Previous Section Contents Page Next Section


7) The narrativisation of the object

[173] So far in this thesis the relationship between narrative images and the objects that bear them (be they buildings, books or whatever) has been a rather distant one. Of course there is always a connection of some sort between the context and the content of a narrative. A cathedral bears images in glass or stone which depict universal Christian narratives as well as stories of more local interest (foundation myths, patron saints, etc), while a reliquary will more often show the life and death of the saint whose relics it contains - but normally the connection is rather a generic one.  This chapter deals with a very special form of relationship between an object and its images; specifically it concerns those objects that depict themselves and their own stories as a mise-en-abyme.

In the curious language of heraldry, the abyme or fess-point is the exact centre of an escutcheon. To place something en abyme literally means to depict it in the middle of the shield. The term is usually reserved however for the practice of placing at that central spot a smaller shield with its own bearings, which modified the meaning of the bearings on the main shield. A commonly cited example is the so-called ‘escutcheon of Pretence’ in which a man who marries an heiress may temporarily add a miniature version of her family arms to the centre of his own escutcheon, while a nation’s arms may take en abyme the bearings of territories to which they claim rights. The expression mise en abyme might well have remained an obscure terminus technicus had not André Gide, himself a keen student of heraldry, used it to describe a form of self-reflexive embedding he had observed in various art-forms. As he explained in his journal:

... in paintings by Memling or Quentin Metzys, a small dark convex mirror reflects, in its turn, the interior of the room in which the action of the painting takes place. ... what would explain better what I’d wanted to do in my Cahiers, in Narcise and in La Tentative, would be a comparison with the device from heraldry that involves putting a second representation of the original shield ‘en abyme’ within it.1

Gide’s use of this term ‘mise en abyme’ was picked up by Claude-Edmonde Magny in his history of the French novel and soon became part of the analytical lexicon of literary historians.2

It subsequently entered the discourses of art-history, particularly amongst scholars working on the early-modern period, who saw in its self-reflexivity an index of artistic ‘self-awareness’.3 [174] The phenomenon has however been largely ignored by students of medieval art, despite its frequent occurrence in presentation scenes, author portraits and many other contexts.4 Rather than engaging in a general discourse about mise en abyme as a strategy of medieval artistic production, this chapter will concentrate on how the device functions specifically in the context of visual narratives.

Returning to the origins of the term, what mattered for Gide (as is plain from the examples he chose to illustrate his point) was not the mere presence of an embedded image or narrative within a larger whole but the fact that the thing thus contained resembled that which contained it - and moreover that this resemblance in some way informed the viewer or reader about the form or meaning of the whole (the effect for which he was striving in his own writings). Thus in Shakespeare’s Hamlet, the Prince has the players perform The Murder of Gonzago (the plot of which resembles his own predicament) that it might ‘...prick the conscience of the King’ and thereby provoke the latter into revealing his guilt.  In a similar fashion, in Las Meniñas, Velàsquez includes the King and Queen as reflections in the mirror at the back of the painter’s studio in the Alcazar - a reflection of what is outside the picture, which helps to contextualise and explain what is within it (or rather, to judge by the historiography, helps to problematize it).5 It is precisely this meta-narrative function of mise en abyme that is of interest in relation to medieval narrative art - that is to say the ways in which the use of the device contributes to a viewer’s reception of the narrative context in which it occurs.6

Since the term is often misunderstood it is worth stressing that mise-en-abyme does not necessitate an implicitly endless recursion of ever-smaller images placed en-abyme inside each other (sometimes also known as the Droste effect after a Dutch drinking chocolate manufacturer who used such a device on their packaging). As the preceding explanation of the term’s origins and the ensuing discussion of its uses should make clear, a single level of self-reflexive imagery is enough.


7.a) first issue with iconicity; the nature of  the referent

One issue that must be refined somewhat before turning to some practical examples of mise en abyme in narrative art is the exact nature of the iconic sign - since fundamental to mise en abyme is the principle that the thing contained must ‘resemble’ its container. All of the examples discussed here can be said to be ‘iconic’ (in the Peircean sense) or ‘iconic signs’ as Wallis called them (see chapter 1f) but the nature of the visual resemblance and the issue of what exactly constitutes the object of that resemblance (its ‘referent’) is often ambiguous.  Let us consider for example some images of books that appear within a particular illuminated manuscript, the Parma Ildefonsus, (Parma, Palatine Library ms.1650), which was made at Cluny around 1100.7 As well as the Ildefonsus text itself (a defence of the virginity of the Virgin) the manuscript includes a brief biography of the author, a eulogy composed by one Bishop Julian and a lengthy scribe’s colophon copied from an earlier, unillustrated Ildefonsus manuscript written by the Spanish monk Gòmez for Bishop Gotiscalc of Le Puy in 951.8 Of the thirty-five miniatures in the Parma codex, codices are depicted in no less than fifteen. All of these ‘books portrayed within a book’ can legitimately be considered as mises-en-abyme to some extent, since they are contained within an object recognisably of the same type as themselves. The functions and the semiotic modes of these images can however be separated into distinct groups depending on what the image of a book represents:

