodd, in the preface to his early nineteenth century metrical English translation, reported that the earliest known French prose version was written by one ‘Jehans’, at the behest of a prisoner of Philip Augustus (d.1223), something which suggests that popular interest in the text was thriving at the start of the thirteenth century, around the same time Chartres Cathedral was under construction. In 1929, Joseph Bédier proposed that the text was composed to fit into a larger mid-twelfth century compilation known as the Book of St James, a complete edition of which survives in the Codex Calixtinus (Santiago de Compostela, Cathedral Archives) and which also includes accounts of the miracles of St James, the story of his translation to Santiago and the famous ‘Guide for Pilgrims to Compostela’.9 As will become clear in the sections that follow, I regard this association with St James and pilgrimage as highly significant.

Finally, segment 4 of the window’s narrative (The Mass of St Giles) is a well-known story found in most of the many versions of the Vita sancti Aegidii, which were in circulation from the tenth century onwards.10

[98] Although these three are the most obvious sources for the narrative discourse found in bay 07, it is important to remember that by the start of the thirteenth century, the mythology surrounding the historical figures of Charlemagne and Roland was already well developed and diverse. As well as the widely circulated ‘official biography’ of the Holy Roman Emperor, Einhard’s Vita Karoli magni, there were also several variants of the Old French Chanson de Roland (a Chanson de Geste probably composed in the eleventh century). One must assume that such a popular story would also have been well represented in those vernacular traditions that leave no documentary evidence. For every person who knew Charlemagne from personally reading a manuscript of Pseudo-Turpin, there must have been hundreds who knew him from a range of oral and visual sources, the details of which may have matched that text to a greater or lesser degree. 11 In other words, what we are probably dealing with, when looking for the ‘sources’ of the Chartres window’s narrative is the kind of complex ‘implied text’ discussed in Chapter 1. Emile Mâle’s search for a single manuscript that contained all these story segments in the correct form led him to a text entitled  De la Sainteté, des mérits et de la gloire des miracles du bienheureux Charlemagne (Paris Bibl.Nat. mss. lat. 4895A and 6187), which included both the Jerusalem and Spanish crusades but which made no mention of St Giles. Mâle’s conclusion from this was that there must be some lost, or as yet undiscovered thirteenth century manuscript that contained all three stories. To be fair, his automatic presumption of a single manuscript source was not without reasons. As he pointed out, the famous Charlemagne reliquary at Aachen features the same basic story segments as the window (in a series of repoussé copper gilt plaques - see Fig. 4:28A for an example) and for Mâle it was obvious that ‘ close an analogy between works of art from two regions so remote from each other proves that they both derived from a common written source’. 12  Leaving aside the question of just how physically and culturally remote Chartres and Aachen actually were from each other - and from other major sites on the commonly trod pilgrimage routes, this over-reliance on a single textual source (even an unknown one) can be dangerous. Quite apart from our not knowing what other forms of transmission may have been operative in this area, the assumption that we are looking at the visualization of a specific text can all too easily lead us to mis-interpret narrative images in order to meet our preconceptions. As we will see in the detailed discussion of the individual panels below, this is a trap into which most of the studies of the Charlemagne window have fallen to a greater or lesser degree. [99]

4.b.2) Significance of the story for thirteenth century audiences

Charlemagne was canonised (as a Confessor) on the 28th of December 1165 by the Antipope, Paschal III, presumably at the behest of Frederick Barbarossa. Although most of Paschal III’s actions  were subsequently annulled by Rome, the sanctification of Charlemagne appears to have been left unchallenged. In an age when the Church was actively concerned with promoting crusade and pilgrimage, his story was perhaps too good an exemplum to waste. And indeed, when one considers the version of Charlemagne’s legends depicted in the window at Chartres, crusade, pilgrimage and devotion to the Church are the themes which feature most strongly. When compared to the multi-faceted character who emerges from Pseudo-Turpin, and even more so from the later chansons, the Charlemagne presented in bay 07 is a simple, one might even say Bowdlerised, figure. The only note of controversy is the inclusion of the Mass of St Giles and even here there is no hint of what the unnameable sin might be - we only see the episode as an exemplum of divine forgiveness. The King’s failings as reported in Pseudo-Turpin - his occasional naivety, and incidents such as his ill-treatment of the poor in chapter 12 (which loses him the opportunity to convert Argolander) are omitted from the chansons as they are from the window, while the earlier text’s repeated accounts of  individual battles are elided or generalised. Instead, Charlemagne’s piety, service and submission to the Church occupy centre-stage (literally so in the case of the window, where the donation of relics, the rebuilding of a church and the Mass of St Giles are all sited on the central axis). Régine Lambrech has argued that the message of the Charlemagne narrative presented to contemporary monarchs in later prose versions such as the one that appears in the  Grandes chroniques de France was straightforward:

...Charlemagne was a servant of the Church; look at his pious conduct, at his generosity to the Church, at the miracles God worked to reward him for his religious zeal. You should emulate his devotion to the Church.13

As in the Grandes chroniques, so in the window at Chartres - where it is noteworthy (though not always noted) that the discourse is a very different one from the earlier verse-narrative traditions that grew up around Charlemagne and Roland, as it also is from the various secular image cycles.14 Confronted by this relatively peaceful depiction of the life of Charlemagne, one can only imagine what fun the artists who made the Morgan Picture Bible might have had with Pseudo-Turpin’s lengthy and frequent descriptions of warfare. [100]

4.c) the Charlemagne window as visual discourse

4.c.1) Extent, arrangement, directionality

The Charlemagne window is a polyscenic narrative, in which one extra-narrative element (the signature panel) and four narrative segments of unequal length are spread over twenty-four separate panels (see the plan and narrative flow diagram on Fig. 4.00). The first three segments are themselves depicted in polyscenic fashion, whilst the fourth (the Mass of St Giles) is monoscenic; a single narrative panel (22) flanked by two smaller, hypotactic panels (23+24). Each panel presents a single, self-contained scene, except for panel 19; the dramatic climax of the third segment, in which the hero Roland is repeated left and right of the central axis, performing two distinct actions which closely follow each other in the narrative (see detailed description for this panel below).

The panels are arranged in a complex pattern within a tall, thin lancet. The central axis comprises a half-circle at the bottom plus seven equal-sized circular medallions contained within alternating circular and lozenge-shaped openings in the armature (this shows the danger of relying on armature patterns alone when considering geometry - on bright days the hard outlines of the armature dominate but when the light is more diffuse, the viewer becomes more aware of the circles within those lozenges). Flanking the lozenges are larger lateral semicircles, divided vertically into two panels each. In one of the more imaginative studies of the Charlemagne window, Stephen G. Nichols claimed that these various shape elements were themselves bearers of complex meanings.15 However there is no correlation between these supposed meanings and the contents of the panels concerned, either in this, or in the myrian other windows that use such shapes, and his theory has attracted little scholarly support. Duncan Robertson has argued that the ‘centrifugal’ arrangement of panels adopted in this window, with its visually dominant central axis and slightly weaker lateral medallions, is ideally suited for typological windows (which are intrinsically hypotactic) but presents certain challenges for straightforward narrative.16 As we will see, the designer here has risen to those challenges well, positioning his heroes upon this central spine in such a way that the narrative unfolds around them.

The general direction of reading follows the usual French cathedral glass pattern of ‘bottom to top’. Within this, the reading is broadly boustrophedonic, though the three main segments differ slightly in how they fit the central frames into this pattern. [101]

The border spaces surrounding the medallions are filled with predominantly blue and red vegetal ornamentation, scale-like patterns and rinceaux. All of the narrative action is contained within the medallions, with no image elements over-spilling the outer limits of the {picture} frame, although in several cases body parts and other elements overlay the concentric red and blue bands that mark the interior borders of those frames (see the individual panel descriptions for details).

4.c.2) Anisochrony and frequency

One of the most interesting features to emerge from a comparison between the discourses in the texts and in the window is in the treatment of frequency (see Chapter 2.b.4). Particularly in Pseudo-Turpin, several actions or happenings re-occur in multiple episodes. Some of these repetitions, such as departure for war, the taking of an enemy town, battles against the main foe, the building of churches, etc. are only to be expected in a crusade narrative whilst others are more noteworthy; the two miracles of the flowering lances (before the Battle of Sahagún in chapter VIII and then again before the Battle of Saintonge in chapter X), or the repetition of St James’ appearances to Charlemagne over three successive nights.  In the window however, each of these repeated events is depicted once and only once; the army setting out for Spain (panel 10), the taking of a city by force (panel 12), the construction of a church (panel 13), a battle between crusaders and Moors (panel 14) and the flowering of the lances (panel 15).

Most of the authors who have written about the window have described each of these panels as representing one or other of the specific events described in the sources, even if they disagree over which specific event. Thus Mâle, Delaporte and Robertson all identify the city taken in panel 12 as Pamplona, despite the obvious mismatch with the texts (see the detailed discussion of the panel below), Lejeune & Stiennon uniquely regard it as the city of Noples.  Maines takes a more flexible approach in this instance, arguing that ‘ seems reasonable to suggest that panel 12 is a generalized scene representing all those cities taken by force’.17 Maines similarly regards panel 14 as a generalised representation of a battle scene, although for reasons which are not entirely clear, he considers the church construction site in panel 13 to be specifically that of Saints Facundus and Primitivus at Sahagún (the other authors all regard this as the rebuilding of Santiago Cathedral).

The broader question of whether a particular scene in a polyscenic narrative represents a generic event, or a specific instance of that type (not a generalised battle within the campaign in general for example but this specific battle), is a difficult one. Perhaps the greatest advantage of the indeterminacy of visual narratives is that in most cases it can represent, or at least evoke, several different [102] narrative referents;  this specific battle, all the battles in that campaign - and extending the chain of signification still further, all similar battles (between an army of type X and an army of type Y), right back to the wars of the Old Testament. Whilst such looseness may be annoying to modern scholars, it was manna for the typologically inclined medieval theologian, for whom a figure like Charlemagne was an open invitation to discuss Joshua (consider the collapsing walls of Pamplona...), David, Saul, Abraham, and so on.

Perhaps the strangest of the repeated elements in bay 07 is the cloud frill that appears at the top of several of the panels. In panel 08, where it clearly represents the subject of Charlemagne’s nocturnal contemplation, there can be no doubt that it denotes the Milky Way. Most authors have conveniently overlooked its rather erratic appearance in the other panels. Clark Maines tried to explain these appearances in terms of the Pseudo-Turpin reference to the cloud being not just a geographical sign of the direction of Charlemagne’s journey but also a sign of his greatness to come and of his eventual expulsion of the ‘perfidious pagan tribes’. He then stretches this point slightly to see its appearance in the scenes of Constantinople (panel 04) and the Holy Land (panel 05) as representing the recent Latin conquest of the former in 1204 and the still longed for victory in the latter.18 The only problem with this over-interpretation is that these cloud frills are extremely common throughout the windows at Chartres in contexts where such an interpretation would be meaningless. Indeed, approximately one quarter of all the panels in all the narrative windows at Chartres feature a cloud-frill of some kind. Some of these are for specific purposes, as markers of oneiric content or as the Deus ex machina from which God’s hand of blessing appears. A great many of them however have no discernable narrative purpose or pattern at all. A much deeper analysis is required if there is ever to be a systematic taxonomy of cloud-frills in early thirteenth century stained glass.19 Frustratingly, for the moment it may have to suffice to say that some cloud frills are significant - but that any educated guesses about their meaning are just that. It may however be significant that in the window adjacent to this one, the ‘Life of St James’ window in bay 05 (see Fig 4.31), the topmost half-panels each contain nothing but a cloud frill, something to which I will return at the end of this chapter.

4.c.3) Structural/Functional repetition

Although as mentioned above the Charlemagne window reduces most repeated scenes to a single image, at the level of the narrative function (see chapter 2.e) this visual discourse does [103] feature two repeating nuclei; the prophetic dream (panels 03 and 09) and the consultation with learned men (panels 02 and 08). These panels actually repeat themselves structurally as well as functionally - and in so doing help to clarify the otherwise debatable divisions between narrative segments.

In panel 03, which opens the first narrative segment, it is the Emperor’s dream about Charlemagne that prompts him to send envoys to Aachen (panel 02), where they consult with the king and his trusty Archbishop. This combination of dream and consultation is repeated at the start of the second segment but here the sequence is reversed - Charlemagne consults his astronomers (panel 08) but they are unable to explain the Milky Way, whose significance is then revealed to him in a dream (panel 09). Significantly, Pseudo-Turpin makes no absolutely no mention of the events depicted in panel 08 - he tells how ‘...night after night he was wont to contemplate it, meditating on what it might signify’ (Chapter II) but never mentions the King discussing it with anyone. panel 08 has therefore been invented by the artist for some other purpose.

