Friday, December 12, 2014

Iconature - 2.) Image cycles with deeds of exceptional individuals

*** I am sorry that the footnote links don't work. I made the mistake of switching back and forth between the Compose and Edit HTML view as I was writing this post and apparently that messes up the footnote code. I suggest you open this post in two tabs - one for the main text, one for the footnotes. That should allow for pretty smooth reading. ***

In trying to convince you that it is not irrelevant how we define the field of our study and what terminology we use, and that to adopt the term 'iconature' would not be equal to buying a white elephant or building a bridge to nowhere, I will start with an example. While, admittedly, it is not an example of where the mere existence of the term 'iconature' would have solved all the problems, it certainly is an example of how wrong terminology can create wrong concepts of thought, which in turn create interpretative puzzles and convoluted theories and theoretical impasses.

The main part of this example is taken from Byzantine art history, which, I am told, is not everyone's cup of tea. If you do not share my passion for it, please bear with me, focus on the main message, and I will try and make your journey through it as painless as possible. If, on the other hand, you are knowledgeable in the field, please excuse what to you may be unnecessary examples and explanations. This is not a specialist blog, but one which I would want to be accessible and comprehensible to anyone. I will try and put the lay person's satisfaction in the main body of this text; the specialist should, of course, look for his/her satisfaction in the footnotes.

This will be a long post, and a very academic one to boot. If you want a short version of it, here it is: iconature, just like literature, comes in genres. And if you just flip through the images below, you will see one of its genres - something bigger in the centre, surrounded by something smaller on the sides. It has made its appearance in very different periods of history, in very different cultures, and in very different media. And therefore it can best be understood as a genre of iconature, rather than as a heap of unconnected appearances to be studied within the domains of icon painting, ivories, sculpture, book art, movie posters, comics, and whatnot.

And now here's a 30,000-word version of that.

I. On the origin of the vita icon

      I.1 Ševčenko's puzzle

In a pioneering work on the vita icon1 , Nancy Patterson Ševčenko has left little to doubt concerning her definition of the object of study: 'In the middle of such icons is the image of a saint – sometimes the figure is full-length, sometimes a bust – and around him, usually on all four sides, are little compartments containing scenes from his life'.2   (see Figures 1 and 2)

Fig. 1 Vita icon of St George, early 13th c., Monastery of St Catherine

Fig. 2 Vita icon of St Panteleimon, early 13th c., Monastery of St Catherine

She has thus delineated what she viewed as a distinct category of artworks, one which scholarship should understand and treat separately from all other icons incorporating some narrative element. Pictorial sequences illustrating texts such as the Akathistos hymn would not qualify for this definition, even though they are sometimes organised around a central portrait of the Virgin (Fig. 3), as they do not exclusively depict scenes from the holy figure's life.3

Fig. 3 Icon depicting praises to the Mother of God and strophes from the Akathistos hymn, middle 15th c.; originally from the Dormition Cathedral of the Cyrillo-Belozersk Monastery, now in the Russian Museum of Saint Petersburg 

And some icons that would satisfy this requirement, as for instance the illustrations of the Nativity or the Passion of Christ (Fig. 4 and 5, respectively), would not qualify on the grounds that the narrative sequence is not organised around a central portrait.

Fig. 4  Icon with the Christmas cycle, ca. 1100; Mount Sinai, Monastery of St Catherine

Fig. 5 Icon with Passion scenes, third quarter of the fourteenth century; the holy monastery of Vladaton, Thessalonike

Finally, certain narrative icons that would seem to meet both of these qualifying conditions, as is the case with the thirteenth-century icon of Hodegetria Aristerokratousa from the monastery of Saint Catherine on Mount Sinai (Fig. 6), have also been left out of the scope of Ševčenko's study, presumably on the grounds that the narrative images appear on the triptych wings rather than being directly adjacent to the central portrait.4

Fig. 6

 Fig. 7.1 

Fig 7.2

It is likewise very clear that Ševčenko felt the analysis should be limited to the artworks dating before the fall of Constantinople, as well as to areas under direct Byzantine influence. Contemporary vita icons from Italy constituted, in her opinion, only a ‘variant on the traditional Byzantine vita icon,’5 and everything created after the fall of Constantinople only an afterlife of this essentially Byzantine invention. The Slavic examples, regardless of their date, have been left out of the picture except for one St. Nicholas icon in Skopje.

Finally, the author has also argued that ‘[i]cons with framing scenes from the life of Christ or the Virgin’ should likewise be considered a special and separate case, as ‘[t]hey first appear somewhat later than the hagiographical ones, and most likely derive their form from them.’6

The object of study (for art historians focusing on Byzantine art at least, and especially if their interest was in the origin of this genre) was thus to be limited to the ‘less than two dozen surviving Byzantine examples’ ‘of the familiar sort, that is, one having scenes framing a central portrait.’ Thus defined, the vita icon has, according to Ševčenko, appeared rather suddenly as a novel form in Byzantium, sometime in the late-twelfth or early-thirteenth century, with no clear formal precedents. As ‘no one really [knew] how or where the form originated,’ nor what its function might have been, it presented something of a puzzle. Further complicating factors were the 'apparently unprogrammatic tone of these panels' and the lack of revealing dedicatory inscriptions. Ševčenko did note some strong formal parallels for this icon type in a number of works of Late Antique, Early Christian, and even Carolingian art, but, quite reasonably, she asserted that the vita icons were 'unlikely to have depended directly on such models.'7 And since she was, it would appear, firmly convinced that a direct, exact model was indeed needed, she at least briefly explored several possible options.

The idea that the scenes around the sides might have been thought of as wings, and that they would thus have had their source in ivory triptychs, did not seem convincing to her, because 'on our vita icons the scenes almost always go around all four sides, for which [she knew] of no parallel among Byzantine ivories.' As has been stated above, Ševčenko was very much aware of the fact that the very earliest artwork that could satisfy her definition of a vita icon was precisely a (painted) triptych (Figures 7.1 and 7.2). We can only speculate as to why she disregarded this detail in 1992, but in 1999, at least, she briefly explained that there seemed to be no later parallels for the triptych form among the surviving narrative icons. The latter only holds true, of course, if we accept her proposition that the examples involving the life of Christ and the Theotokos ought to be considered a separate case. We have already seen, in Fig. 6, that we do possess other icons where the narrative panels appear on the triptych wings rather than on the same panel as in the class of images that Ševčenko would call vita icons 'of the familiar sort.' And if we remember that the only reason why Ševčenko thought that the narrative icons of Christ and the Virgin should be omitted from our study was that they 'most likely derive their form from them,' it becomes obvious that this whole line of argument is in need of some revision. Given that the St. Nicholas icon on Mt. Sinai (Figures 7.1 and 7.2) is both the earliest surviving example, even by Ševčenko's definition, and that later occurrences of the triptych form do exist (Fig. 6), we should perhaps not so easily dismiss the possibility that the vita cycles were indeed originally viewed as wings.

Having rejected the idea of wings as a solution to her puzzle, Ševčenko offered a similarly cursory inspection of the epistyle cycles (painting cycles appearing on epistyles/architraves, such as this one) and declared this explanation to be equally implausible. Since we could not reconstruct exactly how 'such a strictly horizontal row of scenes end[ed] up wrapped around a central portrait,' she concluded that these cycles 'may have influenced the content of our vita icons, but (…) can scarcely have been the immediate source that we are seeking.' It must again be noted that the whole argumentation rests on the notion that the defining characteristic of this class of objects is that a central portrait is surrounded, ideally on all four sides, by narrative scenes. This means that all those objects where a central portrait was surrounded by any other type of an image but the narrative scene (Fig. 8) cannot have served as a model, because, although their formal structure is so often identical to that of the vita icons, we cannot reconstruct exactly how the leap from, for instance, portrait images, to narrative scenes in the frame of an icon might have occurred.8

Fig. 8   Mosaic icon of St Nicholas, 12th c., Monastery of St John, Patmos

In a word, we are asked to believe that it is beyond probability that the Byzantines would have drawn on their constructive imagination to such an extent as to play around with and combine some basic artistic forms. Additionally, we are also asked to set aside all the hagiographical icons that did not include a central portrait at all (Figs. 4 and 5). As they do not conform to the layout that the author regarded as a defining and specific characteristic of this class of objects, they are to be considered nothing more than rare departures from the norm. It is certainly to be admitted that epistyle cycles could not have provided direct and exact formal models for all hagiographical icons, but they might very well have done so for some, and we cannot discard such a possibility based on nothing but an artificially constructed definition of the vita icon's ideal type.9 What is more, if we allow the Byzantine artist and patron some freedom of creative thought, such as could combine the hagiographical sequence of scenes that we find, for instance, in epistyle cycles with small compartments that we so very often find in the icon's border decoration, then the innovative leap from epistyle cycles even to the 'familiar type' of the icons is perhaps not so implausible.