  1. The book as symbol of the divine Logos, as in f.5r (Fig. 7.01B) and f.44v (Fig. 7.02A). When Christ is seen as Judge in Apocalyptic imagery, the book He holds may be interpreted as The Book of Revelation seen by John in his vision. However, in all other cases, as here, it represents Christ’s divinity as ‘The Word’ - the Logos mentioned in the opening line of John’s Gospel. Often the book is shown open with the words Ego sum lux mundi (John 8:12), though whether this should be seen as a prosopopoeic utterance of the book or as a kind of speech bubble for He that holds it is a moot point.
  2. The book as metonym for the words of the person depicted in association with it - Ildefonsus himself in f.15v (Fig. 7.02B), Bishop Julian in f.1v (Fig. 7.02C), the Evangelists Mark and Luke in f.44v (Fig. 7.02A), plus various of the Old Testament prophets whose words Ildefonsus cites in support of his arguments (Isaiah in f.26r, David in f.27v, Ezekiel in f.28r, etc.)
    Take for example f.36r (Fig. 7.01C). Above the image is a rubric; Dicit Malachias. The image itself shows Malachi holding up a codex and pointing to it as he relates his [176] prophecy to two figures, one of whom points off to his left as if he plans to obey the instructions in the text written below the picture; Veniet ad Templum ... (Malachi 3:1). Thus we have the rubricated words ‘Malachi says’, an image of Malachi speaking, plus the actual words of his prophecy. Clearly in this case the book held by the prophet does not represent books in general, nor the book in which this image appears, but specifically (and metonymically) his words as they appear in the Old Testament. However, by showing us Malachi brandishing his book as if it were a weapon against ignorance or heresy, the designer of the image is also providing a model of how books (including this one) should be used - not as a physical bludgeon but as a source of true arguments against sceptics, heretics and unbelievers.
  3. An actual book (i.e. an individual physical codex) such as the Parma manuscript, or the earlier copy made in 951 by the scribe Gòmez. If the previous mode is associated with author portraits, this mode is characteristic of scribal portraits and presentation scenes. A special case of this mode (and the truest form of mise en abyme) is when the book depicted in such scenes is (or is meant to be interpreted as)the self-samephysical codex in which its image appears. Such images sometimes appear as a ‘visual colophon’, such as the famous scribal self-portrait of ‘Hugo pictor’ at the end of Oxford ms. Bodley 717 (Fig. 7.03D) or the example that closes the Toledo Bible moralisée discussed below. Interestingly, this ‘specific codex image’ is the one mode of signification that does not appear in the Parma manuscript. However one could argue that by copying the lengthy colophon from the 951 copy (Paris Bib. Nat. ms Lat. 2855) and adding two new miniatures, one showing Gòmez making his copy (f.102r - Fig. 7.02D) and another of him presenting it to Bishop Gotiscalc,  the maker of the Parma manuscript is using these mises en abyme to show the story of this particular text (its autobiography, as it were), and thereby to assert that his is a true and faithful copy of that earlier, and hence authoritative, manuscript. 

What I have, for convenience, presented as a classification into three distinct groups, should more properly be thought of as a continuum, ranging from the purely generic (a general type of object) to the very particular (a specific physical instance of that type).  If all these distinctions between different images of codices seem overly pedantic, it is worth remembering that they fit rather neatly into medieval scholastic distinctions between quidditas and haeccaetas, between ‘what-ness’ and ‘this-ness’ - and hence into the contemporary disputes between neo-Platonism and neo-Aristotelianism. An image of a codex, as a sign, may signify the Platonic ‘idea’ of ‘book-ness’ (quid est res? ), or it may signify a specific instance of ‘book’.


7.b) second issue with iconicity; the nature of the likeness

As well as asking what exactly the depiction depicts, one might also consider the manner of that depiction - the way in which the image resembles its subject. Chapter 1.e described the two main modes of visual signification - the iconic sign which resembles its referent mimetically and the conventional sign, whose relation to the referent is instead based on socially-determined conventions. In these terms the various depictions of codices in the Parma Ildefonsus are problematic. One would not necessarily read the object held by Malachi in Plate Fig. 7.01C as ‘a book’ without at least some prior knowledge of the conventions of Romanesque painting, still less when these painted marks on a page are removed from their context, as shown in Fig. 7.03A. Herein lies the flaw in the ‘iconic sign’ model - which elides so much of the complexity underlying mimesis. At the heart of the problem is the fact that resemblance itself is to some degree socially determined, with local conventions dictating the relative contributions of shape, colour, scale and perspective towards the recognisability of the supposedly iconic sign.  Moreover, even if we still had the original manuscript in its original binding, there is no way we could confidently state from its appearance alone that the codex depicted in f.102r  (Fig. 7.02D) is actually the manuscript made by Gomez or its later copy or else some other book entirely. The object depicted is simply a generic book - since that is the medieval convention for depicting books. It is only its context that identifies it as ‘this book’, i.e. the physical codex in which the image appears.

At the opposite end of the iconicity spectrum (and well outside my chronological scope) is the famous presentation miniature in the Chroniques de Hainaut attributed to Rogier Van der Weyden and dating to around 1453 (Fig. 7.04).9 Here, the volume being presented to Philip the Good is depicted with such precision and apparent naturalism that one would feel quite confident about being able to pick it out from amongst its peers in the Ducal Library. Yet we can be equally confident that the tanned and tooled leather binding depicted so carefully in the miniature (Fig. 7.04B) looks nothing like the actual binding that originally contained it, since this same book was described in a 1504 inventory as ‘Ung livre ... couvert de satin figuré noir...’.10

Generally then, it is not a detailed physical resemblance that identifies the depicted object with the object containing that depiction. It is rather the combination of a generic resemblance (the [178] type of object) and the context in which it is depicted (presentation/donation scene, author portrait, etc).