Wolfgang Kemp, in his broader discussion of typological narratives remarked on the ‘doubling’ of events in this window:

...identical motifs begin the first two episodes, the Jerusalem Crusade and the first Spanish expedition: first, the dream of the Byzantine emperor Constantine, whose message is communicated to Charlemagne by Constantine’s ambassadors; second, the dream of Charlemagne, the vision of the Milky Way, which is pointed out to Charlemagne by two companions. After the dream or sign and its interpretation follows the deed of the hero...20

In fact the repetition on the structural and functional levels may be even more complex than this, as Fig. 4:23 should reveal. At the functional level, panels 03 and 08 each begin a narrative segment (initiation of a quest) and furthermore they each serve to initiate an uncertainty; will Charlemagne answer the Emperor’s plea? What do the starry signs mean? In each case, the uncertainty is then resolved in the following panel (02 and 09). If one extends the form of these first two sets of panels (02-07  and 08-13) upwards, the next group with this shape would begin with panels 14 and 15. Although these panels depict somewhat different nuclei at the functional level (although the sleep-theme is still there), structurally they repeat the same pattern - the initiation of uncertainty in panel 15 (what do the flowering lances forebode?) and the resolution of that uncertainty in panel 14 (the deaths of those thus singled out for martyrdom). Whilst on its own this might not be enough to argue for a particular grouping of panels 14-21 as the third narrative segment, they certainly strengthen the case when considered alongside the actual events depicted in those panels (see the individual descriptions below).

[104] As Fig. 4:23 shows, the initiating panels for the three segments alternate, left and right, creating a double chiasmus. Whilst this may be pure coincidence, chiastic compositions were certainly popular elsewhere in medieval art, so there is always the possibility that this arrangement was intentional (see the discussion of the Bourges Good Samaritan window in chapter 9 for more on the significance of the chiasmus).

4.d) panel by panel breakdown

NB. In this section the panels are presented in their narrative order. The accompanying plates (without which this text will make even less sense) are arranged by panel number.

Panel 01: Signature panel (a furrier and his customer)

In keeping with the {interpretive} macro-frame common to most of the lancets at Chartres, the bottom panel  contains an extra-narrative ‘trade’ scene. A furrier holds up a full-length vair garment (the alternating black and white triangles, now rather faded, function as an iconic sign for the particoloured squirrel fur used for lining rich garments at the time; a conventional sign for such garments and also a widely understood symbol for wealth/high status) for inspection by a customer, who wears a full-length hooded gown and carries a staff. To the left of the figures is a portico. To the right is an elaborate chest out of which spills another fur garment, whist a third is hanging from some kind of support obscured by the upper right {picture} frame.

The scene is enclosed within the semi-circular arc of the metal armature, inside which are thin bands of blue and then red glass. The latter overlay and obscure the top of the fictive architecture but are themselves overlaid by the heads of the two figures, contributing a basic sense of depth and suggesting that the figures are standing in front of the portico. The frame is thus being used to suggest that the scene is taking place in front of a furrier’s shop.

Clearly, as with all signature panels, this scene is extra-narrative, playing no part in the story depicted above it. The observation by Elizabeth Pastan that in this scene, ‘a cloth not unlike [the Sancta Camisa] is held up for viewing’ is tantalising - but perhaps a little disingenuous, partly because this is clearly fur, not cloth, but also because this composition is really no different from the signature panels of furriers in other Cathedrals, where the connection with that relic would be meaningless.21   It has been speculated that Charlemagne was chosen as a suitable theme for the sponsorship of the ‘Furrier’s Guild’ (along with St Eustace in bay 43) owing to his fondness of hunting. As always however this assumption has to be tempered by an awareness that [105] there were no established guild structures in place in Chartres at this time. 22 Despite the absence of any underlying fabula, the scene is presented with a certain amount of pseudo-narrative richness, as is typical for the signature panels at Chartres. This is not a static image of a furrier but a dynamic scene of interaction between a tradesman and his client.

Panel 03: Constantine’s dream (Jerusalem Crusade, scene 1 of 6)

Fictive architecture denoting an interior setting  - a tall slender portico topped with crenellations, a gold-tiled dome and green-tiled pitched roof. The left foreground is dominated by a figure sleeping on an elaborate bed. The crown worn by this sleeping figure indicates that he is a ruler - in this case the Emperor Constantine (though there is nothing in the image to identify him explicitly). One has to assume that medieval kings did not generally sleep with their crowns on, though they are invariably depicted thus - see for example the dreams of Pharaoh (Fig. 4.24B) and of the other Constantine elsewhere at Chartres (Fig. 4.24C). The crown here is therefore functioning as part of a conventional sign (not an iconic sign), of a sleeping ruler. To the right is an armoured man on a horse, seemingly riding through the sleeper’s bedroom. This mounted knight is helpfully identified with a titulus reading ‘CAROLVS’, which floats in a strip of black-painted glass against the blue-ground in front of the rider (this semantic enclave is subtly different from the ones in panels 02 and 08, where the titulus is separately framed). He also wears a crown over his closed-faced helm.

Helmets and shields function as consistent internal indices throughout this window; the Christian knights all carry kite-shaped shields and wear closed-faced helmets, while the Moors all have round shields and open, conical helms. Another interesting index (both internal and external) to make its appearance here is Charlemagne’s banner - the Oriflamme. Although it supposedly originated as the banner of the ancient kings of the Vexin,in medieval Europe the Oriflamme became closely associated with Charlemagne. In the Descriptio it features in the Emperor’s dream as the banner carried by the king who will defeat the Saracens.  Similar banners appear in panels 05, 11 and 12, although the colour varies. This cultic symbol of Frankish royalty was stored at the Abbey of St Denis and given to the French King’s standard-bearer (the porte-oriflamme) whenever he set out for war - an event depicted symbolically in one of the transept clerestory windows at Chartres (see Fig. 4.24A), where the Constable of France, Jean Clement, is shown receiving it from St Denis himself. Phillip Augustus, recognising its value in the manufacturing of Capetian identity, adopted the Oriflamme as his [106] battle standard, most famously carrying it at his victory over the Imperial alliance at Bouvines in 1214. 23  Given its subsequent importance to French national identity, it is likely that any image of a red or golden pennant of this type would have suggested the Oriflamme to a contemporary audience.

Overall, the scene is a representation of dreaming, shown in a conventionalised format that was common in medieval art;  a sleeping figure denoting the dreamer (see chapter 5.b regarding embedded narratives and the representation of dreams) juxtaposed with the contents of the dream. A less common feature is the angel floating over the sleeping King’s head and pointing towards the vision. This detail is specifically mentioned in the Descriptio:

While I was asleep, an angel appeared, who said to me “Behold Charlemagne, the King of France, your defender”.24

Ontologically, the angel does not belong entirely to the level of the dream content nor to that of the dreamer but acts as a hinge-function between the two worlds. His narrative role is that of a modality marker, denoting that the dream is divinely inspired (and therefore trustworthy).

In terms of {picture} framing, the scene is bounded within the lower left quadrant of a circle, with the upper left corner sliced off by the adjoining lozenge (panel 04). The right edge is formed by the metal armature only, with no internal border. The upper and top-left boundary is marked by a thin red border, while the lower left border (corresponding to the outside of the circle) has thin concentric bands of white, then blue, then red glass. The head end of the Emperor’s bed sticks out  over these bands, pushing him into the foreground (a conventionalised rather than iconic system of perspective). All other elements at the edge of the scene are overlaid and cut off by the frame. A small red cloud-frill motif at the lower right corner, beneath Charlemagne’s horse, may be intended to hint at the oneiric status of that part of the scene. The curious white horizontal strip at the bottom right corner may be a stylised representation of floor tiles - a similar pavement appears at the bottom of panel 22.

As a narrative function, this scene marks the opening nucleus of segment 1. A helpful magically or divinely-inspired dream suggesting a possible answer to the problem of how to obtain the ‘longed-for thing’ is a topos found in many narrative traditions, including the Old Testament, classical literature and folk tales. In the Christian narrative tradition, it commonly functions as the initiator of a quest, as is the case here. Because the dream is shown as being divinely inspired, the scene is also proleptic with regard to Charlemagne’s later success - since angelic messengers do not lie, the presence of the angel pointing towards the dream-content indicates that it [107] will indeed take place. Nevertheless, this episode can be seen as the initiating of an uncertainty (will Charlemagne heed the Emperor’s call?), which will be resolved in the next panel.

Note that in the Descriptio, this segment begins in medias res, with the envoys arriving at Charlemagne’s court. The dream then enters the discourse analeptically through the reading of the letter from the Emperor. This has led some authors, including Mâle to regard panel 02 as the opening panel in the visual discourse. Personally I prefer to see panel 03 as the opening partly because this allows for a more consistent boustrophedonic pattern of reading and partly because, in visual terms, the leftwards momentum of the various elements in this panel encourage an onward reading in that direction. Although anisochronies are sometimes respected in visual discourses, their meaning is never as clear as in a text. Hence it is not unusual for this type of analepsis to be unravelled in polyscenic narratives to restore the natural fabula sequence, as happens here.

Panel 02: Charlemagne receives an emissary from the Emperor Constantine (Jerusalem Crusade, scene 2 of 6)

Extensive fictive architecture suggesting palatial interior setting - crenellated walls and a pitched roof over an arcade of trilobate arches. Three seated figures, their hand gestures indicating discourse. Central figure has a titulus reading “CAROLVS” directly beneath him. He also has a nimbus (sanctity), a crown (royalty) and a footstall (conventional sign for any authority figure). All of these elements contribute towards his identification as Charlemagne. Unusually he does not have crossed legs (a very common informant indicating royalty) though this may be because of his full frontal pose. The other two figures in the scene wear mitres, indicating their episcopal status. As Clark Maines pointed out, the differences in their poses suggest that the figure to our left is a member of the King’s court (his general posture, gesture and head angle all echo the King’s) whilst the other bishop is an emissary from the Emperor (he sits slightly apart, faces towards the other two and is making a more expressive hand gesture, as if imploring aid).25 The letter from Constantine, which is a key detail of this scene in the Descriptio, is not shown - the books carried by the two priests are just conventional attributes of clerics.

All of the pictorial elements are contained within the innermost borders - nothing overlaps the frame, contributing to the appropriately static feel of the composition. The titulus (which, as always in this window, is actually formed of clear-glass lettering emerging from a black-painted [108] ground) occupies a thin horizontal black strip, separated from the main pictorial field by a pair of red bands. The extensive area of black to the left of the titulus proper is filled with ornamental rinceaux, reminiscent of the pen-work used by scribes to fill incomplete text lines in manuscripts. The overall effect is to isolate this semantic enclave from the rest of the image field far more than is the case with the titulus in panel 03.

Following on from the Emperor’s dream in panel 03, his emissaries have come from Constantinople to Aachen (Aix-la-Chapelle) to ask for Charlemagne’s help in re-conquering the Holy Land. The key nucleus in narrative terms is his decision to embark on the quest, marking the resolution of the uncertainty initiated in the preceding panel. Although this segment does not come from the Pseudo-Turpin chronicle, the figure on the left is generally taken to be Archbishop Turpin. Given the medieval tendency to ascribe identities and meanings to everything and everyone in an image (a mental habitus shared by some twentieth century historians), this seems a very reasonable assumption and marks the first appearance within the visual discourse of the homodiegetic narrator (even though his narrative proper does not begin until panel 08).

Panel 05: Charlemagne and his army re-conquer the Holy Land (Jerusalem Crusade, scene 3 of 6)

Stylised bushes and uneven lumps painted with grass and weeds at the bottom of the image field serve as a conventional sign denoting an exterior scene. Above this, a dense, visually confused composition has been used to evoke the chaos of a mêlée. Very strong movement from left to right (indicated both by the poses of the horses and by the way the pennant and foremost knight’s surcoat billow out to the left) suggesting a rout. There is the usual contrast between closed helms and kite-shields (over the foremost knight’s left shoulder) for the Christians and conical open-faced helmets and round shields for the Muslims. A related internal index is the contrast between the ‘modern’ chain-mail hauberks of the Christians and the anachronistic lorica squamata armour worn by the Muslim cavalrymen here and also in panels 12, 14, 17 and 19. Lorica squamata, a type of scale mail used mainly during the (pre-Christian) Roman Republic, often features in medieval narrative images as a conventional sign of ‘otherness’ - particularly of pagan identity but sometimes just as an index of antiquity, as for example with the soldier attending Abraham and Melchizadek on the inside of the west façade at Reims [109] (see Fig. 4.25C). 26 In reality, throughout the period of the Crusades, the Saracens wore the same type of chain mail as the Europeans.