Fig. 9.1                                                                                                  Fig. 9.2  
Byzantine book covers with Crucifixion and Anastasis, 14th. c, Biblioteca Marciana, Venice (Cl. Gr. I, 53)  

Finally, Ševčenko mentioned even the option of this being a purely Western invention, pointing out that '[v]ita icons do crop up in the East and West at just about the same time,' and that 'in the East the surviving Byzantine icons derive almost exclusively from areas subject to considerable Western influence: Sinai, Cyprus, and Kastoria.' However, this option has received even less exploration, and the author would seem to have dismissed it purely on the grounds of this being too 'radical' a solution. (N. P. Ševčenko, ʻIcons and “Decorated” Icons of the Komnenian Period’ (as in n. 1), p. 61.)

Having thus concluded her brief survey of the possible, but in her opinion improbable, alternative answers to her puzzle, Ševčenko has turned to the solution that she favoured, namely the understudied case of the κεκοσμημέναι είκόνες ('decorated icons'). In these relatively rare artworks (which, in Ševčenko's opinion, happen to be much closer in time, context, and form, to the proposed invention of the vita icon genre), ‘a holy figure is surrounded by various smaller images’ on a metal frame. And while the author had to admit that she knew of only one example where these smaller images on the frame really appeared as contiguous boxes, and not separated by panels of ornament, and was therefore unsatisfied that she had discovered the origin of the form, she believed that she had at least reconstructed how this addition of a frame would originally have been perceived. We may not have come any closer to answering the question of how the form came into being, but, in the author's mind, we are at least closer to understanding how it would have been understood when it was first used.10

Ševčenko's claims, as stated above, together with her puzzle, have been largely accepted in all subsequent scholarship, beginning already with A.W. Carr, who published her article in the same volume that featured Ševčenko’s first generalised treatment of the form.11 Paroma Chatterjee, for instance, has challenged it only by reminding us that some scholars date the Cypriot vita icon of St. Marina to a considerably earlier period than that proposed by Ševčenko, and has not pushed this point any further.12 Hans Belting was content to add only a short clarification: ‘Biographical icons combine the portrait and the legend in a way that, despite parallels in ancient cult images of Mithras and Hercules, would seem hybrid, if they did not confirm the tendency to summarize and standardize. The portrait allowed the saint to be viewed as a person, and the accompanying short biography gave an idea of his or her life and miracles. (…) The small, shorthand scenes are merely an aid to memory; they presuppose a knowledge of the texts that were read aloud. The narrative element, which otherwise has no place in icons, annotates the physical portrait of the saint with the “inner” portrait of his or her virtues.’ 13

The present study does not aim to provide an alternative date for the appearance of the Byzantine vita icon, nor will it challenge Ševčenko's assumption that the direct origin of this form could be in the decorated (κεκοσμημέναι) icons. The problem with Ševčenko's thesis was that it sought to resolve the riddle of the origin of one class of objects by simply pointing to another class, whose formal origin was equally unknown, which brought us no closer to proper understanding of the matter.14 The rest of this post will, therefore, examine the possibility that the format of a central (iconic) image surrounded by a sequence of narratively connected scenes did not have to be used very close in time to provide a direct formal origin for the so-called vita icons. Its existence in various historical periods, cultures, and a wide range of media, which we shall examine here, seems to indicate that the form comes very naturally to the human mind. Having restricted her analysis to less than two dozen Byzantine images that conform to the rigid definition of the vita icon, Ševčenko has effectively played not only an assisting but indeed a formative part in the creation of the very problem that her study then set out to resolve – that of the possible formal origins of the vita icon format. Given that she had chosen to disconnect her corpus of artworks not only from all other types of icon that incorporate some narrative elements, but also from narrative art in general, it is perhaps only natural that she was then puzzled with the question of how a 'row of scenes [would] end up wrapped around a central portrait.' When, however, analysis is not limited by such restrictive definitions, and especially if it takes the longue durée approach, as well as prolonged looks at other media, and is on the lookout not so much for direct formal precedents, but rather for conceptual models, then Ševčenko's conundrum looks less difficult to resolve. It may even, in final analysis, be an entirely abstract question of the kind that is more proper to intellectual amusement than to actual reconstruction of historical developments. This study will attempt to show that, even in a culture as resistant to change as Byzantium is sometimes thought to have been, new image types did not invariably need clear, exact, and direct formal origins. Specifically, what shall be demonstrated here is how the vita icons fit into the long and wide tradition of narrating, describing, and praising the deeds of exceptional individuals. Within that tradition, we shall not look for exact and direct formal origins of the Byzantine vita icons, but for conceptual models and universal narrative strategies. Guided by the conviction that what is universal in terms of both time and space can be expected to (re)appear with no immediate predecessor, I shall attempt to show that Ševčenko was wrong to underrate the link between the Byzantine vita icons and the very similar artworks from other media and/or periods.

     I.2 Central iconic image framed by narrative scenes – a brief historical survey

As has been noted above, Ševčenko did acknowledge the existence of similar formal arrangements from at least as far back as the first century AD.15 She even offered a limited catalogue of pertinent examples, some of which will necessarily reappear in our survey here. However, this study will not only add to this catalogue but also draw quite different conclusions from the collection and examination of it.

          I.2.1 Tabulae Iliacae

We may begin our survey in exactly the same place where Ševčenko chose to begin hers – with the so-called Tabulae Iliacae. 16 Most of these twenty-two tablets have been dated to sometime between the first century B.C. to the mid-second century A.D. and thus present the earliest known examples of our form. However, although the so-called Tommassetti tablet in the Vatican (Fig. 10), which is the only one that Ševčenko mentions, is indeed the closest parallel to the form of the vita icons, I believe that the whole class of objects, rather than only this single example, is relevant to our study.

Fig. 10  Tabula Odysseaca Tomassetti, uncertain date between 1st c. BC and mid-2nd c. AD, 
Vatican Museum

Even on the most casual examination, it becomes perfectly obvious that, given the large number of narrative scenes included on these tablets, the larger central panel, whatever formal arrangement it may contain, invariably serves a very basic visual purpose – that of preventing monotony. It is beyond dispute that, for instance, the so-called Tabula Capitolina (Fig. 11), the Tabula New York (Fig. 12), or the one usually referred to as the Veronensis I tablet (Fig. 13) do not incorporate an individual portrait, or even a prominent, unattached figure, in their central fields, and that they, therefore, cannot be regarded as direct relatives of the vita icons.