Notwithstanding these caveats, there are plenty of cases where the artist has included information to help individuate the depicted object. In the signature panel of the ‘Relics of St Stephen’ window at Chartres (Fig. 7.05A), dating from around 1210, the miniature window being offered by the Cordonniers contains a carefully drawn geometric pattern which perfectly matches the armature pattern of the window that contains it - one which (of the many patterns deployed at Chartres) is unique to this window. Even if the contents of the miniaturised panels had to be elided, the window itself is unmistakeable. A similar example can be found at the bottom left of the St Maurice window (bay 201, panel A1) of the Cathedral  of St Gatien in Tours, where the donor, Bishop Geoffroy Freslon of Le Mans, offers his window (see Fig. 7.06).11 The latest fashions in window design are perfectly reflected in this example, which dates from around 1260 - both the real window and its self-image in the bottom left hand panel consist of three slender lancets surmounted by three trefoil oculi with glazed spandrels. The rayonnant bar tracery shown in the miniature window presented by the Bishop exactly matches that of the real window that contains it, even down to the pierced cusping  at the top of the lancets - hard to spot from the ground but just visible in Fig. 7.06. Perhaps the best known of these ‘self-reflexive presentation scenes’ in stained glass is the image of Abbot Suger offering a miniaturised Jesse window in the bottom right of the real Jesse window at St Denis - see Fig. 7.05D. Sadly this one really is too good to be true. Although most of the central panels are original, the entire border of this window, including the detail of Suger, was invented by the nineteenth century restorers.

One may contrast the carefully self-referential windows from the cathedrals of Chartres and Tours with the many examples of donation scenes in the surviving thirteenth century windows at the Cathedral of St Julien in Le Mans.12 In most cases, the proffered panel is depicted without internal detail, as in bay 109, panel C1 (Fig 7.05B). Even more interesting however are the two donor panels in the ‘de Cormes Family Window’ (bay 207) made around 1270 (Fig 7.05C). The Abbot of Notre Dame de la Couture, who appears in panel B1, and the unidentified knight in panel E1 each hold up lancets with clearly marked medallion patterns; yet in both cases these patterns of quatrefoils and lozenges are nothing like the plain lancets of the window that [179] contains them. Instead they are of a type more characteristic of narrative windows made half a century earlier, such as those of Chartres and Bourges, or which once filled the radiating chapels at Le Mans. I suspect that by using such retardataire patterns the artist was not suggesting that either of these donors were responsible for the earlier windows but was simply adopting an iconic form that was more familiar to his audiences and hence more easily recognised as windows when viewed from a distance.

Generally speaking then, it is enough for the record of the donation of some object en abyme to indicate the generic type of object thus gifted. Context and convention are enough to denote that it is this window that was given, rather than some other, earlier one.

7.c) mise en abyme as a strategy for narrativising the object

Turning now from the forms to the functions of mise en abyme, the most frequent use to which it was put in medieval art is the narrating, or more precisely the narrativising, of a key event in the life of the object (this distinction between narrating and narrativising is a subtle but important one - the former is the telling of a story but the latter is the creation of that story;  the turning of actions or happenings into narrative proper - see chapter 2.a). Of the events in the biography of an object, the most important for medieval owners, donors and audiences alike were normally its creation and its donation. To permanently mark upon an object an autobiographical scene showing either its origin or the purpose of its existence served to reify its foundation-myths - and thus embed them in the viewer’s repertoire of remembered narratives in a far more potent way than conventional prosopopoeic inscriptions of the [X] me fecit form.13

Self-reflexive scenes depicting the creation of the object are generally restricted to authorial or scribal portraits in manuscripts, perhaps the oldest extant example in a Christian context being the image of Ezra (and/or Cassiodorus) the scribe in the early eighth-century Codex Amiatinus (Florence, Biblioteca Medicea Laurenziana, MS Amiatinus 1, f.5r). 14 Images like this often find a new life in books on manuscript illumination as depictions of the working practices of scribes, despite the fact that these scribes and illuminators are shown rather implausibly writing in bound codices, rather than on loose sheets as would normally have been the case. From the viewpoint of the original user (if not the art historian), instead of devaluing the narrative truth-claim of the image, such ‘modifications’ enhance it by making clearer the relationship between the depicted act of writing and the physical object (a finished book) that contains it. [180] Authorial portraits could also be enhanced to make additional claims for the truth-value of the associated text by suggesting divine inspiration. This was particularly common for the Evangelists and the Fathers of the Church. Thus the frontispiece of the Hartker Antiphonary (St Gall Stiftsbibliothek ms.390-391) shows the nimbed Pope Gregory I, dictating to a seated scribe, while the Holy Ghost, in the form of a dove, whispers in his ear (Fig. 7.03B). Antiphonaries contained the sung portions of the Divine Office, supposedly compiled into Canonical form by Gregory at the end of the sixth-century. Four hundred years later, there were still a number of variant Offices in use across Europe, despite several attempts to standardise the form. By opening a manuscript  of Hartker’s Antiphonary with a narrative image showing the text’s foundation myth - not just saintly but divinely inspired - the maker is suggesting that this particular version is to be trusted more than others. When one considers the extent of textual discrepancies between manuscript copies within any given stemma, the value of such images of ‘authorship in action’ as modality markers becomes clear.