The yellow cloud-frill at the top of the panel is an oddity, as discussed above.

The artist has used the {picture} frame to contribute towards a sense of the chaos of battle.  The axe of one of the Saracens trying to save his king and the head of another are protruding over the border upper right, while the hind feet of the Saracen king’s horse overlap the white border strip lower left, lending emphasis to its already over-extended pose as its rider attempts to flee. The position of this panel, one of the left hand quadrants, is similar to that of the other pitched battle scene, in panel 14 - one of a number of ‘positional echoes’ throughout the window.

Although the Descriptio details multiple battles, the visual discourse has reduced these to a single scene (i.e. the frequency is nF:1D - see chapter 2.b.4), focussing on the decisive moment when Charlemagne slays his opposite number. In addition to frequency, the narrative pacing also shows a relative disinterest in the battle scenes - they occupy a significant proportion of the textual narrative but only one panel out of six in the visual discourse (this corresponds to Gerard Genette’s ‘summary’ mode of anisochrony - see chapter 2.b.3). One could compare this with the Morgan Picture Bible (Pierpont Morgan Library, Ms M. 638), in which almost every battle in the Old Testament is lovingly detailed, in some cases over multiple {picture} frames.27

Panel 04: Constantine welcomes the victorious Charlemagne to Constantinople (Jerusalem Crusade, scene 4 of 6)

Three figures (two crowned, one younger and bare-headed) stand alongside the conventional depiction of  a town. Fresh from the re-conquest of Jerusalem, Charlemagne (central figure) is welcomed by Constantine (right), who places a fraternal arm around him, suggesting that the Emperor sees the northern king as his equal. Both men wear identical crowns (the cross-bar of Charlemagne’s is present - just rather obscured by the restored leading). With his left hand, Charlemagne fingers the strap of his cloak;  a very common but poorly understood  (and probably multivalent) gesture, though certainly not one ‘...which in the 13th century was reserved almost exclusively for royalty’.28   Often this strap-tugging seems to denote modesty or unease - for example, a little further around the north ambulatory in bay 21 the (far from royal) [110] young St Julien and his wife both make this gesture in the scene of their marriage  (see Fig. 4.25B), as does one of the Pharisees being put in his place by Christ at the start of the Good Samaritan window (bay 44, panel 04 - see Fig. 5.01). To the left of the two rulers is a young male figure, bare-headed and nobly-attired. Although there are no internal clues as to the identity of this character, for reasons discussed below under panel 07, it is normally assumed to be Roland.

The right hand side of the panel is dominated by fictive architecture denoting Constantinople. Whilst it would be tempting to see the powerful encircling wall as an iconic depiction of the Theodosian walls and the dome over the portal as suggesting the Hagia Sophia, this would be wishful thinking. In fact the details of the fictive architecture seen here are purely generic; the formula of arched portal and crenellated encircling wall also appears in panel 12, as well as in several other windows at Chartres and elsewhere. The dome over the portal is also a generic element within the conventional sign complex denoting a city - see for example the depiction of the city of Sens in panel 13 of bay 17 at Chartres (Fig. 4.25A).

As he is welcomed into the city, Charlemagne is still wearing his spurs - a subtle analepsis suggesting that he has just arrived (the artists at Chartres were adept at adding such clues as to relative chronology, perhaps the most effective being the scene of the newly parricidal St Julien wiping the still wet blood from his sword in bay 21, panel 19 - see Fig. 8.28C). The spurs are gone in panel 06 (suggesting he’s had time to settle in) but reappear in panel 07, where the King is freshly arrived at Aachen.

The cloud-frill is present once again (at the top-left of the frame), as is the titulus (‘CAROLVS’), in a black band beneath the King’s feet. This time the titulus has no additional framing bands around it, although it does have decorative rinceaux filling the space to either side of the text. It is partly obscured however by the feet of Charlemagne and Constantine, slightly destabilising the separation between the two semiotic fields.

To the left of the frame, ‘Roland’ is touching the red band encircling the panel with his right hand and is stepping out of the frame with his right foot (he also gestures off stage right with his left hand). All this seems to indicate a general movement towards panel 05 which at least raises the possibility that the correct sequence of reading is actually 02-04-05-06 - and therefore that this scene is Constantine welcoming Charlemagne on his way to the Holy Land.

In narrative terms, this image is unusual in that it is not a nucleus but just a catalysis - it involves no decision points in the plot, simply a continuation. Its importance for the overall story, and presumably the reason for its inclusion, is that it shows Charlemagne being received as the equal of the Byzantine Emperor - a proleptic external index of his own eventual [111] coronation as Holy Roman Emperor on Christmas day 800. It would be hard to imagine any other reason why this scene is accorded such a dominant position on the central axis.

Panel 06: Constantine offers Charlemagne great riches - but he will only accept some holy relics (Jerusalem Crusade, scene 5 of 6)

Elaborate fictive architecture indicating an interior scene. The prominent golden cross top right denotes a church or cathedral setting, whilst the oil lamp suspended upper centre is an iconic sign that frequently serves as a conventional sign for a place of special sanctity - either a chapel or a shrine.

In the left foreground is an altar or shrine on which rest three gable-ended reliquaries, partly covered by a cloth (though it seems not to have attracted scholarly attention, this cloth may be a rare depiction of the performative aspect of ‘showing relics’, suggesting that the reliquaries were normally kept covered and were then unveiled before the expectant gazes of pilgrims or honoured guests). The outer {picture} frame is as usual used to create simple planes of depth - the fictive architecture passes behind it whilst the reliquaries in the foreground pass in front of it.

To the right of the scene are two crowned figures - the one in the centre is identified by the ‘CAROLVS’ titulus beneath his feet (unframed and without decorative scroll-work) as Charlemagne. He touches his companion with one hand and gestures towards the reliquaries with the other. Although Bulteau and others have identified this figure as Constantine (in the centre) offering the relics to Charlemagne (on the right), Maines correctly points out that in all other panels, where a titulus is beneath a character it serves to identify that person.29

This identification of the central figure gesturing towards the reliquaries as Charlemagne also fits more closely with the Descriptio version of the story. Constantine offers Charlemagne great riches and new territories but he refuses all these and instead asks only for some relics - here we are seeing the second part of that dialogue. Not only does this interpretation fit better with the details of the Descriptio, it also suits one of the broader moral implications both of that text and of the window, namely that a king should pay more heed to spiritual riches than to financial ones.

Overall, this is one of the key nuclei in the ‘Jerusalem Crusade’ segment. As well as serving as an index of Charlemagne’s modesty and piety, it sets up the all-important donation scene in panel 07. [112]

Panel 07: Charlemagne presenting relics to the Church (Jerusalem Crusade, scene 6 of 6)

Within a rich fictive architectural setting, Charlemagne (identified by his crown and titulus) kneels to present something at an altar, alongside which stand two bare-headed and tonsured clerics (one of whom carries a book and a crosier). Behind the king stands a young man plucking at his cloak-band, who is not explicitly identified but whose pose and appearance match those of the figure occupying the same position in panel 04 (directly below this scene). Hanging from a tie-beam above Charlemagne is an oliphant (an elephant-tusk horn) with gilded bindings. Given this object’s special association with Roland in the Pseudo-Turpin and elsewhere, its inclusion here greatly strengthens the case for identifying the figure on the left, and in panel 04, as the king’s nephew. In which case the horn here serves as a proleptic internal index, referencing Roland’s tragic fate as shown in panel 19 (which is also positioned on the central axis, four registers above this scene).

The microarchitecture above the altar is unusually elaborate (when compared to the other windows at Chartres) and depicts a large baldachin with trilobate arches, over which are displayed three gabled reliquary chests. Claudine Lautier has argued that this structure is a ‘simplified representation’ of the tribune des Corps-Saints in Chartres Cathedral; a high platform that once stood behind the high altar, and which is described in early-eighteenth century accounts as ‘a platform on which compartments containing chasses and reliquaries were arranged in a pyramid’. 30 She goes on to claim that: can recognize in the architectural setting of the scene not an image of the Emperor’s palatine chapel, but indeed a representation of the cathedral of Chartres, with the tribune des Corps-Saints, the main arcades and the colonnettes of the triforium behind the high altar and tribune.31

Owing to the limitations and conventions of the medium, there are always problems when trying to associate fictive architecture in stained glass with specific real buildings. In this particular case, there is the fact that the gabled reliquaries on the platform seen in panel 07 are very obviously not arranged in a pyramid (seemingly one of the distinctive feature of the tribune des Corps-Saints). Instead the reliquaries are displayed in niches on a single level, beneath a domed roof resting on colonnettes (just visible in the blue glass between the reliquaries - this detail is much clearer in an enlarged and slightly overexposed photograph - see Fig. 4.26A). Given the practical considerations of work on a medieval building site, it is questionable whether such an elaborate microarchitectural structure as the tribune would have been erected so early in the building campaign that it could have been seen and drawn by the painters of the choir windows. [113] On the other hand, the intention to construct such a feature may have been sufficient for the window’s designers to anticipate its general appearance. Under the circumstances then, we should not expect the relic platform in panel 07 to be an accurate picture of this now-lost structure, even if that was the original intention - and nor should it be treated as such.

The fictive architecture behind the baldachin is also quite distinct from what appears in all of the other windows at Chartres - nowhere else do we find columns on top of an arcade like this. As Claudine Lautier suggested, this may well be an attempt to depict the very building in which the window was to be set. Set against these Chartrain architectural elements however are a couple of details which point back to Aachen as a narrative setting. Behind the altar is a very distinctive processional cross with a circle at its centre and elaborate finials on the arms. This is quite unlike the many other crosses which appear in the windows at Chartres and seems to be a reference to Aachen’s famous Crux gemmata, the Lothar Cross, with its distinctive Augustus cameo in the centre. Like the oliphant, this detail would locate the setting as Charlemagne’s Palatine Chapel and not Chartres, or indeed the Abbey of St Denis, to whom, according to the Descriptio, Charlemagne donated the relics brought back from Constantinople. Most likely therefore we are looking at a fictive architectural setting which is meant to resemble Chartres (and thus help re-localise the story for the viewer) but to represent Aachen.

Contrary to what has been written elsewhere, and despite their similarity, the reliquaries shown on the baldachin here are not identical to the ones in panel 06 (see the enlarged details of both sets in Fig. 4.26 A and B) and it does not necessarily follow that ‘...they must refer to those depicted in the preceding panel’.32  Indeed the fact that the king is shown in the act of presenting something quite different from any of those reliquaries would seem to mitigate against this conclusion.

What exactly Charlemagne is presenting in panel 06 has been the subject of some disagreement in the past, thanks in part to the poor quality images from which most authors have had to work. Paul Durand, who wrote a commentary on this window to accompany his line-drawing in Jean-Baptiste Lassus’ gargantuan nine-volume Monographie de la Cathédrale de Chartres (Paris, 1842-65), suggested that it might be the crown of the Saracen king slain by Charlemagne in panel 05. 33  This is a perfectly reasonable suggestion, which would make the offering a trophy rather than a relic, however when seen in close-up the two objects are quite different (compare Fig. 4.26C&D). Others, including Maines, correctly identified this as a crown-reliquary - perfectly suitable for containing a part of the Crown of Thorns. However it seems to have been generally missed that Charlemagne is actually holding two separate objects - a crown in his [114] right hand and something smaller in his left (see detail in Fig 4.26C). His hands and the offerings are all painted on a single piece of glass with no intervening leading, so it is clear that the artist has deliberately painted two separate objects. According to the Descriptio, as well as the Crown of Thorns, the relics which Charlemagne brought home from Constantinople included a piece of the True Cross (for which one would expect a cross-shaped reliquary), various fabric relics and one of the nails from the Crucifixion. The last of these would seem to be the most likely identification for the object in his left hand.

All of the elements in the image are fully contained within the {picture} frame except for Charlemagne’s right foot, which protrudes lower left, helping to accentuate his elegant pose - and perhaps also suggesting that he has rushed here to present his treasures to the Church as quickly as possible (an interpretation helped by the fact that he is, again, shown wearing spurs). The titulus (‘CAROLVS’) is positioned directly beneath the kneeling figure, its beginning slightly obscured by the frame, implying that the visual text belongs to the fictive world enframed and not to the artifice of the frame itself.