Fig. 11   Tabula Capitolina, uncertain date between 1st c. BC and mid-2nd c. AD, 
Rome, Musei Capitolini

Fig. 12   Tabula New York, uncertain date between 1st c. BC and mid-2nd c. AD, 
The Metropolitan Museum of Art, NY

Fig. 13   Tabula Veronensis I, uncertain date between 1st c. BC and mid-2nd c. AD
Paris, Cabinet des MédaillesBibliothèque nationale de France

In that respect, the Tommassetti tablet, which Ševčenko has singled out, is indeed unique. These other tablets can, however, and perhaps better than the Tommassetti one, teach us a valuable lesson on the spatio-topical system17 of the narrative sequence. When an artist is presented with the task of displaying a significant number of narrative scenes, on any surface that can be taken in with a single glance, he necessarily runs into problems of balancing dynamism and clarity. While, on the one hand, a uniform encadrement coupled with a chronologically organised sequence18 will certainly ensure the easiest comprehensibility, the repetitiveness of such an approach may, on the other hand, generate monotony and cause the artwork to quickly lose the viewer's attention. In order to counter that, certain simple artistic mechanisms can be employed, and dividing the visual field into dynamically counterpoised units is perhaps the most obvious among such solutions.19 The Iliac tablets make use of the large central panel precisely for such reasons. Even in cases where the figures and the scenes in the central panel are no larger, and the action no less densely packed, than in the borders, the mere existence of the central field breaks the visual monotony of the uniform framework and compressed narration. It allows for greater legibility and visual variety by introducing a sense of differentiation, and perhaps hierarchy, among the scenes. It gives visual expression to the fact that certain parts of the story deserve a greater degree, or a different kind, of attention than others. In a certain sense, it even corroborates the viewer's estimation that this hoard of images should in fact be read as a story, rather than be viewed as a purely decorative network of figures. If the limits of the present study allowed, it could even be explored as one of visual art's equivalents of a literary narrator's voice.

The universality of this artistic device, employed for the same basic purpose, is attested to by its existence in various cultures, time periods, and media. We find it in archaic Greek art on an amphora from Canosa, now in Munich (Fig. 14), bearing a figural composition based on the Medea of Euripides. 

Fig. 14   An archaic amphora with scenes from Euripides' Medea, uncertain date, 
Antiquarium in Munich

We likewise find it in Carthage, in the late fourth century A.D., in the famous mosaic of Dominus Julius (Fig. 15). 20

Fig. 15

It is also the organisational principle in a sixth-century Egyptian textile panel depicting Dionysios and the labours of Heracles (Fig. 16). 21

Fig. 16

In the East, we find it as early as the beginning of the ninth century, in a painting on silk, found in the Chinese caves of Dunhuang (Fig. 17). 22

Fig. 17

We even find it in an embroidered handkerchief (‘rumal’) from the sequestered Indian district of Chamba, depicting an episode from the Ramayana (Fig. 18). 

Fig, 18

And, even though among these five examples, picked out basically at random from a plethora of similarly structured artworks, only the Egyptian and the Chinese one actually contain central portraits, they all, in addition to the Iliac tablets, illustrate for us the compositional role of such a central panel. Since there can be no talk here of dissemination from a common origin, we can confidently conclude that its role indeed is primarily a compositional one and that its appearance does not necessarily depend on any culturally determined factors. The choice of the form stems rather from universally human laws of visual perception and should be considered one of, so to speak, natural solutions to the task of organising a large number of narrative panels in a single visual field. As has been very convincingly argued already by Rudolf Arnheim, our minds display concurrent needs for both complexity and order23 and the introduction of such clearly demarcated central panels into a clutter of virtually homogeneous narrative frames is clearly a case of achieving balance between these two competing tendencies.24 Even when the visual contents of the central panel are in no way different from those of the surrounding areas (which now inevitably encourage us to perceive them as a sort of border), the mere existence of a framing device, no matter how unpretentious, introduces a degree of order that balances out the cacophony of the sometimes many dozens of miniature narrative scenes.

The parallels between the Illiac tablets and the Byzantine vita icons, then, are much more than purely formal ones and should not be examined only insofar as we can establish a direct historical lineage between these two classes of objects. As with all the artworks that shall be included in this survey, the relationship between the Illiac tablets and the Byzantine vita icons is not of genealogical nature. What we are looking at here is reincarnation rather than descent and, though this realisation cannot provide any answer to Ševčenko's puzzle delineated above, it can certainly help us to better understand the Byzantine vita icon as a work of art. Once we have realised that this formal scheme appears as a natural consequence of the intrinsic laws of visual narration, as well as that these laws, in turn, depend on universally human laws of visual perception (rather than on historically and culturally conditioned formal arrangements that need clear trajectories of dissemination), a new understanding of the vita icon naturally presents itself to us. It becomes apparent that, though it can be considered a hybrid form, it should not be considered a genre of icon painting in the sense that an iconic portrait has somehow attracted an accretion of narrative panels. When we start to perceive the vita icon as a genre of iconature (or, at least, of narrative art) we see that it is rather the series of narrative panels that asks for some sort of structural device that would prevent visual monotony, counter complexity with order, and essentially connect a series of images into a meaningful work of art. For this purpose, a central frame, and especially when it contains something like the narrative protagonist's portrait, which so dynamically differs from the rest of the visual field, is so plainly adequate that its universality should come as no surprise.

          I.2.2 Image cycles with deeds of exceptional individuals

If the Illiac tablets illustrate for us only the general structural interplay of the small but numerous narrative scenes and the prominent, detached central panel, our next group of objects will bring to fore the particular narrative role that such a panel served within a very specific genre. We shall now begin to examine ancient cycles depicting the deeds of exceptional individuals, which is the genre of iconature that the vita icon will become heir to. Out of this class of objects, Ševčenko has singled out as directly relevant but a handful of examples, again only as passing remarks and in the sense that, while there is certainly some superficial similarity in the formal arrangement between these artworks and the vita icons, the connection cannot, in the last analysis, be very strong. 25 A prolonged look at the existing material, and one that pays attention not only to formal similarities with the vita icons, but also to the narratological reasons for which these similarities appear, will, once again, prove that the connection is in fact much stronger than it might be expected.

Ancient art knows several characters whose deeds have been developed into cycles of images. Out of these, we shall pay particular attention to the cycles depicting the labours of Hercules and only briefly touch upon those dedicated to others.26 The deeds of Hercules will serve our purpose best for several reasons. Firstly, because the Hercules cycles seem to be the earliest in this line. Secondly, because some of these artworks afford the closest formal parallels to what Ševčenko would call the canonical vita icon form. And thirdly, because their pictorial formulas have been canonised in a way very similar to that which we later find in the vita icons.

Although depictions of Heracles, and even shorter image cycles of his deeds, can certainly be found earlier, the canonical Dodekathlos makes its first apperarance in the metopes of the fifth-century temple of Zeus at Olympia. The number of twelve deeds represented there was, in all probability, a result of nothing more than space limitations, but this famous cycle will, nevertheless, provide an influential prototype that will remain in use for many centuries afterwards. It will appear essentially unaltered in different media and across wide ranges of space and time.27 Among these numerous and variegated manifestations, at least four examples, in addition to the Egyptian textile fragment mentioned above (Fig. 16), are directly relevant for this study. 28 These are the Roman votive relief of Cassia Priscilla (Fig. 19)29 , two floor mosaics in Spain – one found in Liria (Fig. 20), the other one in Cartima (Fig. 21), and the medallion of Gordianus (Fig. 22). All of them include both a central portrait (sometimes of Heracles alone, sometimes with companions) and depictions of the hero’s labours arranged around it.30

Fig. 19   Votive relief of Cassia Priscilla, 2nd c. A.D.
Museo Nazionale, Naples.

Fig. 20   Floor mosaic with Heracles, Omphale and depictions of Heracles' Twelve Labours,
found in Liria, 3rd. c. AD, National Archaeological Museum of Spain, Madrid

Fig. 21   A reconstruction of the Hercules mosaic found in Cartima, uncertain date
As in: E. Hübner, Die antiken Bildwerken in Madrid (Berlin, 1862) no. 827, p. 310

Fig. 22  Medallion of Gordianus, Hercules in the centre, with the depictions of his Twelve Labours radiating around him.
As in: R. Bräuer, ‘Die Heraklestaten auf antiken Münzen,’ Zeitschrifft für Numismatik, 28 (1910), pl. V:20

Fig. 23  Textile panel with Dionysios and the labours of Heracles, Egypt, 6th c. A.D.
The Hermitage Museum, Saint Petersburg, Russia.