Apart from author/scribe portraits in manuscripts, donation/presentation scenes are perhaps the most widespread forms of mise en abyme, both in their frequency and in the diversity of object on which they appear. The giving of gifts was a major component of medieval society, in secular and religious spheres alike. The recipient of a gift might be a secular lord or vassal, an ecclesiastical dignitary or, most commonly, the Church. In modern cash-based societies, gifts and bequests are seen as selfless acts and legally held to be ‘alienated’ (that is to say, the giver surrenders all rights and interests in the item). By contrast, medieval secular gifting was a complex two-way affair, closer to the practices described by social anthropologists in ‘primitive’ cultures, whereby acceptance of a gift implied the acceptance of some reciprocal obligation.15   Moreover, such gifts (which typically took the form of land or privileges) could always be revoked if the giver felt they weren’t receiving sufficient service in return. Naturally, ecclesiastical communities were not subject to the same feudal laws on the giving and receiving of gifts, although gifts between religious houses were important as ways of establishing and maintaining hierarchies and affiliations. Gifts between churches may also have  played a major part in the promotion of saintly cults (see for example chapter 8.1 regarding the cult of St Valerie). Finally there was the all important offering of gifts from the secular community to the church, either of lands and privileges (a practice so widespread across Europe that many rulers eventually found it necessary to introduce ‘mortmain’ laws to prevent the gradual and permanent alienation of their kingdoms) or of new buildings and treasures to occupy them. Gifts were the warp around which medieval societies wove their threads of kinship. Within such a gift-oriented society, to have an act of giving permanently inscribed upon the gift itself had [181] obvious benefits, not only in the short term but also as a reminder to future generations of the benefactor’s piety and generosity. Recording a gift on the object thus gifted was one strategy for planning one’s route out of Purgatory by helping to attract the intercessory prayers of one’s successots. In order to explore how artists employed mise en abyme to create a kind of visual autobiography of these gifted objects, I will examine a few examples in more detail.

After his death in 1918, the French connoisseur Victor Martin le Roy posthumously ‘gifted’ to the French nation a large part of his extensive collection of medieval art (the reciprocal obligation in this case having presumably been the relief of death duties). One of the items from that bequest was the top part of a portable altar made in northern Germany around the end of the eleventh century. This object, which I shall refer to as the ‘Rudolfus Altar’, is slightly taller than an A4 sheet of paper and like most early portable altars, was designed to be viewed in ‘portrait’ rather than ‘landscape’ format (Fig. 7.07B). It consists of a rectangular slab of serpentine marble mounted on a wooden panel and surrounded by a border of silver gilt. Just like their full sized fixed cousins, portable altars in the Roman Rite were required to have a chest of some kind below, housing relics, and a stone tablet above. Sadly a high proportion of surviving examples have, like this one, lost their bases, and with them much crucial evidence for how they were originally used (in some cases it seems these portable altar slabs were originally made as independent luxury objects with no provision for a chest and with decoration on the reverse, however this particular example appears to have been the genuine article). The gilded border is inscribed all around - Christ enthroned in the top margin and the Virgin enthroned at the bottom, each flanked by six seated but highly animated apostles, with images of saints on the two long sides. Near the very bottom of this border is the prostrate figure of a tonsured monk, who places an object beneath the Virgin’s right foot (Fig. 7.07A and detail in Fig. 7.07C). The Apostle to Mary’s right, presumably (given his privileged position and smooth-chin) John the Evangelist, is looking down at this donor figure and drawing him to the attention of his neighbours.

Vasselot’s catalogue of the le Roy bequest cautiously described the object being presented to the Virgin as ‘...un livre ouvert (?)...’.16   A careful examination however suggests that it is not a book but an oblong-shaped object with a slightly overhanging top face, just as this portable altar would have been before it lost its base (Fig. 7.07C). Limitations of the medium and style make it difficult to be absolute in this identification but the obvious difference between the presented object and the books held by various of the Apostles elsewhere on the panel (two are visible in Fig. 7.07A) seems to support this identification.17  The implication that the prostrate monk is depicted as presenting this portablealtar (i.e. the physical object on which the presentation [182] image appears) to the Mother of God is strengthened by both the content and the alignment of the lengthy inscription which runs clockwise all the way around the altar. This begins lower left beneath body of the donor, and ends in the middle of the bottom margin, directly beneath the act of donation that it records, with an entreaty to the Virgin to ‘...accept the gift of your servant Rudolfus’.18 Showing the altar itself being offered to, and accepted by, the Virgin makes potent claims about the merit due to the object, as well as to its giver. Placing oneself beneath the foot of a ruler was an age-old sign of fealty. The depiction of such an act here serves not only to indicate Rudolfus’ claim of vassalage but also, crucially, Mary’s acceptance of it (something which the image can suggest though the text cannot). That the Virgin’s foot is resting on the portable altar also sets in train a sequence of metaphors concerning Mary as ‘the altar that bore Christ’ - here herself borne on an altar tablet that will bear the body of her son each time it is used.  By prominently marking this portable altar with its own ‘autobiography’, in both verbal and visual forms, Rudolfus staked a powerful claim for a place in the memories, and intercessory prayers, of all the priests who would use it after his death. Nothing else is known about this Rudolphus but whoever he was, he was a shrewd investor.