If, as seems likely, one of the main objectives both of the composition of the Descriptio and of the making of this window was to authenticate and valorise relics, then clearly this is one of the most important panels. Seeing the story of a relic’s translation visualised in close proximity to the reliquary that contained it must have had a powerful effect on its audiences. Equally important to the overall narrative however is the emphasis in this panel on Charlemagne’s piety and generosity towards the Church. Not only is he giving his hard-won treasures to the Church but he is also humbling himself before the altar.

The presence of Roland in this panel (as suggested by the oliphant) and in the one below it (panel 04) is significant. In the Descriptio, the king’s nephew is never mentioned - Roland only enters our composite fabula in the text of Pseudo-Turpin. His inclusion here in the earlier phase of the visual discourse is perhaps an attempt by the window’s designers to unite the two parts of the legend into a more coherent narrative.34

Panel 08: Charlemagne contemplating the Milky Way (Segment 2 - First Spanish campaign, scene 1 of 6)

Charlemagne is shown seated on a faldstool at the left of the panel, with crown, nimbus and sceptre. He fingers his cloak-band and looks upwards to an elaborate cloud-frill that stretches across the top of the panel. Facing him on the right are two male figures dressed as noblemen, one of whom wears a skull-cap similar to the one in panel 18. All three figures look up at the [115] sky - one of them raises his hand to point at what is obviously the topic of their conversation. Apart from the king’s faldstool and two footstools, there is no microarchitecture, furnishings, trees or other informants to indicate the setting (though since they are all looking at the sky, an exterior setting is implicit). Beneath the king is an unframed titulus reading ‘CARROLVS’ with scroll-work completing the line. Interestingly, the King’s name is written with a single ‘R’ in panels 2, 3, 4, 6 & 7 (Jerusalem Crusade) but with ‘RR’ in panels 8, 11 and 13 (Spanish campaigns), although it appears to be painted by the same hand in all cases (for convenience, all eight versions of the titulus are reproduced in isolation in Fig 4.27). It is conceivable that two different advisors were providing the instructions for the window - one responsible for the Jeruslaem Crusade section who copied down the titulus with one ‘R’ and another for the Pseudo-Turpin panels who wrote it with the double ‘R’ - however a simpler explanation is that the artist made a copying mistake in panel 08 and then carried it over to the later panels.

Chapter II of Pseudo-Turpin opens with an analeptic summary explaining how the body of the martyred Apostle St James was rescued from Herod by his followers and carried by sea to Galicia, which he had converted to Christianity but whose people had subsequently reverted to paganism. In a brief narrative summary, the subsequent seven centuries or so are reduced to few lines of text, which get us to the point where Charlemagne has restored most of Europe to Christendom and is settling down for a quiet retirement. His peace is interrupted by a vision of...

...the starry way in the Heavens, beginning at the Friezeland Sea, [...] and from thence in a straight line over Gascony, Bearne and Navarre, and through Spain to Galicia, wherein till his time lay undiscovered the body of St James...35

This scene of Charlemagne consulting with his wise men about the significance of the Milky Way, which begins the second phase of the visual discourse at Chartres, does not appear in Pseudo-Turpin. As mentioned above (section 4.c.3), this scene must have been interpolated in order to provide an echo of the opening scenes of the first segment and as a means of introducing uncertainty into the plot (before St James’ appearance in the next panel resolves that uncertainty).

On a purely formal level, the upturned faces and gestures of the figures in this panel also help to create an upwards momentum at the start of this new phase in the narrative.

Panel 09: St James appears in a dream (First Spanish campaign, scene 2 of 6)

Fictive architecture suggesting an interior scene. The king (identified by his crown and halo but this time without a titulus) lies asleep on his bed in the foreground. Behind him stands a nimbed, bearded male figure with no obvious attributes - the piece of glass on which his left hand is [116] painted has been inverted at some point during restoration, but even in close-up it is not possible to identify the object it holds (the piece may just be a stop-gap). This figure’s identity as St James can be inferred only from the narrative context.

According to Pseudo-Turpin, night after night Charlemagne ponders the significance of the starry band that stretches from east to west across the sky, until one night St James appears to the king in his dream and introduces both himself and the next phase of the quest:

‘My body now lies concealed in Gallicia, long so grievously oppressed by the Saracens, from whose yoke I am astonished that you, who have conquered so many lands and cities, have not yet delivered it. Wherefore I come to warn you [...] to prepare my way and rescue my dominions from the Moabites, so that you may receive a brighter crown of glory for your reward.’ [...] Thus did the blessed Apostle appear thrice to the Emperor, who, confiding in his word, assembled a great army, and entered Spain to fight the infidels 36

Note that as is often the way with dreams and visions, in the text St James’ message is repeated thrice, though as usual it is only depicted once in the window.

From a narrative point of view this episode has several functions - it resolves the uncertainty initiated in the preceding panel (the significance of the Milky Way), provides divine sanction for the new quest and justifies the forgiveness of Charlemagne’s unspeakable crime (see panel 22). Formally it also echoes the position and composition of  the dream sequence in panel 03 that opened segment 1 (see Fig. 4.23).

Panel 10: Charlemagne’s army set out for Spain (First Spanish campaign, scene 3 of 6)

A tightly clustered group of men on horseback ride from right to left across the panel. In the foreground is Charlemagne, crowned but without nimbus or titulus, riding a chestnut stallion (in the latter anatomical detail, the artist has been unusually specific). He turns in his saddle to talk to a companion wearing a mitre who, by implication, must be our putative narrator, Archbishop Turpin. This is a relatively rare example of the inclusion of the homodiegetic narrator within a visual discourse.

Lejeune and Stiennon identified the youthful rider at the head of the group as Roland on the grounds that he carries a ‘curved sword’, which they suppose to be the legendary weapon Durandel. As Maines points out however, the weapon wielded by Roland elsewhere in the window (panels 17 and 19) is anything but curved.37  In fact, a more careful examination of the rather narrow and insubstantial object held by this rider reveals that it is almost certainly a riding crop (and so at most an index of the rider’s enthusiasm for the crusade to be underway). [117] Furthermore, the head of this figure is clearly not original so there is no guarantee that he was originally so youthful in appearance. Nevertheless, given the prominence of Roland in Pseudo-Turpin’s account and elsewhere in the window, his presence here at the head of the group can probably be assumed.

The only informants relating to setting are the stylised trees at the right hand edge of the panel and the green wavy line of the ground at the bottom, indicating a generic external scene. In terms of framing, the hooves of Charlemagne and Roland’s horses overlap the frame slightly, whilst Turpin’s horse is cut off by it, indicating the relative depth of the visual planes but also contributing to the sense of movement across the frame from right to left.

The artist has constructed his group with an admirable economy of effort. We can actually see part of the heads of six riders, on three horses, which share between them eight legs. It is a measure of the success of visual gap-filling and the conventionalism of medieval art generally that the viewer simply registers this as ‘a group of horsemen’ rather than worrying unduly about the occluded details.

All other things being equal, one normally finds that journeys depicted within a visual narrative proceed from left to right. When this direction of travel is reversed it is normally because the artist wants to guide the viewer leftwards to the next panel in the sequence, as is the case here.

Panel 11: Charlemagne praying in front of his army (First Spanish campaign, scene 4 of 6)

Charlemagne, dressed in full chain mail and surcoat, identified by his crown, nimbus and titulus has dismounted (the saddle of the chestnut horse behind him is empty) and kneels in prayer, facing upper-right. Behind him a tightly pack group of mounted knights with closed helms and kite shields are waiting impatiently - note how the second horse paws the air while his rider lowers his lance, from which flutters a red, multi-ribboned banner. Lejeune and Stiennon identify the latter as the Oriflamme (and by extension, identify its bearer as Roland). 38 Clarke Maines questions this identification on the basis that the banner here is different from the one in panel 03, although this criticism overlooks the fact that in early thirteenth century stained glass, objects do routinely change their colour and appearance between panels (something which is itself an interesting reflection on the nature of iconicity in the period). A stylised tree indicates an exterior setting, while the cloud-frill continues to hover at the top of the panel. Unusually for the panels in this window, Charlemagne’s titulus floats unframed and slightly askew, in the empty space above him. [118]

The narrative role of this panel will be addressed below since it cannot be assessed in isolation from panel 12.

Panel 12: The taking of a city - possibly Pamplona (First Spanish campaign, scene 5 of 6)

To the right is the standard image of a fortified city seen from outside, comprising a jumble of walls, roofs and turrets, with a crenellated enclosing wall (this highly conventionalised depiction is extremely common across several of the windows at Chartres and is very similar to depictions in the windows of other cathedrals and also in other media). A yellow pennant (presumably this time not the Oriflamme) flutters over the gate and a man blows a horn in alarm (again, presumably not the oliphant) from one of the turrets. Meanwhile, a mounted Saracen (lorica squamata armour, open helmet, round shield slung over his back) is fleeing into the city gate, turning as he does to look at a Christian knight (closed helm, full chain-mail and red surcoat, kite-shield), who has pierced him with a lance. The crusader’s horse is sharply cut-off by the frame from which it emerges on the left which, together with its galloping front legs frozen in mid-air, helps to give the scene a strong sense of frenetic movement. The pursuing knight has neither crown, nor nimbus, nor titulus and hence is unlikely to be the king (who sometimes appears without one or other of these attributes but never without his crown).

All parts of the image are contained within the frame, except the end of the horn being sounded from the turret, which overlaps the border slightly. As with the preceding panel, there is a cloud-frill occupying the very top corner of the frame. As several people have pointed out, although other parts of the composition differ, the detail of a rider fleeing into a city gate and turning to look at his attacker is almost identical to the one in a now-lost panel in the mid-twelfth century ‘Crusader window’ from the Abbey of St Denis, recorded in a Montfaucon drawing (see Fig. 4.28C).39

Panels 11 and 12 remain somewhat contentious. The obvious interpretation, that it is the siege and fall of Pamplona (the most important conquest in the campaign) appears initially to be contradicted by some of the details of panel 12. In Pseudo-Turpin, the key point about the taking of Pamplona was its miraculous nature; Charlemagne and his armies besieged the town for three months but were unable to overcome its invincible walls. The king then prayed for help, whereupon  ‘...God and St James hearkening to his petition, the walls fell utterly to the ground of themselves; but Charles spared the lives of the Saracens that consented to be baptised, the rest he put to the edge of the sword’. 40 One of the important consequences of this divine [119] intervention was that the Saracens were so impressed by the manner of Charlemagne’s victory and the honourable behaviour of his troops afterwards that ‘...they gave them therefore a peaceful and honourable reception, dismissing all thoughts of war’ (ibid). The narrator goes on to tell how Charlemagne proceeded westwards to the sea and that the ‘Pagan nations’ were converted (by him, Turpin), and baptized, while those who refused to convert were killed or enslaved. The artists who depicted the scene in repoussé copper on Charlemagne’s shrine in Aachen followed Pseudo-Turpin diligently (see Fig. 4.28A). There  the conventionalised city is surrounded by siege tents; on the left of the plaque Charlemagne is shown kneeling in prayer, just as he is in panel 11, but on the right his prayers are seen to be answered as the hand of God appears from the sky and sends the walls of Pamplona tumbling. By contrast panel 12 appears to show a city being taken by force.

Emile Mâle, following Bultheau, interprets panels 11 and 12 as the taking of Pamplona, although he does note in passing that ‘the artist of Chartres, unlike the Aix-la-Chapelle artist, did not represent the walls of Pamplona crumbling under God’s hand’. 41  Lejeune and Stiennon agree with the identification of panel 11 as Charlemagne praying before Pamplona but for rather dubious reasons choose to interpret panel 12 as Roland’s conquest of the city of Noples (subject of a tangential reference in the Chanson de Roland and supposedly the theme of a now lost Chanson de geste).42 Clark Maines on the other hand sees panel 12 as a generic scene of conquest:

The chapter following the capture of Pamplona in the text contains a list of those cities taken by consent and those taken by force. In the absence of any textual basis to interpret the scene specifically, it seems reasonable to suggest that panel 12 is a generalised scene representing all those cities captured by force. The choice of the more violent of the two possibilities is entirely in keeping with the generally violent tenor of the panels.43

Although there is much to be said for this position, the ‘generally violent tenor’ of the Charlemagne window needs to be approached with caution; only five of the twenty-two main panels in this window depict acts of violence, compared with almost every panel in the St Denis ‘Crusades’ window. As mentioned elsewhere, the general emphasis in this window is on Charlemagne’s piety, rather than his martial prowess. Even the list of cities ‘taken’ by Charlemagne is of limited help. The largely achronic chapter found in some versions of Pseudo-Turpin (Chapter V in MS.17656) and cited by Maines, is essentially a gazetteer comprising a long list of fourteen cities and around a hundred other locations taken by Charlemagne, but the verb used throughout that list is adquirere (to acquire or win), rather than vincere (conquer or defeat), which does not readily match with the violence of panel 12. [120]

In reality, the window does not contain sufficient information to make any confident claim about the nature of panel 12. It may be a generic representation of the towns taken by force - but then again there is nothing in the image specifically showing that the town has been taken successfully. This may simply represent the unsuccessful three month siege of Pamplona (so the sequence of reading the panels could be 09-10-12-11-13 instead of  09-10-11-12-13).