The argument that these monuments stand at the beginning of a genre that the vita icons will perpetuate in their time – namely that of narrative cycles coupled with the portrait of an exceptional individual, whose deeds they promote and acclaim – should, of course, not be taken without a few standard reservations. First of all, it goes almost without saying that the term 'genre' is used here, like many others in this study, primarily for lack of a more developed and more suitable terminology. It should not be taken, therefore, in the sense that its creators and their original audiences would have recognised it as a distinct category of artworks defined by clear stylistic conventions. What is meant here is rather that the formal arrangement we have observed already in the Iliac tablets was now being recognised as particularly suitable for this subject matter. Its suitability is grounded in the simple fact that a prominent central portrait not only captures the viewer's attention more easily than any narrative sequence can but also focuses it on the narrative's main character. While the sequence of narrative panels invites our gaze to move around the artwork, inspecting the various deeds of the protagonist, the central portrait stands both at the beginning and at the end of our viewing process, affording us a closer look at and a more immediate connection with this exceptional individual. That there is symmetry in the arrangement of the narrative scenes, or that they are arranged around this portrait, rather than to be clustered, say, only above it, has certainly nothing to do with any particular precedent such as was sought by Ševčenko. It is simply that we find such arrangements more harmonious and pleasing to the eye than any other variation. This genre (or rather this bond between the formal arrangement we are inspecting and this particular subject matter) appears, therefore, without much conscious deliberation on the part of the creator, and is kept in existence simply because, while successfully performing its purpose of offering visual encomium, it also manages to please our eyes and minds. 

Our second reservation regarding the claim that we are looking at the beginning of a genre might be that the basic signification of these artworks was by no means unfluctuating. Quite to the contrary, their meanings showed a great deal of unsteadiness both synchronically and diachronically. The case of Hercules is again a very suitable illustration of this. Not only would a cycle depicting his labours in the metopes of a Greek temple carry a vastly different meaning than one appearing on a Roman sarcophagus, but the passage of time would dramatically affect their meaning even within the same class of objects. What is more, this holds true even for cycles that show not only the same selection and order of scenes, but also the exact same iconography. For instance, as Peter Jongste has noted,31 the labours of Hercules adorning the Roman sarcophagi from around 150 A.D. can be safely interpreted as metaphors for the owner's quotidian struggles and an expression of his hope of posthumous reward. By the year 170 A.D., though, we find that the situation has changed dramatically. Even though the cycles differ only in quality, while the iconography remained perfectly unaltered, the customers' only interest was now usually in finding a decent container for their remains. The iconography of the labours of Hercules in this age, thus, no longer pointed to anything more than the ambitions of the nouveau riche middle class to imitate the aristocratic practices of the preceding period. Such questions of semantics should, however, not overly concern us here. Whatever the intended meaning of some particular utilisation of it might be, we are examining this image type primarily as a formal arrangement. And, since we have defined the genre under examination here as the depiction of the deeds of exceptional individuals, we should not have too many qualms about bracketing together the works that do fit this relatively simple criterion but differ in their particular meanings. We see that sort of semantic variation in any genre of any art, and within all of their subcategories, and, as long as it is understood that calling this a genre does not imply a set expression of its connotations, we should not consider this in any way problematic. 

Finally, the very fact that we are bracketing into a single genre something that appears in such a wide variety of media could, of course, raise similar concerns. In addition to what has been stated above (that the very term genre should be taken with the proverbial grain of salt, and that we are examining this image type primarily as a formal arrangement), it must be underlined here that there is nothing truly unconventional in recognising the same image category in different media. While it is certainly true that a representation of the labours of Hercules in a floor mosaic and one in a private funerary relief would not be executed either for the same purposes, by the same workshop, or perhaps even for the same audience, to hold that they could, therefore, not belong to the same category of images would amount to claiming that the ancient viewer could see no connection between them at all. In short, as long as it is understood that this genre (like any other, for that matter) is not to be considered perfectly monolithic and undifferentiated, a very strong case can be made that these images indeed form a distinct and recognisable category.

These reservations being, perhaps somewhat overcautiously, stated, we may now proceed with the analysis of our five Hercules cycles. Out of this group, the closest parallel 32 to the vita icons is probably afforded by the relief of Cassia Priscilla, which consists of a central frame with standing figures of Hercules and Omphale, depictions of the twelve labours of Heracles on the top and lateral sides, and the two main figures' attributes at the bottom of the assemblage (see Fig. 19)33. The relief is unique in this group of objects in that the faces of Hercules and Omphale are, in all likelihood, portraits. We may thus suppose that its patrons had at least some active role in its creation and have not simply picked one of the templates from the workshop's standard repertoire. 34For whatever reasons the Heracles theme might have been chosen, however, it is by no means clear on what textual or pictorial source the narrative cycle was based. Although all the deeds match the selection from the temple of Zeus at Olympia, their sequence does not seem to follow either the order of the metopes as described by Pausanias35 (Table 1), or the order of the deeds that has been preserved and canonised by the Bibliotheke of Pseudo-Apollodorus36 (Table 2).

It should be noted, though, that while these two sources are the most well known to us, there is no reason to believe that ancient artists and their patrons would have considered them to be prescriptive, or even that they would always have been directly familiar with them. The labours of Hercules were not told in any single ancient work – they were, rather, retold in numerous versions, none of which point to any authoritative source.37 The artists could have used for their model any of these other sources that are known to us today, or perhaps even some that have not been preserved or identified. For instance, although Hyginius’ thirtieth Fabula might seem to us like a much less important source than Pseudo-Apollodorus’ Bibliotheke (on the basis of which we have established what we now recognise as the canonical Dodekathlos and its order of deeds), this source seems to offer the closest known parallel to the sequence we find in our relief. If we suppose that the deeds were not arranged around the central panel at random, but that they were meant to be read in a logical way, then their order would, in all probability, be either as in our Table 4 or as in Table 5. The top row would be meant to be read first, from left to right, offering the first six labours, and then the sequence would continue either down the left side, followed by the right, or across the artwork, in a double-Z path, from top to bottom. And if such a scheme (and in particular the one in Table 4) is compared to the order of the deeds we find in Hyginius, it is apparent that the order differs in only one place, whereas it differs from Pseudo-Apollodorus’ account in as many as three places. We thus have plenty of reason to believe that the sequence of scenes in this relief was indeed based on some textual account, which is, however, closer to that of Hyginius and which, as far as I am aware, is not known to us today.38

The other artworks in the group have an even weaker connection to any of the textual or pictorial sources known to us today and the sequence of the episodes in them can, to the present-day beholder, accustomed to set sequences of events, sometimes seem entirely erratic. 39In light of what we have just seen, and considering the fact that there are many works of ancient art that lay out their narrative scenes in exactly the fashion that we would today perceive as logical and natural40 , we should not, however, too lightly dismiss the possibility that at least some of them did follow textual models that are simply not known to us.