Another example of a presentation-scene, one which perhaps takes mise en abyme to a second level of embedding, can be found on an orphrey now in the Musée de Cluny. This opulent textile in coloured silks and golden thread originally adorned the back of a now lost chasuble. It was embroidered in England or Northern France around the end of the thirteenth century and was used in the monastic church at Vergy (Burgundy).19

The orphrey itself is divided vertically into three rectangular fields, each with detailed embroidered captions explaining their contents (Fig. 7.08). In the upper panel is a conventional iconic/hieratic scene of the Virgin and Child enthroned, flanked by Saints Viventius and Peter. The middle panel shows a tonsured and bearded monk kneeling in front of an altar (equipped with a cloth-covered chalice, ready for the Mass), towards which he holds out a chasuble. The scene is set within a fictive architectural frame representing a transverse section through a three-aisled basilica, through the upper parts of which are threaded a Latin inscription reading:

Before the altar, Brother Peter offers this complete sacerdotal vestment.20

[183] Beneath this scene, in the lowest register, are two kneeling secular figures, also clearly identified by an inscription:

Count Manasses and Countess Ermengarde, founders of the monastery known as Vergy, offer it to God, to Saint Viventius, to the Blessed Virgin and to Saint Peter.21

These kneeling figures, who are not themselves contained within a microarchitectural frame, are shown with their hands raised aloft, not in conventional attitudes of prayer but supporting the architectural stage-set within which ‘Brother Peter’ is making his own offering. There is thus a double mise en abyme here - the presentation of the chasuble is shown on the back of that garment itself, within an image of the church where that presentation was made and where the chasuble was subsequently to be used. In terms of a conscious strategy of narrativising the gift, this image thus has a twofold function. Firstly, the ‘conventional’ mise en abyme of the garment shown within the garment trumpets to posterity that one ‘Brother Peter’ (whose identity remains otherwise unknown) personally presented it to the church. Secondly, it situates Brother Peter’s act of generosity within the broader narrative of ‘pious giving’, both in general and specifically within the monastic community at Vergy, positioning the donor within the chain that leads from the first founders up to their ultimate dedicatees; God, Saint Viventius, the Blessed Virgin and Saint Peter. Indeed, when one looks at the orphrey as a whole, Manasses and Ermengarde are lifting Brother Peter skywards to help him reach their mutual celestial patrons - if not standing on the shoulders of giants, Brother Peter is at least raised up by the hands of the primi fundatores. It is true that the orphrey of the chasuble being presented by Brother Peter does not closely resemble the one on which it was mounted, nor does the church seen in cross section resemble what is known about the Romanesque structure built at Vergy in the early twelfth century.22 These mismatches however simply serve to demonstrate how little an accurate resemblance between object and image mattered in constructing the latter to be read as mise en abyme, since the deictic pronoun in both embroidered inscriptions (‘this church’, ‘this chasuble’) makes the identifications explicit.

The Manasses Orphrey also serves as a timely reminder of the temporal complexity inherent in mise en abyme, combining as it does the historical past of the founders’ gift (analepsis), the immediate-present of Brother Peter’s gift (though strictly speaking, donation scenes are always proleptic since the object must be made, and the presentation depicted upon it, before it can be thus presented) and the eternal or timeless state of the dedicatees in the uppermost register.

[184] Another example of this more explicit approach to mise en abyme may be found in the fascinating visual colophon that rounds off the six hundred or so pages of the Toledo Bible moralisée (Fig. 7.08A).23 The lower right hand quadrant of this page depicts a lay scribe working on a sheet of parchment on which he has already laid out the borders of his mise en page (Fig. 7.08C) - an arrangement of eight roundels in two columns with a blank column between, which instantly identifies it as a page of a Bible moralisée.  In the lower left quadrant of the page, a seated cleric reads from the open book on his lectern and, with commanding gaze and forefinger raised, directs the work of the artist in the adjoining frame.  Thus, as with the earlier example of Hartker’s Antiphonary, the artist of the Toledo colophon has employed mise en abyme to create a narrative image that not only shows the act of making but which tells a deeper story of the book’s creation - not necessarily as it actually happened but at least as its makers wanted it to be understood. The presence of the enthroned king and queen figures in the upper quadrants of this page also extend the book’s causal chain upwards, in much the same way that the  subject [x] in the phrase [x] me fecit denotes the patron as often as it does the artificer. Thus this visual colophon depicts the books autobiography in terms of the three levels of agency responsible, as if saying; ‘monarchs commissioned me, clerics designed me, craftsmen illuminated me’.

7.d) mise en abyme as a strategy for contextualising the object

As well as helping to visualise an object’s own narratives of creation and consumption, mise en abyme could also serve to re-contextualise the object by depicting it being used within a different visual narrative. In particular, depicting an object (or simply one that loosely resembles it) in some privileged position might enhance the perceived value of the object itself.