Certain elements in the image suggest a possibility that the Chartres artist had seen the Aachen panels or else that both artists were drawing on a common (and lost) source. On the other hand the visual topoi associated with scenes of warfare may be sufficient to account for such similarities on their own.

Panel 13: The building of a church (First Spanish campaign, scene 6 of 6)

The right hand side of this circular medallion is filled with a church-construction scene showing the disregard for scale that is normal for the period. A conventional stylised church building (but with a distinctively Gothic flying buttress system) is being assembled by two masons working at the clerestory level. In the upper centre of the panel is the master-mason, marked out by his surcoat, who is fitting a masonry block in place (this is not, as has been suggested, a piece of parchment or bifolio with the plans of the church being shown to the patron - it is simply a block of masonry with string-course mouldings around its base, exactly the same as the ones on the stone at the level of the mason’s knees). To the right his assistant, wearing a close-fitting coif and no surcoat, checks the position of a now-missing block using a plumb-level. In the foreground, two labourers (bare-headed and wearing short cottes) are carrying another dressed stone up the wicker scaffold using a hand barrow. To the left of the panel is Charlemagne, now dressed in rich robes rather than armour and distinguished by his crown and nimbus, who sits on his horse and directs operations. His gaze and raised index finger seem addressed to the master mason. At the top of the frame once again is the cloud-frill while near the bottom in a black band marking out the ground level of the scene is the titulus (‘CARROLVS’), with decorative scroll-work filling the space to the right.

According to Pseudo-Turpin, after liberating Pamplona, Charlemagne remained in Spain for three years, during which time

...with the gold given him by the kings and princes, he enlarged [augmentavit] the church of the blessed St James, appointing an Abbot and Canons of the order of St Isidore, martyr and confessor, to attend it: he enriched it likewise with bells, books, robes and other gifts.44

He then adds that with the money left over, on his return from Spain, Charlemagne built many churches and founded ‘innumerable abbeys in all parts of the world’. [121] Conceivably then, this image could represent either the ‘enlarging’ of Santiago Cathedral, the building of some other specific church, or perhaps a generic image representing Charlemagne’s church-building activity in general.45 Most authors opt for the first of these, although some mistakenly describe it as the ‘building’ of the church of St James, rather than its rebuilding. Clark Maines however strongly disagrees, arguing that:

...according to Pseudo-Turpin, Charlemagne did not build the church of Santiago. He only ‘enriched’ it and installed there the order of St Isidore.46

Unfortunately this is a rather selective reading of the text, as the longer quotation given above makes clear. In medieval church-building terms, ‘to enlarge’ (augmentare) could imply fairly substantial changes, if not complete rebuilding. For example augmentavit is also the word Abbot Suger uses to describe his changes to the choir of St Denis in his De Administrationis. Rather than Santiago Maines wants panel 13 to represent Charlemagne overseeing the construction of a Benedictine Abbey dedicated to Saints Facundus and Primitivus (on the site of the Battle of Sahagún - see following panel),  an establishment now relatively unknown but apparently the heart of Cluniac influence in Spain at the time. On the whole though this seems unlikely. Apart from the re-ordering of panels it would require, Facundus and Primitivus have no obvious connection to Chartres, while the story of St James (whose church at Santiago was one of the main pilgrimage sites in Europe) occupies the ambulatory window immediately to the right of this one (bay 05). Crucially though, it was the nocturnal visitation by St James (panel 09) and his request that Charlemagne wrest his resting place from Pagan hands that initiated the narrative segment of which this is the conclusion. The identification of panel 13 as the rebuilding of the Cathedral at Santiago de Compostela therefore seems secure and, as the fulfilment of the quest with which St James charged Charlemagne in panel 09, it provides a satisfying degree of narrative closure to this phase of the story.

Panel 15: The Miracle of the Flowering Lances (Second Spanish campaign, scene 1 of 7)

In a distinctive and highly original composition, the artist has shown a jumbled group of knights at rest, sleeping against their large kite-shields. The variety of poses is remarkable and a world away from the more conventional images of sleeping soldiers (such as are often shown beneath Christ’s tomb in Resurrection scenes). Two of the sleepers have their faces uncovered, revealing bearded chins and closed eyes - the others wear closed-face helms. Behind the sleepers, seven lances are lined up, their points upwards, with leaves sprouting from all along their shafts. A larger than usual cloud-frill appears at the top of the frame, echoed at the bottom by the flicker [122] of a red pennant. The impression of a group of soldiers sleeping uneasily before battle is enhanced by the figure stretched out lower left, whose surcoat flaps out over the edge of the frame.

According to Pseudo-Turpin, after rebuilding the church at Santiago, Charlemagne and his armies returned home to France. A Pagan king, called Argolander, then recovered Spain, driving out the Christian garrisons and prompting Charlemagne to march his armies west again, thus beginning the second Spanish campaign. The two great armies met at Sahagún (on the site where Charlemagne would later build the Abbey of Facundus and Primitivus) and after three indecisive days of set combat, prepared for an all-out battle on the fourth day. Pseudo-Turpin’s account of what happened that night is worth quoting in full:

Then did this miracle happen. Certain of the Christians, who carefully had been furbishing their arms against the day of battle, fixed their spears in the evening erect in the ground before the castle in the meadow, near the river, and found them early in the morning covered with bark and branches. Those therefore, that were about to receive the palm of martyrdom were greatly astonished at this event, ascribing it to divine power. Then cutting off their spears close to the ground, the roots that remained shot out afresh and became lofty trees, which may be still seen flourishing there, chiefly ash. All this denoted joy to the soul, but loss to the body; for now the battle commenced, and forty thousand Christians were slain...47

Thus the knights seen sleeping here in front of their flowering lances are those who will die in the ensuing battle. Any attempt to ascribe particular identities to the individual sleepers would be purely speculative, though naturally this has not discouraged certain authors from doing so. Any putative identification as Charlemagne or Roland would however be particularly perverse since neither of them die the following day. Indeed the only major character named by Pseudo-Turpin among the forty thousand Christians who die in that battle is Milo; Charlemagne’s General and Roland’s father. Since Milo does not explicitly feature elsewhere in the visual narrative, it would be naive to argue that he must be represented by one or other of the sleeping figures. As discussed in Chapter 2, visual discourses can operate with far greater indeterminacy (or ‘gappiness’) than their textual counterparts. For a written chronicle it would be important to list the identities of the participants but the visual story-teller can paint his discourse with broader brush-strokes, leaving the viewer free to create identities for his anonymous characters at will.

In terms of the overall narrative progression, this scene marks the start of Charlemagne’s second Spanish campaign - the third narrative segment. As with the first two, this segment begins with a nocturnal miracle, this time a communal experience rather than an individual vision (see Fig. 4.23). [123] Once again, this miracle initiates an uncertainty in the plot - an uncertainty that will be bloodily resolved by the events shown in panel 14.

The topos of the dead branch that sprouts fresh life has rich and multi-layered significance in Christian narratives but the primary function of the miracle is nearly always to indicate that the bearer has been chosen by God for some purpose; as when the flowering of Aaron’s rod (Numbers 17:8) denotes that only the Tribe of Levi may provide the Priests, or when Joseph’s rod marks him out as the chosen partner of the Virgin Mary (an alternative function was to indicate a specially appointed place, as with St Etheldreda’s staff). However a deeper meaning of the topos - that God’s power can bring life from death - was particularly relevant when the unexpected flowering is indicative of impending martyrdom. Pseudo-Turpin’s claim that the ash-trees which grew from the spear-stumps ‘may be still seen flourishing there [at Sahagún]’ would have given the panel a special significance to any long-range pilgrims passing through Chartres on the Chemin de St Jacques - a preview of one of the treats they might get to see as they passed through León towards Santiago (the modern-day town of Sahagún lies on the A-231 or the Autovia Camoni de Santiago as it is better known).

Panel 14: The Battle of Sahagún (Second Spanish campaign, scene 2 of 8)

A pitched battle scene - charging in from the left is a group of mounted knights (kite-shields, full-face helms, chain mail hauberks), who are attacking a group of mounted Saracens (round shields, open-faced conical helmets, lorica squamata). In the foreground one of the latter falls limply from his tumbling horse, pierced beneath the arm-pit (the weak-point of most armour) by a crusader’s lance. A man falling from his horse is a common motif in medieval art as a personification of pride (one appears on the south porch at Chartres) but the artist here has made a clear distinction between those vigorously tumbling figures and this very dead rider. The group of Saracens, although fleeing, present a solid cluster of helmets, very different from the disorganised trio who try in vain to defend their king in panel 05. One of them aims a sword blow at his pursuers. The overall impression is that although the crusaders have the upper hand, it is not the one-sided rout depicted in the earlier battle scene. None of the crusaders are crowned or nimbed and although (inevitably) the lead rider has been identified by Lejeune and Stiennon as Roland, this is pure invention.48 Although the armorial bearings on shields in medieval art often function as a semantic enclave (expressing meaning through the ‘language’ of heraldry), that is not the case in this window, where the shield markings are inconsistent and [124] essentially decorative - generic markers of ‘shield-ness’ rather than heraldic identifiers of their owners.

A conventionalised tree in the background and some undulating ground at the bottom of the panel are included as rather redundant informants indicating an exterior setting. The cloud-frill is missing from this panel, which also lacks a titulus.

Functionally, this scene is the resolution to the uncertainty initiated in panel 15 - the knights whose lances flowered overnight are about to achieve their martyrdom. Clark Maines takes a different view from most commentators over this and the preceding panel. He sees panel 14 as a generalised battle scene, representing various battles that precede the flowering of the lances in panel 15. As he points out, the Christians in this battle seem to be getting the upper hand - and claims this would negate the narrative point of the battle, which ought to be the slaughter of those martyrs whose lances flowered. This argument is problematic however. For one thing, in Pseudo-Turpin there are no descriptions of any preliminary triumphs between Charlemagne’s return to Spain and the Battle of Sahagún. Nor is the fact that the fight appears to be going the way of the Crusaders necessarily a problem. Generally in the visual tradition one does not find scenes depicting the deaths of large numbers of crusaders during a battle which they are ultimately going to win (Charlemagne loses 40,000 men at Sahagún but emerges victorious). The simplest and most plausible conclusion is that panels 15 and 14 represent the miracle of the flowering lances before the battle of Sahagún and then the battle itself - neatly encapsulating the two key nuclei in chapter VIII of Pseudo-Turpin.

Panel 16: Combat between a Christian and a Saracen (Second Spanish campaign, scene 3 of 8)

Two mounted knights engaged in single combat, with lances levelled at the point of impact. Approaching the centre from the left is a Christian knight, with kite shield and closed helm (but no crown, nimbus or titulus), while from the right comes his Saracen counterpart with round shield and an open-faced, conical helmet, around which is a crown. Unusually, both fighters are wearing chain-mail and surcoats. The Christian has landed a blow on the shield of his opponent, who in turn has broken his lance on his. This presents an interesting conundrum for the modern iconographer. A broken lance normally has negative connotations (for example when held by Synagoga) but was a positive indicator in a joust, the result of a well-aimed blow. Thus to viewers versed in matters martial the meaning of this image would be clear; it shows a fight between two evenly-matched opponents.

A cloud-frill is visible at the top of the panel. Near the bottom is a strip of stylised ground, with a couple of very conventionalised plants - beneath this is a thin horizontal strip of red glass [125] which resembles the one used as a frame for the titulus in panel 02 - though here there is no titulus (nor any evidence that the plain blue glass beneath it is not original). Compared with the busy clutter of some of the surrounding panels, the simplicity of this symmetrical composition, together with its position on the central axis, helps it to stand out. Of all the panels in bay 07, this is the one which makes greatest use of overlapped borders; the hind quarters of both the charging horses protrude right out to either side over the red, blue and white encircling bands of the frame, greatly enhancing the sense of movement and dynamism. Interestingly, the frame violation stops there; the horses are cut off level with the lead cames containing the white band and do not intrude into the foliate patterns of the spandrels. This rather suggests that these central panels were originally conceived and designed as circles, rather than as lozenges - otherwise the artist could have enhanced the dynamic effect by continuing his horses further into the border before stopping, as usual, at the armature.