In any case, one regularity certainly can be found in all these sequences, if we are but willing to accept that the slaying of the Nemeian lion was established already at that time as the first deed in the sequence and that the last one was either the capturing of Cerberus or the picking of the apples of Hesperides.41 All the objects place the Nemeian lion scene somewhere on the left side of the composition, and the last scene (whether it is Cerberus or the golden apples) on the right side. They thus conform to the order of reading that is natural in the Western culture, and resemble the vita icons in yet another important aspect. As we shall see in more detail in the following sections, many of the vita icons also do not consistently follow any literary text or visual model known to us and the order of their scenes has, therefore, also been perceived as surprisingly erratic and jumpy.42 They do, however, invariably place the logically first scene of the sequence (often that of the saint’s birth or baptism) on the left side of the composition and the last one (often that of the saint’s death or interment) on the right. With very few exceptions, the first scene is also placed where the present-day viewer would most likely expect it – in the upper left corner of the composition, and the last scene is placed in a correspondingly predictable place – the lower right corner. This regularity, which we find both in the Hercules cycles and in the vita icons, is all the more interesting if the sequences indeed do not otherwise follow any textual sources. In that case, we would have a somewhat peculiar, not to say paradoxical, situation of artists displaying complete lack of reserve in their playful arrangement of the middle parts of the story and, at the same time, a sort of conformism in the placement of the first and the final episodes. We might, of course, also suppose that these artists did not follow actual literary accounts, but worked rather from only very abbreviated summaries, based on unreliable, second-hand information, or that they only used some sort of model books, consisting perhaps of completely separate sheets of writing material, that did not always offer some indication of the ‘correct’ order of scenes. This would not, however, explain why or how the first and the last scenes in the series would still retain their respective places with such regularity.43

Whatever the case with the selection and ordering of the deeds may be, there is no doubt that the parallels we have just noted between these Hercules cycles and the vita icons are much more than mere similarities of form. We once again see that certain principles and conventions are so intrinsic to narrative art that the unexpected links which they provide between artworks have an ability of transcending genealogy. The icon painters in Byzantium cannot even have seen most of these Hercules cycles, let alone have studied their narrative techniques in any depth, and yet their works display not only the same formal arrangements, but repeat even their peculiar narrative inconsistencies. A strong link between these two classes of objects, then, does seem to exist, and we must be careful not to dismiss it only because it may require the kind of attention that departs at times from the domain of a historian and finds its home rather with a narratologist.

But let us turn our attention now from the narrative sequence to the central panel. We have seen in the case of the Illiac tablets that, when we think of this combination in compositional terms, it is in fact the monotonous series of narrative frames that calls for (something like) the central panel, and not the other way around. A very similar observation can be made if we take the same approach to the Heracles cycles. When the Dodekathlos first appeared in the metopes of a Doric temple, the nature of the medium dictated that it should consist of essentially isolated frames whose primary function was to fill the space between the triglyphs. The deeds of Hercules, as a choice of subject matter, may, of course, have had a more meaningful function, but the space for such a cycle was generated by nothing more than architectural conventions.44 And, needless to say, within those conventions, there was neither room nor need for an addition such as a prominent portrait of the protagonist. It was only when this cycle had migrated from the classical entablature into other media that some artistic freedom could be exercised in its formal arrangement. In cases where the nature of the medium did not demand that the deeds follow one another in a single band (such as we might find enveloping the body of a sarcophagus), arranging twelve deeds in a single visual field presented the same compositional problem that we have noted in the case of the Illiac tablets. To put the twelve scenes one next to another in, say, three rows of four scenes each, or some similar combination, meant to create an unimaginative, insipid design that did next to nothing to attract and maintain the viewer’s attention. A much livelier solution was to introduce a prominent portrait of the hero and arrange the deeds around it. On the other hand, the introduction of such a portrait was also a natural choice for any artist that did not only narrate a multi-character story such as we find, for instance, in the Illiac tablets, but was determinately enumerating and commending the deeds of a single hero. Encomium, whether verbal or pictorial, may be accomplished by a recounting of the individual’s deeds, but it is ultimately centred on the individual, not on the deeds themselves. This pictorial arrangement, then, is a very natural visual expression of that reality – just as verbal encomium, whatever it consists of, revolves around the person, the sequence of small panels, setting forth a theme and in this case depicting the hero’s deeds, revolves around the central portrait. It is, not, however, the portrait that attracts an accretion of the narrative panels, but the narrative panels, originating from contexts where they existed entirely on their own, conjure a portrait.

But what reason do we have to believe that this is anything more than speculation? What proof can be offered that this form of visual encomium indeed arises from such compositional and narratological reasons, rather than from models such as were sought by Ševčenko? The answer to this has already been hinted at when we called this a recognisable genre, and when we said that it appeared in cultures and periods so diverse that there can be no talk of dissemination from a common model. Comparable examples simply appear at too many different places and in too many periods of history for us to conclude that this image type comes into being in any other way but the one stated above.

The earliest, but by no means the closest, parallels can be found in Mithraic art, where the central tauroctone image is often coupled with series of smaller panels on the top and lateral sides (Figures 24 – 28).

Fig. 24  Mithraic fresco, Gardens of Palazzo Barberini in Rome, 3rd c. A.D.
As in: A. Grabar, The Beginnings of Christian Art, 200-395 (London, 1967), fig. 73.

Fig. 25   Mithraic shrine in Dura Europos, uncertain second- or third-century date.
As in: L. A. Campbell, Mithraic Iconography and Ideology (Leiden, 1968), pl. 1.

Fig. 26   Mithraic relief from Neuenheim, Germany, uncertain date.
Badisches Landesmuseum, Karlsruhe.
As in: L. A. Campbell, Mithraic Iconography and Ideology (Leiden, 1968), pl. 25

Fig. 27   Mithraic relief from Neuenheim, Germany, uncertain date.
Badisches Landesmuseum, Karlsruhe.
As in: L. A. Campbell, Mithraic Iconography and Ideology (Leiden, 1968), pl. 26

Fig. 28   Votive plaque of Mithra Taurokton, from the village of Kurtovo Konare, Bulgaria, 3rd century A.D.

Much closer comparisons, in terms of both overall form and nature of the sequential imagery, can be found in Early Christian art, owing much to the fact that Christianity, unlike Mithraism, is an inherently narrative religion. The Early Christian examples appear primarily as ivory book covers that were erroneously identified as ‘five-part diptychs’ in much of the literature of older date.45 Four complete, or nearly complete, sets of these book covers are known to us (Figures 29-32), and there are fragments of at least five more. Both the front and the back leafs always contain a central iconic image and narrative scenes surrounding them on at least three, and sometimes all four, sides. The group seems to date from the mid-sixth century and originate from a single, as yet uncertain, source.

Fig. 29.1 (front)

Fig. 29.2 (back)
Figures. 29.1-2   Ivory covers reused on the Ejmiadzin (Etchmiadzin) Gospels, 989 A.D.
Matenadaran Library at Erevan

Fig.  30.1 (front)

Fig. 30.1 (back)
Figs. 30.1-2   Ivory book covers reused on the St. Lupicin Gospels (the front and back have been exchanged), 9th c.
Paris, Bibliothèque nationale de France, MS lat. 9384.

Fig. 31.1 (front)

Fig. 31.2 (back)
Figs. 31.1-2 Reconstruction of ivory book covers from Murano, Ravenna, 8th c.
Ravenna, Museo Nazionale; Manchester, John Rylands University Library; Paris, Musée du Louvre; St. Petersburg, State Ermitage Museum; Berlin, Museum für spätantike und byzantinische Kunst.

Fig. 32.1 (front)

Fig.  32.2 (back)
Figs. 32.1-2 Ivories from Milan, 6th c.
Cathedral treasury in Milan, Italy

Given that the great majority of the narrative scenes present stories from the life and ministry of Jesus Christ, who is invariably represented in iconic mode in the middle of the ensemble, it goes almost without saying that they could be perceived as a Christian variation on the theme of narrative encomium offered to an exceptional individual.46 The essence and character of Christ’s deeds are, needless to say, vastly different from those of a classical hero, as are their import and message for the beholder, the Dodekaorton (the twelve great feasts of the liturgical calendar) is substituted for the Dodekathlos, but the main constituents of our genre remain unaltered. Scenes from an illustrious life elucidate the importance of an individual and, to make the design more compelling and give the figure more eminence, an iconic image is introduced in the centre. In the Christian context, this combination serves to affirm the meaning of the Gospel texts, and offer at once an icon fit for receiving veneration and a narrative conspectus of the economy of salvation.

It is by no means unimportant for the future of the genre that these ivories functioned as book covers. The form will indeed continue to make appearance in book art not only throughout the Middle Ages, but also deep into the modern period.