Although the narratives of Limoges enamel chests are discussed in detail elsewhere (chapter 8.1), I would like to mention briefly an enamelled object of a very different type - one of a small group of ciboria, which are believed to have been made in Catalonia or the Roussillon in the mid-fourteenth century. One of these, supposedly from the Aragonese royal foundation of Santa Maria de Poblet (hereafter the ‘Poblet goblet’) and dating from around 1330-40,  is now in the National Museum of Washington (Fig. 7.10).24   These vessels generally consist of a bowl mounted atop a footed stem, rather like a chalice but with the addition of a domed cover surmounted by a knop and/or a cross. Normally the cover is hinged, with a clasp on the other [185] side to lock it closed. Inside the cup was a  smaller container (the capsis) which held the consecrated host. Unlike a monstrance or a pyx cage which served (at least after the Fourth Lateran Council)  for displaying the consecrated host over an altar, these ciboria were used primarily for delivering the host to those physically unable to attend mass in person (particularly to the sick or dying - the sacrament ad viaticum infirmis).

Most of the surviving ciboria of this type are decorated with Christological imagery and the Poblet goblet is no exception.  One segment of the hexagonal foot is decorated with an isolated extra-narrative image of the enthroned Christ offering his blessing, whilst a single narrative scene of the Adoration of the Magi, arranged centrifugally, spreads around the other five (Fig. 7.10B). In this scene, the eldest king kneels to make his offering of gold - which takes the shape of a footed and covered cup just like the vessel on which the scene is depicted.  In itself this is hardly surprising since fourteenth century depictions of the Epiphany routinely depict the Magi’s gifts in the form of such covered goblets - see for example the scene from the Très Belles Heures de Notre Dame (Fig. 7.10D) - however in this context it takes on an extra resonance.  The self-reflexive image on the foot of the ciborium, as is generally the case with mise-en-abyme, gives an immediacy to the scene, but it also validates the physical form and the function of the vessel that bears it. When it was in use, the incarnated Christ - to whom we see a precious cup being offered in the Epiphany image - would have been incarnated again (in the form of the transubstantiated host) and presented to the sick or dying communicant from just such a cup. Emphasising further this association with the two bodies of Christ, the lid of the cup has four more scenes from the Incarnation cycle (Annunciation, Visitation, Nativity and Annunciation to the Shepherds), as well as two scenes from the Passion positioned to align with the faces of the cross mounted on the finial; the Crucifixion (to remind the viewer of Christ’s own final suffering) and the Resurrection (a reminder of the eternal future promised to those who eat of His body). The Resurrection scene on the lid is positioned directly above the image of Christ receiving the goblet in the Epiphany scene on the foot. Thus when considered in the context of this object’s use as a container for the viaticum, the narrative scenes it carries serve to ‘frame’ the ciborium and its contents - helping to condition the communicant into the right receptive mode, not just for the story of Christ’s incarnation, sacrifice and rebirth - but also for receiving his body, perhaps for the last time.

Examples where mise-en-abyme was used within narrative images to condition the reception of an object and its stories are surprisingly common once one starts looking for them, particularly in manuscript illuminations. However, for the sake of variety, I will end this particular thread with some embroidered hats.  There are at least three surviving Opus Anglicanum mitres from [186] the early thirteenth century which are embroidered with the martyrdom of Thomas Becket, two of which are shown in Fig. 7.11.25

The first is preserved in the Trésor des Soeurs de Notre Dame at Namur and is said to have belonged to Cardinal Jacques de Vitry (author of the Historia Hierosolymitana, who died in 1244). The front (not shown) depicts the martyrdom of St Lawrence, held down on his griddle by two executioners with pitchforks. On the back is the martyrdom of Becket (see Fig. 7.11B). To the right of the scene is an altar dressed with two candlesticks and an uncovered chalice, indicating that mass is taking place (although of course Thomas was not celebrating mass at the time of his death, the majority of representations show him doing so). Thomas is shown in the centre,  falling forwards with his hands extended in prayer towards the altar. Three knights are shown - the second is striking a blow to Becket’s head (the sword is intact and no blood shown). The hand of God emerges from a cloud-frill at the top, through a titulus reading  SCS TH.OMAS.   My second mitre, now in Munich, has lost its lappets. One side shows the martyrdom of St Stephen; the consul is seated far left with a sword over his right shoulder, directing three executioners who stand with stones raised aloft.  Stephen is shown kneeling, hands orant, with three stones around his head. The other side again shows the martyrdom of Thomas in a similar fashion to the Namur mitre (Fig. 7.11A). This time Thomas faces away from the altar and the first knight strikes the top of his head, from whence gush three spurts of blood (originally picked out in bright red silk threads). Again the hand of God emerges from a cloud frill at the top and again the titulus reads SCS TH.OMAS. The third mitre, in the Treasury of the Cathedral of Sens, is almost identical to the one in Munich but is less well preserved. Given how few mitres of any kind survive from this period, to have three showing the martyrdom of St Thomas is a remarkably high number, suggesting that it may have been a relatively common narrative for inclusion in this context.