Following the Battle of Sahagún, Pseudo-Turpin tells how the Pagan King Argolander assembles a great army and takes the city of Agen (in Acquitaine), from whence after a long series of sieges and retreats, Charlemagne drives him back over the Pyrenees and defeats him near Pamplona.49 After various further skirmishes, Charlemagne receives news of ‘...a certain giant, of the name of [Ferracutus], of the race of Goliath’, who has come to fight him. When the two armies meet, Ferracutus challenges all comers to single combat. Having seen various of the King’s champions routed effortlessly, Roland takes the field himself and after both are unhorsed, he and the giant fight on, on foot. After several hours of inconclusive combat the two agree a truce until the next day. Meeting again the following morning, Roland asks the giant how he can be so strong, to which Ferracutus rather naively admits that his is vulnerable only through the navel. The two champions then have a very lengthy and civilised theological discussion (an example of the mode of anisochrony referred to in Chapter 2 as ‘stretch’) in which the unknown author of  Pseudo-Turpin uses his characters as mouthpieces to set out the key tenets of Christianity in dialectical form. Unconvinced, Ferracutus proposes they should settle their theological disagreements by trial of combat and the fighting resumes (panel 17). Roland, temporarily disarmed, seizes his opponent’s sword and slays Ferracutus by thrusting it into his navel.

Panel 16 has presented something of a conundrum to researchers and restorers alike. Prior to 1921 it was mounted in the topmost frame of the armature, in the position now occupied by panel 22 (see Fig. 4.29B for a digital reconstruction of the previous arrangement). The two panels were swapped around by Atelier Lorin during their restoration of the window in 1921, because [126] it was believed that this scene represented the combat between Roland and Ferracutus - and that it had been misplaced during earlier repairs (a very common occurrence with stained glass).50 The combination of a jousting scene followed by combat on foot also occurs in another supposed visualisation of the story, on the famous bronze doors of San Zeno in Verona.51

The counter-argument, proposed by the historian Alphonse Vétault in 1877 revolves around an alternative interpretation of  panel 16 as representing Roland’s combat with the Saracen King Marsile. As Clark Maines (one of the few to agree with this view) explains:

Vétault based this identification  on the crown worn by the Saracen and on the Pseudo-Turpin passage which describes Marsile as a king who carries a round shield. All subsequent scholarship has adhered to Bulteau’s earlier identification of the figure as the giant Ferracutus, presumably disregarding Vétault’s criteria because all Saracens in the window carry round shields and because the battle between Roland and Ferracutus  is of greater didactic importance in the written narrative. Vétault’s hypothesis, however, is confirmed by the roan or reddish horse on which the Saracen is mounted: according to the text, King Marsile is identifiable on the field of battle by both a ‘red horse and a round shield’.52

This ‘red horse’ is a rather weak argument, colour in stained glass generally being an unreliable signifier. Charlemagne is shown on reddish horses throughout the window, but so are other riders, including the Saracen king slain during the Jerusalem crusade (panel 05), while as Maines accepts, the round shield is a red herring. This identification of panel 16 moreover has to be considered alongside that of panel 17.  If that panel represents the killing of Ferracutus and panel 16 is in the correct position, then the obvious conclusion is that the two panels represent the single combat between Roland and Ferracutus (16), followed by the killing of the giant (17) - a pairing that produces a far more satisfying level of narrative closure. Two other objections that have been raised to this interpretation are that a) the Saracen in these panels is not obviously a giant, and b) he wears a crown over his helmet. Neither of these are serious problems. Differences of scale emphasised in texts are often ignored in visual discourses and the texts themselves are prone to hyperbolic topoi regarding the sizes of heroes and their enemies. For example, in the Codex Calixtinus copy of Pseudo-Turpin, Charlemagne himself is described as a ‘handsome giant, eight feet tall’, who could ‘split an armed mounted knight and his horse together’ with one blow of his sword, could ‘easily bend four horse-shoes at once, and could stand a knight-in-armour on the palm of his hand and lift him as high as his head’.53   As for the crown, although Pseudo-Turpin never explicitly describes Ferracutus as a king or prince, he [127] does describe him as leader of ‘twenty thousand Turks of  Babylon’ (Pseudo-Turpin, Chapter XVII); a sizable enough force, which might excuse the artist for adding a crown. Hampered perhaps by poor photographs, several authors failed to notice that the crown appears in panel 17 as well.

A more serious objection to the identification of panel 16 as the contest of single-combat between Roland and Ferracutus however, is that in Pseudo-Turpin, almost all of that fight takes place on foot, the two horses having been killed in round one. Furthermore, why does the knight in panel 17 have a nimbus but not the one in panel 16? Ultimately it seems unlikely that these two scenes will ever be entirely resolved to everybody’s satisfaction. Instead the safest interpretation seems to be that; a) panel 17 (see below) does indeed show the killing of Ferracutus by Roland, and b) that panel 16 may be the beginning of Roland’s battle with Ferracutus or his later fight with Marsile, or a conflation of the two. Perhaps the best interpretation is that panel 16 is an epitomic image which can represent either or both of those specific happenings - but which can also serve to evoke generic themes of combat against the Muslims/Pagans - and even more generally the virtue of the Miles Christi (for which Roland and Charlemagne were presented as exemplars). This would also explain why it is in one of the most visually charged positions in the overall composition of the window - two-thirds of the way up on the central axis. On the basis of these arguments it seems likely that panel 16 is indeed in the correct place (a point to which I will return when discussing panel 22).

Panel 17: Roland kills Ferracutus (Second Spanish campaign, scene 4 of 8)

A Saracen soldier (lorica squamata armour, conical open-faced helmet with crown) lies on his back, his sword limp at his side, while a standing knight (closed-faced helm, chain mail and surcoat, nimbed) drives a sword into his belly. The knight is shown leaning forward, with his left hand on the sword’s pommel, to emphasise the force with which he is pushing it into his foe. This particular form of coup-de-grace (stabbing through the navel) is hardly an honourable blow for either party and in the literature of the period is probably unique to the death of Ferracutus, something which aids greatly in the process of identification. It also appears in several other visualisations of the story in various media.

Conventionalised bushes/trees are shown to left and right of the frame, indicating an exterior setting. Two cloud frills appear at the top of the panel. The Saracen’s crown encircling his helmet strengthens the case for regarding this scene and panel 16 as consecutive events.

(For the further comments on this panel, see panel 16 above) [128]


Panel 18: Charlemagne heads homewards (Second Spanish campaign, scene 5 of 8)

The King (with crown, nimbus and robes instead of armour) is shown riding towards the right of the frame, together with four companions. Two of the latter resemble the men with whom Charlemagne converses about the Milky Way in panel 08 (and one of them wears a similar green skull-cap), though given the limited range of facial types in the artists’ repertoire, it would be unwise to read too much into this. Charlemagne is engaged in conversation with his companions - he gestures back in the direction from which they have come and half-turns that way. The figure to the right raises his right hand palm outwards (suggesting disagreement) and places his left hand on the neck of Charlemagne’s horse. The usual cloud-frill appears at the top of the panel, while the cutting off of two of the horses by the right hand frame accentuates the rightward movement of the group. Perhaps the most unusual element of this panel is the mass of stylised plants and undulating ground which rises almost vertically up the left side of the frame and from behind which the riders emerge. This extraordinary feature illustrates rather nicely the flexibility of the sign-systems of medieval art. Horizontal undulating pieces of glass, with grass or low vegetation painted onto them, are the standard way of illustrating open ground in medieval windows (the usual system of iconic signs functioning as conventional signs). Here in panel 18, those elements have been tipped up through ninety degrees in order to represent mountains. Their clear meaning is that Charlemagne and his companions have just crossed the Pyrenees on their way back to France.

After having re-conquered Spain for a second time and accepted oaths from the remaining Saracen kings, Pseudo-Turpin tells how Charlemagne began his return to France, leaving his nephew Roland to guard the mountain pass at Roncesvalles. Secretly however, one of his nobles, Ganelon, made a deal with the Saracen King Marsile, betraying his friends for money and leaving the rearguard in great danger. Since Turpin and Ganelon are the only companions mentioned explicitly in the text, most authors have assumed that the figure to the left of the panel is Turpin and the one on the right is the traitor Ganelon. Whilst the former identification is uncertain (Turpin is shown with his mitre in panel 10 and that would be a more obvious attribute for him), the actions of the other character, who seems to be dissuading the king of something and restraining his horse, are highly relevant to what happens in panel 19 (see below), suggesting that this is indeed Ganelon.

Panel 19: Roland’s fight at Roncesvalles (Second Spanish campaign, scene 6 of 8)

Although not the final panel of the third narrative segment, this is clearly its climactic scene in the visual discourse, having the same formal relationship to the rest of the segment as panels 07 and 13 (the other two fully circular medallions on the central axis) do to segments 1 and 2. [129]

The panel is divided internally into three main parts. In the upper-left section, a knight in a chainmail hauberk, with closed-faced helm and nimbus, is slicing a rock (on which stylised plants are growing) with his sword.  In the upper right section, the same knight has removed his helmet (it is the green object visible in front of his left foot) and is blowing a horn alongside further stylised trees. The entire lower part of the roundel is thickly strewn with what have been rather charmingly described as ‘...truncated Pagans in heaps’.54 Amidst the carnage, one can pick out two heads (one of which has been detached from its body) in the Saracens’ distinctive open-faced helmets, plus a further disembodied head, sans-helmet. Perhaps the most gruesome detail is in the foreground, where the lower half of a body clad in green lorica squamata armour disgorges neat coils of spilt intestines onto the ground - the red-helmeted head and torso immediately to the right are the upper half of this body. One of the literary topoi of  the romance epics is the opponent ‘cleft in twain’ by the hero’s sword - precisely what the artist has illustrated here.  At the top-centre of the panel the hand of God emerges from a small cloud-frill, conferring a blessing.

Pseudo-Turpin tells how, after Charlemagne’s main force had passed through the mountains, Marsile and his armies swept down on Roland and the rearguard at Roncesvalles, where after a bloody battle only Roland, Theodoric and Baldwin were left alive. Roland, separated from his companions, slays Marsile (which some authors unconvincingly argue is the subject of panels 16 and possibly 17) but is greatly outnumbered and is mortally wounded. Finding himself alone and surrounded by Saracens, Roland is reluctant to see his beloved sword Durandel fall into the hands of some ‘slothful timid soldier’ and so, after what amounts to a love-song to his sword, tries in vain to smash it against a block of marble - instead cleaving the rock in half. This is clearly the event depicted in the upper left section of panel 19. He then takes up his famous horn, the oliphant (which we last saw hanging proleptically in the church like an ex-voto in panel 07). Pseudo-Turpin describes what happens:

He now blew a loud blast with his horn, to summon any Christian concealed in the adjacent woods to his assistance, or to recall his friends beyond the pass. [...] The sound reached the King’s ears, who lay encamped in the valley, about eight miles from Ronceval, towards Gascony, being carried so far by supernatural power. Charles would have flown to his succour, but was prevented by Ganelon, who, conscious of Roland’s sufferings, insinuated it was usual with him to sound his horn on light occasions. ‘He is, perhaps,’ said he, ‘pursuing some wild beast, and the sound echoes through the woods; it will be fruitless, therefore, to seek him.’ O wicked traitor, deceitful as Judas! What dost thou merit?55

This second part of the story now explains the exchange of gestures in panel 18 whereby Ganelon seeks to restrain the King from answering Roland’s horn (the additional external index, [130] to the treachery of Judas, being an extra bonus for the narrator, or to anyone interpreting the window).

In terms of visual narrative technique,  panel 19 is clearly an example of the simultaneous or complementary mode, with multiple {chronotopic} frames and extra-narrative elements depicted in a single {pictorial} frame. The corpses in the foreground serve as informants, giving clues about the physical setting of the actions shown above them -  although they are also operating as an analeptic index, telling the viewer about the fierce fighting that preceded these actions. The hand of God is more difficult to explain and does not obviously correspond to anything mentioned in Pseudo-Turpin. Lejeune and Stiennon imagine it to represent the supernatural power that carries the horn’s sound to Charlemagne but this seems unlikely given its position at the top (rather than on the right) of the panel.56 I would argue instead that the hand of God here is a proleptic index, showing divine acceptance of Roland’s imminent death as an act of martyrdom. Even though he is still alive in the next scene, this panel, marking the climactic moment in the narrative, is a far more suitable place in which to mark Roland’s accomplishment.