We see it in one of the two surviving illuminations from a Latin manuscript (Lib. MS. 286) now kept in the Parker Library of Corpus Christi College, Cambridge (Fig. 33).47

Fig. 33  Illumination from the St. Augustine Gospels, 6th c.
Cambridge, Corpus Christi Ms. 286, f.129v.

Judging from palaeographical evidence, this manuscript seems to have been produced some time in the sixth century48, and it thus presents at least a near contemporary of our ivory book covers. The style of its miniatures, admittedly, has very little in common with that of the ivories, and the narrative scenes only flank the central portrait on the lateral sides, but the functioning of the ensemble remains identical - with one peculiar variation. The central portrait is not one of Christ, whose works are narrated in the lateral compartments, but of the Evangelist who put them in writing. This does not detract from our basic contention, if we remember that icons are not portraits in the modern sense of the word. Even when representing the holiest of subjects, such as Christ and the Mother of God, they are not meant to portray the individual personages, with accurately recorded outward appearances and distinctive human characteristics, but mediators of God’s grace and agents of his work. In the case of saints, furthermore, the portraits do not glorify their own human excellence. Since their illustriousness is defined only by the degree of their participation in God's grace and work, whatever honour is paid to them through an iconic portrait is channelled through to God. A prominent portrait of an Evangelist, thus, does not compete for glory with miniature narrative depictions of Christ but acts only as a witness to the truthfulness of the account and as a pointer to the one true source of glory. We thus have what to the best of my knowledge is a unique situation in the history of art - that the person who is portrayed is not the one whose works are narrated in the surrounding scenes and who is, in the end, praised by the artwork as a whole. At the same time, however, the person portrayed does have a share in the overall illustriousness of the narrative, not through his own deeds but through the bearing of reliable witness to the deeds of another.

We next find our arrangement in Giacomo Grimaldi’s drawings and descriptions of a destroyed early eight-century mosaic cycle in the Oratory of Pope John VII (Fig. 34).49

Fig. 34  G. Grimaldi’s drawing of a destroyed eighth-century mosaic in the Oratory of Pope John VII, old St. Peter’s basilica, Rome.
Vatican City, Biblioteca Vaticana, Ms. Barb. XXXIV.

The cycle, Christological in theme and consisting of sixteen scenes divided into seven frames, was arranged around a central figure of the Virgin. I know of no comparable monumental arrangement in the Byzantine world, but the central image at least does suggest a link with the Christian East. Given that the Virgin is represented as wearing the dress and the crown of a Byzantine empress50, we can be certain that the artists were at least familiar with Byzantine iconography, if not in fact imported from the East.51 What is most important for us here, though, is that this appearance of our arrangement in one of the most important churches of Western Christendom presents not only an early but also a potentially very influential precedent, which was completely ignored by all my predecessors.52

Our next example comes from another ivory book cover (Fig. 35), produced this time by the court school of Charlemagne at Aachen, around the year 800 A.D.53  The narrative sequence, consisting of twelve images, here surrounds a portrait of a young, beardless Christ in triumph. The dependence on the Late Antique ivories examined above is obvious not only in the overall form, but six of the scenes are in fact direct copies from that model. 54

Fig. 35   Ivory cover, early ninth century, inset in the eighteenth-century leather cover of a Latin Gospel Lectionary produced around 800 A.D.
Bodleian Library, Oxford MS. Douce 176.

The silver-gilt reliefs decorating the covers of a fourteenth-century lectionary that we have seen in Figures 9.1 and 9.2 show us that the form found its way into Byzantine book art as well. Inserted between the narrative scenes from the lives of Christ and the Theotokos are enamelled medallions with portraits of prophets and saints, as well as one depicting the Hetoimasia, but the overall scheme is again essentially unaltered.

An engraving of Pope Sixtus V, published in 1589, ensured that his deeds would be remembered by surrounding his portrait with a series of labelled images, depicting primarily his extensive building projects in Rome (Fig. 36).

Fig. 36   An engraving of Pope Sixtus V
Published in Ioanne Pinadello Taurisino: Inuicti quinarii numeri series. Quae summatim a superioribus pontificibus et maxime a Sixto quinto res praeclare quadrienno gesta adnumerat ad eundem Sixtum quantum pont. Opt. max. (Rome, 1589)
As in: The Vatican: Spirit and Art of Christian Rome (New York, 1982), p. 21

After the Renaissance, this design will live on in book art largely dissociated from its original uomini illustri context and appearing particularly often in the frontispieces. To pick a few examples out of many, we may mention, for instance, that of Robert Burton’s The Anatomy of Melancholy (Fig. 37), where, instead by some illustrious deeds, the side compartments are taken up by the various manifestations of the mental disease the book is concerned with. Insomuch as the author has conducted extensive research into these matters, and has provided us with a book, these can, in a certain way, still be understood as his deeds, but the design itself has by now become well-established and emancipated from its original context.55

Fig. 37  Frontispiece to R. Burton, The Anatomy of Melancholy 6th edn. (1651, Oxford)

Nearly contemporary with this treatise on melancholy are Richard Braithwaite’s two courtesy books, The English Gentleman and The English Gentlewoman, first published in 1630 and 1631, respectively. Their frontispieces (Figs. 38.1-2) also make use of this design, surrounding the title characters with pictorial representations of the qualities the books hope to instill. Those pertaining to the illustrious gentleman include not only ‘Education’ and ‘Moderation’, but even ‘Perfection’ itself, and interestingly never coincide with those proper to a gentlewoman. She becomes noteworthy through the development of ‘Decency’ and ‘Gentility,’ but ‘Apparel’ is not to be neglected either (‘Comely not gaudy’).

Fig. 38.1   Frontispiece to R. Braithwaite, The English Gentleman (London, 1631)

Fig. 38.2 Frontispiece to R. Braithwaite, The English Gentlewoman (London, 1631)

While these examples still retained at least some connection to the image type’s origins, the basic scheme remained in use even when both the centre and the periphery were deprived of a prominent, illustrious individual. If we examine the cover of a weekly instalment of Charles Dickens’ A Tale of Two Cities, published in 1859 (Fig. 39), we find that both the portrait and the deeds are missing from the compound. It has been reduced to its barest form and retained only for its visual interest. It is true that it had, by now, become quite familiar, almost expected, to both the book printers and their audience, but the primary reason why it is still retained is its visual effectiveness. It is simply more attractive to the eye than would be a title placed above or below a simple, uninterrupted sequence of scenes. The design that originated from purely compositional reasons is also retained for those reasons and at no stage of this development does it need, though it certainly can have, direct formal precedents.

Fig. 39   Cover of a weekly instalment of C. Dickens, A Tale of Two Cities, vol. v (London, 1859)

It will be used for precisely those reasons in myriad different examples, from both high and popular culture, as illustrated by our Figures 40-43. In Fig. 40, for instance, we see how the members of the Rothschild family made creative use of this format to promote not only their entrepreneurial deeds and industrial prowess, but also benefaction and altruism. Although the border still offers us the various notable accomplishments, presented here as the stations on the Kaiser Ferdinand’s Norbahn, or stages in the railway’s construction, the centre of the composition is no longer occupied by a single, prominent figure. Instead of openly presenting the person who is to be admired for these deeds, the painting shows us the act of admiration itself. The opening ceremony, dated to 7 July 1839, is presented as an occasion of wide public interest and delight and the foreground is taken not by any member of the Rothschild family, nor even by the railroad itself, but by the mass of the people attracted to this joyful event. The people are the main beneficiaries of this accomplishment and the deeds celebrated are to be understood as those of public spirit and philanthropy.

Fig. 40   Opening ceremony of the Kaiser Ferdinand’s Norbahn, 7 July 1839
As in: C. Corti, The Origin of the House of Rothchild (London, 1928), pl. 9, opposite p. 111.