The Munich and Sens mitres both show the broken sword tip (see Fig. 7.11A) - an element of the narrative that appears in various media and was included to emphasise the violence of the blow aimed at Becket’s head. Less common however is the detail of Becket’s own mitre, which all three versions show lying on the floor beneath his falling body, having been knocked off his head by his attackers’  first blow. In this respect, these images differ markedly from the depictions of Becket’s martyrdom in other contexts, such as on reliquary chests (see Fig 8.06 for examples)  which, although they often show a mitre on the Archbishop’s head, never depict it at his feet. In most narrative settings, the mitre rightly belongs on Becket’s head as an attribute to help identify the hero of the story. Yet when Becket’s mitre appears within the visual narrative embroidered onto a real mitre, it can be shown dramatically knocked to the ground by the first [187] sword-blow. In this context, the mitre en-abyme becomes a lens, focussing not just the violence inflicted on the Sainted Thomas but also the outrage done to his office and to the Church at large (represented metonymically by the insulted mitre). The Namur mitre retains its lappets (Fig. 7.11B) and with them its sense of direction - from which we know that the scene of Becket’s martyrdom was embroidered on the back of the item.26 Thus the participants in the masses at which this mitre was worn would have enjoyed a ring-side view of the violence done to Becket’s head, clearly displayed on the back of the head of their own bishop. For this select audience, the small but clearly visible detail of Becket’s own mitre violently dashed to the ground must have added a certain piquancy to events - especially on December 29th, when the saint’s feast was celebrated and when the story, read out during the canonical hours, was fresher than ever in people’s minds.

7.e) conclusions

Reviewing the art historical literature, one might get the impression that both the use and the theorising of mise-en-abyme belong exclusively to the early-modern period and its historiographers. As this chapter has shown however, the deliberate use of self-reflexive images has a much longer pedigree than this. Picturing an important event or activity involving an object upon that object itself (or upon an object of the same type) has a number of functions, all of which converge on the contextualising or framing of the object specifically in relation to the depicted action (as distinct from all the other subsequent uses it might be put to). In this respect, mise-en-abyme can have what Barthes, in a somewhat different context, referred to as an ‘anchoring’ function, restricting the floating chain of semiosis to a particular set of meanings associated with the specific usage depicted in the image.27 More importantly, the object can be valorised by showing it in a highly honorific setting, for example being used, or received from a donor, by an earthly or celestial patron.

Precious objects are constantly evolving as their ownership and usage changes – a precious cup may start its life as a secular drinking vessel, be converted to a chalice for use in a church, then looted or sold and end its days preserved in a museum as a cultural artefact or objet d’art. Each transition marks a change in the role of the object. If however that cup is inscribed with a scene showing its donation or its intended use, it will retain traces of that original function throughout its subsequent history. Narrative mise-en-abyme is thus a potent guardian of an object’s identity.


Previous Section Contents Page Next Section


1 A. Gide, Journal 1889-1939, Quoted in L. Dällenbach, The Mirror in the Text, Cambridge 1989 [1977], p. 7.

2 Claude-Edmonde Magny, 1950, Histoire du roman français depuis1918, Paris. The first comprehensive attempt at a ‘poetics’ of mise en abyme was Lucien Dällenbach’s ‘Le récit spéculaire: essai sur la mise en abyme’ (Paris, 1977) - later published in English as ‘The  Mirror in the Text’ (trans Jeremy Whitely with Emma Hughes, Cambridge, 1989).  Moshe Ron’s essay ‘The restricted abyss: Nine problems in the theory of mise en abyme’ in Poetics Today (vol. 8:2, 1987) provides a useful summary of Dällenbach’s arguments as well as responses to various criticisms that have been aimed at them. See also, J. J. White, 'The Semiotics of the mise en abyme' in The Motivated Sign: Iconicity in language and literature 2, ed. O. Fischer and M. Nänny, 2001, pp. 29-53 (especially p.33ff).

3 Most famously in V. Stoichita, The Self Aware Image: An Insight into Early Modern Meta-Painting, Cambridge 1997.

4 In an attempt to address this neglect, mise en abyme was chosen as the topic of the first conference in the ‘Medieval Art in Theory’ project, instigated and organised by the author and Dr Laura Cleaver under the auspices of the Courtauld Institute’s Research Forum. The conference was held at the Courtauld Institute on 16th February 2009 and featured papers dealing with the use of mise en abyme in a wide range of medieval artistic and architectural contexts. The present chapter is based on the author’s introductory address to that conference.

5 For various reasons, the literature on its use in Las Meniñas probably exceeds what has been written about all other examples of artistic mise en abyme combined. For an introduction and bibliography see J. R. Searle, '"Las Meniñas" and the Paradoxes of Pictorial Representation', in Critical Inquiry, 6(3), 1980, pp. 477-88.

6 meta-narrative in the sense that the device does not normally form part of the story but rather conditions its reception.

7 The choice of this manuscript was fairly random though rather fortuitous given the artist’s predilection of showing images of books. The only monographic study is M. Schapiro, The Parma Ildefonsus : A Romanesque illuminated Manuscript from Cluny and Related works, New York 1964.  Ildefonsus was a 7th century bishop of Toledo, best known for his De perpetua virginitate Mariae contra tres infideles, which forms the core of the Parma manuscript..

8 For a description of the contents of the Parma manuscript, see Schapiro, 1964 (ibid), pp.7-8. The copy written by Gòmez in 951 is Paris Bib. Nat. ms Lat. 2855.

9 For a general discussion of this presentation miniature see L. Campbell, 'Rogier van der Weyden and manuscript illumination' in Manuscript painting in context: Recent Research, ed. E. Morrison and T. Kren, 2006, Chapter 7.