The chronological relationship between this scene and panel 18 is particularly fascinating if one considers the precise actions involved. In effect we have six distinct {chronotopic} frames bouncing around between two {picture} frames:

  1. Charlemagne and his main army descend from the pass (panel 18). Logically and chronologically, this has to happen before...
  2. ...Roland and the rearguard are attacked and a bloody battle ensues (analeptically indexed by the corpses in panel 19 lower)
  3. Roland tries to break Durandel (panel 19 upper left)
  4. Roland sounds the oliphant to summon aid (panel 19 upper right)
  5. Charlemagne starts to go to his aid but is dissuaded by Ganelon (panel 18)
  6. Roland’s martyrdom is divinely approved (proleptically indexed by the hand of God in panel 19 top)
    (The story then jumps to panel 20)

Although it often goes unnoticed, this kind of chronotopic bounce-back between panels is not uncommon in stained glass (see chapter 3e). This interpretation also explains why the successive {chronotopic} frames within panel 19 move from left to right even though the broader movement between the panels is from lower-right to upper-left; by arranging it this way the artist is [131] able to have Roland blowing his oliphant in the right direction to address the group in panel 18.

Panel 20: Baldwin tending to the dying Roland (Second Spanish campaign, scene 7 of 8)

The panel includes a veritable arboretum of stylised trees, suggesting a wooded grove, (Pseudo-Turpin Chapter XXV, gives the location as the entrance to the pass of Cize). To the right a knight (chainmail hauberk, nimbus, no helmet) is shown laying down, resting against his large kite-shaped shield which in turn seems to be propped against the curving side of the frame. He holds his naked sword loosely in his right hand and with his left gestures towards the object being held out by his companion. This other figure, approaching from the left is similarly dressed and also nimbed. His shield is slung over his back, with the cord around his shoulders. With his left hand he holds a gold-coloured object which may well be a helmet, while his right hand rests on the fallen man’s knee (just above the glazing bar, touching the sword’s cross-piece, is a small patch of green glass that has been mistaken by some authors for Baldwin’s hand, leading them to conclude he is reaching for Durandel - careful examination reveals that this piece of glass is actually Roland’s knee.).

After the battle of Roncesvalles and the sounding of the oliphant, the dying Roland is joined by his friend Baldwin, of whom he begs a drink of water. According to Pseudo-Turpin, Baldwin is unable to find any and instead gallops off to tell Charlemagne what has happened. Mâle correctly read panel 20 as showing Baldwin taking his helmet off to go and look for water (a plan seemingly acknowledged by the gesture of Roland’s left hand). 57 This perfectly straightforward explanation has been much problematized by various authors who have all assumed that Baldwin is offering a helmet full of water to Roland - an action which would clearly not fit with the Pseudo-Turpin account.58

Baldwin’s nimbus remains unexplained. Maines’ suggested that ‘...perhaps the simultaneous representation of two nimbate Rolands in panel 19 led to confusion on the part of lay artisans in the succeeding panels’.59 A mistake or misunderstanding on the part of the artist is of course always available as the ‘explanation of last resort’ for the iconographer. However even if this were valid for panel 20, it would scarcely explain the curious distribution of nimbi in the next panel. Baldwin’s temporary promotion to sainthood in these two panels (and Charlemagne’s demotion in panel 21) must for now remain a mystery. [132]

After the drama of panel 19, the pathos of panel 20 has sadly been diminished by the comical replacement heads that an unknown and particularly callous restorer has inflicted on both Roland and Baldwin - perhaps the worst insults that fortune has inflicted on this otherwise well-preserved window (unfortunately the very necessary secondary glazing that has been fitted to the outside of all the windows in the choir, by obscuring the corrosion on the outer surface of the glass, has made it more difficult to assess which other pieces are replacements).

Panel 21: Baldwin telling Charlemagne about Roland’s death (Second Spanish campaign, scene 8 of 8)

Two riders enter the frame from the right - the foremost is the King, dressed in robes (no armour) and wearing a crown, but this time with no nimbus. In the centre of the panel is a knight in a chainmail hauberk, without shield or helmet but nimbed. He places his right hand on the neck of the king’s horse and raises his left hand in a gesture that seems to denote exposition (c.f. the gestures of the right hand bishop in panel 02, or of the king in panel 18). The king responds with an open palm gesture which, as Maines points out, matches that of the left-hand bishop in panel 02 (although it also matches the gestures of Ganelon in panel 18 and St James in panel 09, which mitigates against a simple interpretation of this probably polyvalent gesture). 60 The usual cloud-frill stretches once again across the top of the frame.

This is clearly the scene described in Pseudo-Turpin, where Baldwin takes Charlemagne the sad news of Roland’s death (note that we don’t actually see Roland’s death in this visual discourse, which may help to account for the presence of God’s blessing hand in panel 19). Quite why Baldwin has a nimbus in this scene and Charlemagne does not is far from clear - both heads appear to be original (the head and flanks of Charlemagne’s horse are later repairs - otherwise the glass is in good condition).

Within a semiotic system in which facial expressions rarely play a consistent or significant role, the artist here has succeeded in giving Baldwin a splendidly mournful demeanour (although since this detail is about sixteen metres above the ground it is scarcely legible to the naked eye).

Panels 22-24: The Mass of St Giles (Monoscenic segment)

Panel 22 shows an interior scene, set within an arcaded space suggesting a church. The artist has included a stylised tiled floor, which is relatively uncommon in interior scenes. Far-right is an altar, seen in side view, on which rests an uncovered chalice (a combination of iconic signs which together function as a very specific conventional sign denoting that mass [133] is currently being celebrated - otherwise the chalice would be shown covered by a cloth). Standing at the altar is a cleric - nimbed, bare headed and tonsured, wearing a cope. Above the altar is the diminutive figure of an angel (also nimbed, wings not shown), who gestures with his right hand (index finger raised) and with his left hand lowers a scroll into the waiting hands of the cleric, who gazes up at him. Behind the cleric stands a deacon, tonsured but not nimbed, who holds a closed book, presumably a missal. Behind him is the figure of a king seated on a throne shown in three-quarter view. This figure is robed, with legs crossed (a very common conventional sign denoting a ruler), crowned and nimbed and is resting his chin in his left hand, a gesture which commonly denotes worry or concern.

To the left and right of the upper half of this main panel are two cut-down quadrant panels (23 and 24), each of which contains an angel emerging from a cloud-frill and facing into panel 22. These angels, shown carrying a shallow dish in one hand and swinging a thurible in the other, are clearly hypotactic elements, ‘commenting on’ the events in panel 22 but not active participants in the story. Although thurifying angels are often associated with the death scene of a saint, this is not always the case. In the St Thomas in India window at Chartres for example they are shown flanking the final image of pilgrims at the Saint’s tomb (panels 26-28), his martyrdom having happened four frames earlier in panel 22. In other cases, as here, they were employed to valorise the depiction of some other miracle, a practice more common with confessor saints (like Charlemagne) rather than martyrs.

The {pictorial} framing of these three panels reveal an interesting characteristic common to most stained glass artists. As we have seen in other panels, the ‘internal’ frames of the panel are often used as part of the sign-system; either to show relative planes of depth, to stress movement in or out of the frame, or to indicate spatial differences.  Here for example the angel with the scroll emerges at the top-right of the panel and completely overlays the red, blue and white concentric rings (within which all other elements of the image are enclosed), to denote that he belongs to, or is emerging from, a different physical space from the rest of the scene. However, in these panels, as is normally the case, the ‘external’ framing system of the armature is inviolable - no part of the image in one panel, such as these thurifying angels, ever spills into the adjoining panel. One can contrast this with the work at Bourges and Poitiers of the ‘Good Samaritan Master’, who routinely uses the space between panels as part of his narrative canvas (see chapter 9) . Compare for example the censing angels constrained by the armature at the top of this window with the angel who floats behind the armature at the top of the Bourges ‘Life of St Stephen’ window (see Fig. 4.30).61

[134] The story of the Mass of St Giles, in which the saint was informed about the king’s ‘unmentionable’ sin via an angelically-supplied scroll, was a well known one in medieval Europe. Indeed the considerable popularity of  St Giles (or Aegidius) as an intercessory saint of greater than usual efficacy was based largely on this specific event in his vita. As it says in the Golden Legend, ‘...anyone who had committed a sin and prayed to St Giles to obtain pardon should have no doubt that by the saint’s merits the sin was forgiven’.62

Emile Mâle rightly points out that Charlemagne does not feature in the legend of St Giles. 63 Since the Golden Legend dates the activity of Giles to around 700AD, the Carolus Rex mentioned in that story is not Charlemagne but his grandfather, the founder of the Carolingian Dynasty,  Charles Martel. Nevertheless, in medieval Europe, the miraculous Mass of St Giles became firmly associated with Martel’s more popular grandson - which inevitably prompted much prurient speculation about what exactly poor Charlemagne’s peccatum innominandum might have involved. Later narratives in the German vernacular opted in favour of necrophilia (a rather unpleasant incident involving his dead queen and an evil charm placed under her corpse’s tongue) - but the French preferred incest. As well as fitting in with the classical literary topos of royal incest, this also meant that Roland, the other great hero of the story, could now be Charlemagne’s son rather than just his nephew. 64

In terms of its theological/didactic functions, the story of the Mass of St Giles has two key messages; firstly that any sin can be forgiven if the sinner is truly penitent, and secondly that forgiveness is only possible when mediated through the agency of a priest. The Fourth Lateran Council of 1215 had tightened up the rules on confession and the confessio Deo soli (in which a contrite sinner may silently ask God’s forgiveness) was no longer enough.65 Instead the sin had to be made fully known to the priest, who was instructed to ‘carefully inquire about the circumstances of both the sinner and the sin’ (Fourth Lateran Council, Constitutio 21). The artist in panel 22 has done his best to conform to these requirements; the king is shown in as penitent a pose as the conventions of the medium will allow - and crucially the full detail of his sin is being communicated to the officiating priest by the angel, who is shown not only handing him a written account of proceedings but also addressing him directly (note that Giles is looking directly at the angelic messenger, not the scroll). Thus panel 22 makes it clear that for all his contrition, generosity to the Church and heroic deeds in the service of the Faith, even Charlemagne could receive forgiveness only through the authority of a priest. [135]

4.e) conclusions

Many of the saints who feature in the windows at Chartres Cathedral were added to the local liturgy in the thirteenth century, subsequent to the installation of their hagiographic windows. Some have seen this as evidence that the narrative windows played an active role in constructing the local cults of those saints. For some reason, despite his spectacular window sited in such a prestigious location, Charlemagne never made it into the Chartrain Liturgy. Indeed the cult of Charlemagne in France only really took off in the second half of the fourteenth century, following the active promotion of the saint by Charles V.66   The obvious conclusion then is that Charlemagne’s adventures are included here not primarily for his own sake but as a pendant for his neighbour in bay 05 (Fig. 4.31).  Charlemagne is in the ambulatory at Chartres as an accessory to St James the Greater. As Elizabeth Pastan explained in one of the most recent and considered studies of this window:

It was Charlemagne who liberated the tomb of James in Galicia from the infidel, thus rescuing him from neglect and instigating pilgrimages to the site. Reciprocally, it was St James [according to Pseudo-Turpin], who decided that Charlemagne’s good deeds outweighed his sins. [...] As this resonant pairing [of windows] indicates, Charlemagne and James are guarantors for each other.67

Whilst it may be overplaying matters slightly to suggest that the reputation of St James the Greater needed much of a boost, it is certainly true that a reminder of the circumstances under which the Chemin St Jacques was first liberated would have helped in the valorisation of the Santiago pilgrimage. Perhaps the prominence given here at Chartres not just to James but also to Charlemagne helped encourage the pilgrims coming from Paris on that journey to favour the route via Chartres over the alternative via Orleans. In bay 05 those pilgrims could read the life and death of the saint whose remains they were travelling to see, while alongside it in bay 07 they could see not only how the path that lay before them was cleared but also get a preview of some of the miraculous places that they would visit en route; Pamplona whose walls were flattened by God at Charlemagne’s behest, and the field of Sahagún where the lances of the martyrs flowered (and where as Pseudo-Turpin assured them, they would still be able to see the ash trees that sprouted from their stumps). In this regard the Charlemagne window at Chartres is a kind of pilgrimage map - not much of an itinerarium perhaps - but certainly a way to narrativise the road ahead.