In Figure 41, on the other hand, the central portrait has been retained, but the surrounding narrative of heroism abbreviated and simplified. Instead of the full sequence of individual deeds or exploits, the border now only has a carefully composed single image. Through a combination of painting and photography, and with clever positioning of the figural masses, the artist has created an image that, when divided into several fields, creates an effect of offering a film- or a strip-like narrative, although in reality there is nothing truly narrative, let alone sequential, about it. The exploits of the couple portrayed in the central panel are perseverance in a time of trial, and tenderness in an age of hostility.

Fig. 41   Poster for the silent film Wings, directed by William A. Wellman and released by Paramount Pictures in 1927

In Figures 42.1 and 42.2 we find two examples of narrating the proceedings of the papal conclave by dividing the process into scenes organised around a central panel.56 There is no glory here to be merited by the exploits, at least not already in this world and, in any case, they are not individual deeds of the future pope, and so the Figure 42.1 shows us the exploits alone. As their results have the potential of influencing the lives of many, they are interesting enough to be delineated and narrated in detail, even without a distinct protagonist that we could identify with. Were it not for the inscription, we might suppose that the engraving was probably executed prior to the new pope’s election, and he was thus a mystery character in this picture, whose absence and as yet unknown identity added to the interest of the artwork. The inscription, however, makes it perfectly clear that the artwork was completed after the election and we have to conclude that the new pope’s portrait was left out purposefully. The engraving in Figure 42.2, though, conforms to the traditions of the genre and does place the new pope’s portrait in the middle, although, as we said, the exploits surrounding him are not his own.

Fig. 42.1   Engraving depicting the proceedings of the conclave to elect Pope Clement X, 1670

Fig. 42.2   Engraving depicting the proceedings of the conclave to elect Pope Alexander VIII, 1691

Finally, just to remind ourselves that the design is by no means unique to the Western culture, we may turn to the example offered by our Figure 43. This incised stone tablet from China, dating from 1989, organises the scenes from the Life of Confucius in exactly the same way that we have observed in its’ Western counterparts.57

Fig. 43   Scenes from the Life of Confucius, incised on stone tablet, 1989.
Qufu, Shandong.
As in: Julia K. Murray, ‘What is “Chinese Narrative Illustration”?’ AB, vol. 80. no. 4 (Dec 1998), p. 607, fig. 10

As we can see, then, the situation is far from that which Ševčenko would seem to have imagined when she wrote that our genre ‘made a startling appearance in 1988 in the Boston Globe.’58 Not only are there much more memorable and influential examples, but they exist in such numbers that providing a full catalogue of them would take nothing less than a dedicated encyclopaedia. And, although some of the examples offered here may, at first glance, seem to be superfluous and unrelated to the study’s main topic, the vita icon, I believe it is vitally important for our understanding of this genre to realise just how ubiquitous it really is. The idea that it is limited to less than two dozen Byzantine examples and the notion that there is something startling about this Boston Globe illustration seem to me to be but two symptoms of the same problem. Both have their origin not only in too narrow a focus, which fails to afford the wider context, but also, and much more importantly, in a defective interpretative model. If we conceive of the vita icon as a genre of icon painting, we necessarily start from the supposition that, in its genesis, the iconic portrait came first, and then it somehow acquired an accretion of smaller, narrative images. We have seen, however, in our examination of the long-term developments in this genre, that precisely the opposite was the case. In the beginning was the narrative, and this narrative has, primarily for compositional and narratological reasons, attracted a prominent central panel which could, but did not necessarily have to, contain the main protagonist’s portrait.59 This basic construction has been employed for a vast range of purposes and has seen many interesting alterations, in terms of both form and content. And the one thing that has remained constant in all of its manifestations that we have observed was that it could always be best understood as a genre of iconature.

As was stated at the beginning, the point of this survey was not to locate possible origins for the Byzantine vita icons, in order to answer how they came into being, but to investigate the universality of the form, in order to understand what they are. The next logical step, now that we do know what they are (namely, a genre of iconature) is, of course, to see what consequences that might have for our study of them. These issues will be explored in the next post on the topic, but let us first cast a brief look over our shoulders and offer an interpretation of what has been observed thus far.

It is, first of all, worth reminding ourselves that Christian art, as well as Christian religion, is inherently narrative. Even images that contain no narrative element per se necessarily point to the events from the economy of salvation and can thus be considered narrative by association. Icons, furthermore, even when they were thought to contain a real, tangible presence of the saintly individual, are not portrayals in the modern sense of the word. As was said earlier, they do not portray individuals, but agents of salvation, whose illustriousness proceeds from their participation in God's grace. And this participation’s most natural artistic expression is precisely narration. In a sense, then, there is nothing innovative about narrative icons, whether they show us a saint’s vita, or belong to any other type. From its very beginning, Christian art has always been, and necessarily always will be, narrative. Vita icons are, therefore, not a hybrid form, as Belting would have it, nor are they essentially a late-twelfth century invention, translated from metal into paint, as Ševčenko has claimed. They are simply a new variant on the universal theme of narrating and praising the deeds of exceptional individuals, very much in resonance with the historical nature of the Christian religion and art. If this variant had only appeared in the late-twelfth century Christian East, we could almost wonder why it had not appear any sooner.60

Additionally, it is worth noting that an outside observer (anyone who is not a Byzantinist, really) might wonder why we are even discussing this at all. Why all this deliberation over the origins of some image type? Cannot an artist simply have invented a design to suit a particular function, independent of any precedents, whether they be direct models, as Ševčenko would prefer, or something more evasive, as I would argue? I do not believe that it would be a very plausible or even a fair answer to this to simply claim that artistic freedom was limited in Byzantium and that innovation only occurred under special circumstances. While the Byzantine culture and society certainly did have a distinctive tendency towards standardisation and canonisation, they were by no means altogether deprived of diversity, freedom, or innovation. New image types did appear, new artistic styles were developed, and that certainly did not happen without innovation, nor did it happen through some peculiar type of it, known only to Byzantinists. A much more honest answer would be that, in this time of increasing specialisation, all academics are necessarily at least suspect, if not altogether guilty, of some type of provincialism. While the modern-age type of academic compartmentalisation has certainly freed us from many errors rooted in the generalism of previous ages, it has not been without faults of its own. And one of its greatest shortcomings has certainly been its recurrent blindness to that which is universally human. Let one example stand for many. There has been considerable debate over the origins of the various practices involving altered states of consciousness in ancient Greece.61 Several convoluted theories of how the Scythian variety of this phenomenon was disseminated were proposed, and many fine scholars proved themselves to be utterly blind to the nature of the puzzle itself. And, as surprising as this may appear in retrospect, many lengthy studies had to be written before scholars could finally grasp the very obvious fact that an existence of the same phenomenon in different cultures does not necessarily mean that the later institution draws on the earlier one, or that both derive from a common historical source. (Or that this type of reasoning is actually a logical fallacy commonly referred to as post hoc ergo propter hoc.) What might very well have been obvious at first glance to an outside observer, took the academic world quite some time to remember – that very similar things can, and very often do, develop independently in several cultural areas. And the situation we have with Ševčenko's puzzle is, while not exactly identical, in many ways similar. In both cases, that is, we are looking at a manifestation of the 'factual and logical inaccuracy of the diffusionist approach.' In both cases we see how the academic world is often so focused on the particular and the local that it is rendered blind to the universally human. For scholars such as E. Rhode, M. L. West, and K. Dowden, there was no question whether a link with the Eurasian steppe, Anatolia, or perhaps India, had to exist in order to explain the Greek variety of shamanism-like practices. The only question was how best to reconstruct it. For Ševčenko and Chatterjee, there was no question whether the vita icons needed some direct model from which to derive their form, the only puzzle was what that model might have been. At several places in her work, Ševčenko expressly stated something like: ‘this cannot have been the immediate model that we are seeking’62, but at no point did she explain why this immediate model would be necessary at all. Focusing, as we now all do, on a narrow strip of land in a short stretch of time, she made her object of study not the creativity of the human being, but the conventionality of the Byzantines. And just as so many fine classicists were blind to the fact that shamanic-like practices, whatever their particular origin may be, relied on the common neurological characteristics of the human mind, Ševčenko was insensitive to the fact that visual narration relied on equally universal laws of visual perception. And while I am not certain that the cognitive or neuroscientific approaches advocated by Ustinova would also be the most productive way of studying narration in Byzantine icons, I do think that certain indispensable lessons can be drawn from such studies. Above all, I think that we historians should remind themselves, at least from time to time, that ‘[t]he analysis of human thought cannot be based exclusively on a research into the culture which gave rise to it’ (Ustinova, op. cit.). And if that does not indeed lead us to such significant paradigm shifts as have been proposed by the recent work of classicists such as Walter Burkert63 or Roger Beck64, it will still do us a good service by reminding us of what our object of study was supposed to be. And if that, in turn, does not lead us to any new discoveries, it should at least lead us out of impasses such as the one that we have been exploring here.