10   Quoted in C. Lemaire, 'Les peregrinations des trois volumes des Chroniques de Hainaut' in Les Chroniques de Hainaut, ou les Ambitions d’un Prince Bourguignon, ed. C. Van den Bergen-Pantens, Turnhout 2000, pp. 29-32. In fact, none of the volumes described in the Burgundian inventories of that period match the binding shown in the miniature. There is no evidence of any previous bindings prior to the black satin cover.

11 An overview of the glazing programme at Tours can be found in CVMA, Recensement II - Les vitraux du Centre et des Pays de la Loire, Paris 1981, pp. 120-32. A full set of panel by panel images of all the surviving 13th century windows in Tours Cathedral will be available from mid-2010 at

12 The CVMA survey of the Le Mans glass is published in Recensement II - Les vitraux du Centre et des Pays de la Loire, Paris 1981, pp. 241-57. For illustrations of the glass, E. Hucher, Calques des vitraux peints de la Cathédral du Mans, Paris 1854-64 is a gargantuan two-volume set of life-sized and scaled plates based on 19th century tracings of the glass but mainly deals with the windows in the chapels. Images of all the windows of the inner and outer clerestories are available at (checked 03/01/2010)

13 The prosopopoeic formula ‘[X] made me’ or ‘[X] had me made’, recording the agency of either the artificer or their patron, has a very long tradition in many cultures. In Europe it was particularly strong within Celtic communities, where certain inanimate objects, such as swords, were routinely imbued with personal identities, though it also survived well into the Christian age. For classical and early medieval uses of prosopopeia see C. B. Kendall, The allegory of the church : Romanesque portals and their verse inscriptions Toronto 1998, p.82ff .

14 C. de Hamel, A History of Illuminated Manuscripts (2nd Ed) London 1994, pp. 19-21 and Pl 10.

15 In Carolingian law, for example, the accepting of gifts from a lord in excess of the value of one gold solidus was taken as evidence of the recipient’s acceptance of vassalage (see M. Bloch, Feudal society: Volume 1, The growth of ties of independence London 1965,  p.163ff. ) The classic study of gifting from the viewpoint of social anthropology is M Mauss, Essai sur le Don, Paris, 1925

16 A. J. J. Marquet de Vasselot, Catalogue raisonné de la Collection Martin Le Roy, Paris 1906.

17 Those who prefer to see it as an open book  (albeit a very thin one held apparently open at a curious angle and faced down) would need to find some other explanation for why the donor should be placing an open book under the Virgin’s foot.

18 The actual phrase is ...RVODOLFI FAMULI SUSCIPE DONATI.Famulus’ is a particularly servile label, meaning slave or servant but especially those attached to a household - I am informed by Dr Joanna Cannon that this is quite a common usage in such dedications of gifts to the Virgin.

19 Musée de Cluny, Inv. CL2158. Described in A. G. I. Christie, English medieval embroidery Oxford 1938, Cat.43. Christie’s description is a little inaccurate, particularly in misidentifying the donor(s) of the chasuble (mid 13th-century) as being the original founders of the monastery (late 9th-century). A more accurate account can be found in O. Brel-Bordaz, Broderies d'Ornements Liturgiques XIIIe-XIVe siecle, Paris 1982, Cat.1 pp.127-29 and pp. 171-72.


Manassès was Count of Autun, Châlon and Dijon, fl. 894-918. His wife Ermengarde was daughter of Boson, King of Provence - see O. Brel-Bordaz, Broderies d'Ornements Liturgiques XIIIe-XIVe siecle, Paris 1982, p.127-8. 

22 The observations about the differences between the original form of the church at Vergy and the fictive architecture depicted on the orphrey were first made by Chanoine Jean Marilier and quoted in O. Brel-Bordaz, Broderies d'Ornements Liturgiques XIIIe-XIVe siecle, Paris 1982, p.172

23 For a deconstruction of this page in the style of medieval exegesis see J. Lowden, 'The Bible of St Louis as a Bible moralisée' in The Bible of Saint Louis - Commentary Volume, ed. R. G. Ruiz, Barcelona 2004, pp.147-51. Although it originally closed the third volume of the Toledo Bible, this page is now part of a separate fascicle of eight pages in the Pierpoint Morgan Library (ms.240).

24 Inventory 1942.9.279(C-3). See R. Distelberger et al, The Collections of the National Gallery of Art: Western decorative arts Washington 1993, pp.41-45. Around a dozen such ciboria are known to survive, including one in a private collection in New York (see Distelberger, ibid, p.44) and another in the Musée de Cluny (inv. CL. 19963)

25 The three are in A. G. I. Christie, English medieval embroidery Oxford 1938, as Cat.20 (Trésor des Soeurs de Notre Dame at Namur), 21 (Bayerisches Nationalmuseum, Munich) and 22 (Trésor du Cathédrale Saint-Étienne de Sens).

26 Lappets are the two long strips of cloth that hang from the back of a mitre - without them it is not always clear which way round the mitre is. The Munich and Sens mitres have lost their lappets and lower borders and with them any sense of direction.

27 For the concept of ‘anchorage’ see; R. Barthes  Image-Music-Text. London 1977, p. 38ff. Although Barthes was talking about the role of textual captions in advertising imagery, in practice any semantic enclave tends to act as an anchor on its container.


Previous Section Contents Page Next Section