So much for Charlemagne – but what does this case-study say about visual narratology? A key lesson, particularly when one reviews the earlier literature on this window, is the importance [136] of striving for an ‘innocent eye’ – to see only what is in the image, rather than one expects, or wants to see there. That goal is of course an unattainable one (see the discussion of reader-response models in section 1.c above) but nevertheless there is some value in self-consciously treating one’s expectations with deliberate scepticism, questioning identifications more rigorously the more conveniently they fit one’s pre-existing theories - and never being afraid to admit that a favoured interpretation is only ‘likely’, rather than certain. The widespread use of visual topoi in medieval art is a trap liable to ensnare any historian who goes looking for examples of a particular episode without considering all the other narratives that also feature that topos (this is the trap into which Lejeune and Stiennon blundered so spectacularly). The other crucial thing is to treat every image cycle ‘holistically’. Even if one is only interested in a specific detail of an individual scene it is wise to consider all of the associated scenes (as well as issues of framing, arrangement and so on). All too often, the correct identification of an individual scene is dependent on a correct understanding of the overall structure.


Previous Section Contents Page Next Section


1 For Panofsky’s ‘hierarchy of readings’ see his essay ‘Iconography and Iconology: An Introduction to the Study of Renaissance Art’, reprinted in E. Panofsky, Meaning in the Visual Arts: Papers in and on Art History, Garden City, N.Y. 1955, pp.26-54.

2 Details as per L. Grodecki, Corpus vitrearum Medii Aevi. Recensement des vitraux anciens de la France, Volume 2 - Les vitraux du Centre et des Pays de la Loire, Paris 1981.

3 Throughout this thesis I have adopted the excellent and widely applicable system for numbering window bays devised by the CVMA in France. In this system, the axial window is bay 0. Other bays are then numbered in order with odd numbers to the left (north side) and even numbers to the right (south). Upper windows, when present, are numbered in the same way but starting from 100 for the triforium (if glazed) and 200 for clerestory. Thus bay 07 is the fourth window around to the north of the central axis on the ambulatory level.

4 In April 637AD, Jerusalem fell to the Rashidun Caliphate and remained in Muslim hands until it was recovered by Christian forces during the First Crusade in 1099 (and then lost again, this time to Saladin, in 1187).

5 Or to give it its full name, the Descriptio qualiter Karolus magnus clavum et coronam Domini a Constantinopoli Aquisgrandi detulerit qualiter que Karolus Calvus haec ad S. Dionysiam retulerit. For background and bibliography, see R. Folz, Le Souvenir et la légende de Charlemagne dans l’Empire germanique médiéval, Dijon 1950, especially p.179.

6 E. Mâle, Religious Art in France: The Thirteenth Century, Princeton 1984 [1898], p.343.

7 H. M. Smyser, Pseudo-Turpin: ed with an Annotated Synopsis, Cambridge, Mass. 1937, p.1.

8 The two main sources for the Latin text and its variants are C. Meredith-Jones, Historia Karoli Magni et Rotholandi, ou Chronique du Pseudo-Turpin, Paris 1936, and  H. M. Smyser, Ibid. The only English translation is T. Rodd, History of Charles the Great and Orlando, Ascribed to Archbishop Turpin, London 1812, however it is not clear which manuscript(s) were the source of the Latin text he used. Although his translation is highly readable and seems reliable, it omits certain sections, which makes the chapter numbering confusing. In the sections that follow, I will use Rodd’s translation and where they differ give separate chapter numbers both for that and for the Latin text in Smyser’s edition.

9 For a summary of Bédier’s arguments and other supporting material, see the introduction to H. M. Smyser (ibid). See also J. Williams and A. Stones (eds), The Codex Calixtinus and the Shrine of St. James, Tübingen 1992.

10 For the Life of St Giles and  its variants, see F. Brittain, Saint Giles, Cambridge 1928.

11 A very similar argument has been made in relation to the legends of Ywain; see J. A. Rushing, Jr., Images of adventure: Ywain in the visual arts, Philidelphia 1995. For the broader issues of literacy and orality during the period, see F. Bauml, 'Varieties and Consequences of Medieval Literact and Illiteracy', in Speculum, 55(2), 1980, pp. 237-65.

12 E. Mâle, Religious Art in France: The Thirteenth Century, Princeton 1984 [1898], p.344.

13 R. Lambrech, 'Charlemagne and his influence on the late medieval French kings.' Journal of Medieval History 14:4, 1988, p.283

14 Several such image cycles are detailed in R. Lejeune and J. Stiennon, The Legend of Roland in the Middle Ages, London 1971 [1966], although some of their identifications must be treated with caution.

15 S. G. Nichols, Romanesque Signs: Early Medieval Narrative and Iconography, Yale 1983, pp.101-05.

16 D. Robertson, 'Visual Poetics: The Charlemagne Window at Chartres.' Olifant 6:2, 1978, p.107.

17 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image.' Speculum 52:4, 1977, p.811.

18 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image.' Speculum 52:4, 1977, p.809.

19 For a discussion of the cloud-frill in (mostly) later art, see S. Ringbom, 'Some pictorial conventions for the recounting of thoughts and experiences in late medieval art' in Medieval iconography and narrative: a symposium, ed. F. G. Andersen, Odense 1980.

20 W. Kemp, The Narratives of Gothic Stained Glass, Cambridge 1997 [1987], p.86.

21 E. Pastan, 'Charlemagne as Saint? Relics and the choice of window subjects at Chartres Cathedral', in The Legend of Charlemagne in the MIddle Ages, ed. M. Gabriele and J. Stuckey, New York 2008, p.120.

22 In 1131, Count Thibaut IV imposed an overseer on the furriers (who paid an annual contribution to the Leprosarium) and similar arrangements were subsequently made for most other trade groups - but the first actual guild statutes at Chartres (those of the wool workers) were not issued until 1268 (see J. Welch Williams, Bread, Wine and Money: The Windows of the Trades at Chartres Cathedral, Chicago 1993, p.11). Thus whatever relationship the tradesmen depicted in the various signature panels had with those windows, it was not a simple case of guild sponsorship.

23 For Philip’s appropriation of the Charlemagne legend as tool of Capetian political identity (which had actively begun around 1200) see R. Folz, Le Souvenir et la légende de Charlemagne dans l'Empire germanique médiéval, Dijon 1950, p.277ff.

24 Quoted in E. Mâle, Religious Art in France: The Thirteenth Century, Princeton 1984 [1898], p.344.

25 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, pp. 806. Emile Mâle, looking as ever for a close match of image to text,  saw the pair of Bishops as being two of the four emissaries mentioned in the Descriptio (E. Mâle, Religious Art in France: The Thirteenth Century, Princeton 1984 [1898], .p344).

26 Lejeune and Stiennon claimed that the famous Reims depiction of Abraham and Melchizadek represented Archibishop Turpin giving the communion to Roland before his battle with Marsile (whom they supposed was represented by the figure in the adjacent niche wearing the lorica squamata). See R. Lejeune and J. Stiennon, The Legend of Roland in the Middle Ages, London 1971 [1966], pp.203-05. This identification of the Reims figures (which is particularly fanciful, even by the standards of that particular publication) has not found wider acceptance.

27 For more details see the excellent catalogue produced by the Walters Gallery in Baltimore; W. Noel and D. Weiss, Eds., The Book of Kings: Art, War, and the Morgan Library's Medieval Picture Bible, London 2002.

28 R. J. Adams, 'The Chartres Clerestory Apostle Windows: An Iconographic Aberration', in Gesta, 26(2), 1987, p.143.

29 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, pp. 807.

30 C. Lautier, 'The Sacred Topography of Chartre Cathedral: The Reliquary Chasse of the Virgin in the Liturgical Choir and Stained Glass Decoration' in The Four Modes of Seeing: Approaches to Medieval Imagery in Honour of Madeline Harrison Caviness, ed. E. Lane, E. Paston and E. M. Shortell, Farnham 2009, pp. 184.

31 C. Lautier, ibid, p.189

32 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, p.807

33 C. Maines, Ibid. p.807, n.21.

34 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, p.807.

35 Pseudo-Turpin, Chapter II (English version based on Rodd’s 1812 translation).

36 Pseudo-Turpin, Chapter III (English version based on Chapter II of Rodd’s 1812 translation)

37 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, p.809, n.30.

38 R. Lejeune and J. Stiennon, La légende de Roland dans l’art du moyen âge Brussels 1966, p.194.

39 See E. A. R. Brown and M. W. Cothren, 'The Twelfth-Century Crusading Window in the Abbey of St Denis: Praeteritorum Enim Recordatio Futurorum est Exhibitio', in Journal of the Warburg and Courtauld Institutes, 49, 1986, pp. 1-40, especially Plate 5.

40 Pseudo-Turpin, Chapter IV (English version based on Chapter III of Rodd’s 1812 translation)

41 E. Mâle, Religious Art in France: The Thirteenth Century, Princeton 1984 [1898] p.347.

42 R. Lejeune and J. Stiennon, The Legend of Roland in the Middle Ages, London 1971 [1966], p.194.

43 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, p.811

44 Pseudo-Turpin, Chapter VI (English version based on Chapter V of Rodd’s 1812 translation). My italics.

45 Carolyn Carty is one of the few to have opted for a generic image of church-building- C. M. Carty, 'The Role of Medieval Dream Images in Authenticating Ecclesiastical Construction.' Zeitschrift für Kunstgeschichte 62:1, 1999, p.70.

46 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, p.811.

47 Pseudo-Turpin, Chapter VIII (English version based on Rodd’s 1812 translation).

48 R. Lejeune and J. Stiennon, The Legend of Roland in the Middle Ages, London 1971 [1966], p.195. They base their identification on the red ‘tunic’ (surcoat) worn by this rider which also features in the fight with Ferracutus in panel 16. This is a nonsense which overlooks the fact that Charlemagne also wears an identical garment in panels 05 and 11. The costumes (and facial characteristics) of characters in stained glass windows are hardly ever a constant and reliable signifier. With very few exceptions, it is only context and/or specific attributes (and occasional semantic enclaves like tituli and heraldry) that can reliably identify a particular character in a stained glass window.

49 Although Maines (ibid, p.814) refers to the subsequent events as a ‘third expedition of the Spanish crusade cycle’, in Pseudo-Turpin,the campaigning is effectively continuous from Sahagún onwards and so it makes more sense to talk about a single ‘second phase’ lasting from Charlemagne’s return to Spain (prior to the Battle of Sahagún) until the eventual retreat to France and the death of Roland.

50 The view that the panels had been misplaced was first expressed in M. Bulteau, Description de la cathédrale de Chartres, Chartres 1850, p.238.

51 R. Lejeune and J. Stiennon, The Legend of Roland in the Middle Ages, London 1971 [1966], p.72.

52 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, p.814.

53 H. M. Smyser, The Pseudo-Turpin, edited From Bibliothèque Nationale, Fonds Latin, MS.17656, with an Annotated Synopsis, Cambridge, Mass. 1937, p.13.

54 D. Robertson, 'Visual Poetics: The Charlemagne Window at Chartres.' Olifant 6:2, 1978, p.107.

55 Pseudo-Turpin, Chapter XXVI (English version based on Chapter XXIII of Rodd’s 1812 translation)

56 R. Lejeune and J. Stiennon, The Legend of Roland in the Middle Ages, London 1971 [1966], p.196.

57 E. Mâle, Religious Art in France: The Thirteenth Century, Princeton 1984 [1898], p.348.

58 See C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, p.816-17 for a discussion of Lejeune and Stiennon’s rather dramatic over-interpretation of this apparent disparity, which prompted them to postulate the existence of a now-lost textual variant of Pseudo-Turpin.

59 Ibid., n.72.

60 C. Maines, 'The Charlemagne window at Chartres Cathedral: New Considerations on Text and Image', in Speculum, 52(4), 1977, p.816

61 I will return to the work of the Bourges ‘Good Samaritan Master’ in my final chapter.

62 W. G. Ryan, Jacobus de Voragine's The Golden Legend: Readings on the Saints, Princeton 1993, Vol.II, p.148.

63 E. Mâle, Religious Art in France: The Thirteenth Century, Princeton 1984 [1898], p.491, n.64.

64 For the medieval speculation on this matter, see S. Hafner, 'Charlemagne's Unspeakable Sin', in Modern Language Studies, 32(2), 2002, pp. 1-14.

65 S.Hafner, ibid, p.2.

66 E. Pastan, 'Charlemagne as Saint? Relics and the choice of window subjects at Chartres Cathedral' in The Legend of Charlemagne in the Middle Ages, ed. M. Gabriele and J. Stuckey, New York 2008, p.99.

67 E.Pastan, Ibid.


Previous Section Contents Page Next Section