II. On the birth of the so-called comics

Most histories of what is variously termed the ninth art, comics or bande dessinée start with what to my mind is a much abbreviated, not to say bowdlerised, so-called prehistory of the medium. It is always obvious that the author has not studied these proto-examples in any depth, or breadth, and that (s)he has uncritically accepted the notion that something entirely novel has come into being with the work of either Rodolphe Töpffer or Richard Outcault. The column of Trajan, the Bayeux tapestry, and a handful of other commonplace examples of visual narration are always mentioned only summarily, by way of preface, before the discussion deftly moves to the nineteenth century. I will challenge this received narrative in a series of posts, proposing that, instead of writing a history of comics/bande dessinée, we should rather write one of iconature. But I want to make it plain from the very beginning that I see the same problem with starting the comics narrative in 1827 (with L'Historie de Monsieur Vieux Bois) or 1896 (with The Yellow Kid and His New Phonograph) as with starting the vita icon narrative in late-twelfth-to-early-thirteenth century. I took the time to write such a long and detailed post on the vita icon conundrum because it is a comparison I plan to milk for all it is worth.  

1 The artworks in question have been variously termed ʻhagiographical’, ʻhistoriated’, ʻvita’, ʻbiographical’, and ʻnarrative’ icons. See T. Mark-Weiner, ʻNarrative cycles of the life of Saint George in Byzantine Art’, vols. I and II (unpublished doctoral dissertation, New York University, 1977); K. Weitzmann, Studies in the Arts at Sinai (Princeton, 1982), passim, but esp. ch. VIII, pp. 211-46; N.P. Ševčenko, The Life of St. Nicholas in Byzantine Art (Turin, 1983); eadem, ʻVita Icons and “Decorated” Icons of the Komnenian Period’ in B. Davezac (ed.), Four Icons in the Menil Collection (Houston, 1992), pp. 57-69; eadem, ‘The Vita Icon and the Painter as Hagiographer’, DOP 53 (1999), pp. 149-65; L. Brubaker, ʻThe Vita Icon of Saint Basil: Iconography’, in B. Davezac (ed.) Four Icons in the Menil Collection (Houston, 1992), pp. 75-93; A.W. Carr, ʻThe Vita Icon of Saint Basil: Notes on a Byzantine Object’ in B. Davezac (ed.) Four Icons in the Menil Collection (Houston, 1992) pp. 95-105; H. Belting, Likeness and Presence (Chicago, 1994), pp. 249-58; H. Maguire, The Icons of Their Bodies: Saints and Their Images in Byzantium (Princeton, 1996), passim, but esp. pp. 173, 185, and 188; P. Chatterjee, ʻNarrating Sanctity: The Narrative Icon in Byzantium and Italy’ (unpublished doctoral dissertation, University of Chicago, 2007); E.C. Constantinides, Images from the Byzantine Periphery – Studies in Iconography and Style (Leiden, 2007), pp. 150-79.

2 N. P. Ševčenko, ʻVita Icons and “Decorated” Icons of the Komnenian Period’ (as in n. 1), p. 57

3 As Efthalia Constantinides explaines, only the first twelve strophes of this hymn are narrative in content and deal with the lives of Christ and the Virgin from the Annunciation to the Presentation. The other twelve are theological and liturgical, with strong allegorical overtones. See E. C. Constantinides: ‘The Question of the Date and Origin of the Earliest Akathistos Cycles in Byzantine Monumental Painting in the Light of the Akathistos of the Olympiotissa at Elasson’ in eadem: Images from the Byzantine Periphery – Studies in Iconography and Style (Leiden, 2007), pp. 42-3. Constantinides is somewhat imprecise in her terminology, though. She uses the term oikoi for all the strophes, although the first twelve are normally referred to as konatkia. Cf. V.D. Lihačeva, ‘The Illumination of the Greek Manuscript of the Akathistos Hymn (Moscow, State Historical Museum, Synodal Gr. 429)’ DOP 26 (1972), p. 256.

4 It should be noted here that Ševčenko was very much aware of the fact that ʻ[T]he earliest extant icon with a hagiographical cycle [was] probably an 11th-century icon of St. Nicholas on Mt. Sinai (now in two parts),’ as well as that she accepted K. Weitzmann’s reconstruction of this icon as originally a triptych. See N. P. Ševčenko, ʻThe Vita Icon’ (as in n. 1 above), n. 2, p. 150; K. Weitzmann, ʻFragments of an Early St. Nicholas Triptych on Mount Sinai,’ Δελτ.Χριστ.Αρχ.Ετ., ser. 4, 4 (1964), pp. 1-23. The icon in question is illustrated here as Fig. 7.1-2.

5 N.P. Ševčenko, ʻThe Vita Icon’ (as in n. 1), p. 154. It is important to note that the author does not offer any explanation of this claim.
6 N.P. Ševčenko, ʻThe Vita Icon’ (as in n. 1), p. 151

7 N. P. Ševčenko, The Life of St. Nicholas in Byzantine Art (Turin, 1983), pp. 163-4; eadem, ʻIcons and “Decorated” Icons of the Komnenian Period’ (as in n. 1), p. 57.

8 It should, however, be pointed out that we do know of artworks that combine the portraits and the narrative scenes in the borders surrounding the central iconic image. See, for instance, the covers of the fourteenth-century Byzantine lectionary now kept in Venice (Figures 9.1-2). In the vita icons we find it in the oldest preserved Russian example – the Pskov school icon of prophet Elijah, dating from no later than the end of thirteenth century and now kept in the Tretyakov gallery in Moscow. Cf. I. Rodnikova, ‘Pskov Icons of the 13th to 16th Centuries’ in A. Rodina (ed.), The Russian Icon (St Petersburg, 2006), no. 104, p. 55.

9 It is perhaps worth reminding ourselves here that we know of no written accounts that would suggest that this genre was seen as a distinct visual category in Byzantium.

10 In a later article, she revised this claim by adding that the form was ‘expressly designed to be understood by the diverse groups that constituted [the East Mediterranean] society.’ It would, thus, serve as an international language of images, rather than as a gift to an icon. See N.P. Ševčenko, ʻThe Vita Icon’ (as in n. 1), p. 161.

11 See A. W. Carr, ʻThe Vita Icon of Saint Basil’ (as in n. 1), p. 97 and n. 9, p. 103. The volume in question is B. Davezac (ed.) Four Icons in the Menil Collection (Houston, 1992)

12 P. Chatterjee, Narrating Sanctity (as in n. 1), p. 75. Ševčenko has, in all fairness, pointed to this weak point of her argument herself – a fact that Chatterjee must have been aware of but failed to acknowledge. See N. P. Ševčenko, ʻThe Vita Icon’ (as in n. 1), p. 150. Chatterjee otherwise focused her efforts on proposing a wider cultural context for the icon type’s emergence and has interpreted the vita icons in relation to the allegedly concurrent appearances of the sanctuary screen and the empsychos eikon.  

13 H. Belting, Likeness and Presence (Chicago, 1994), p. 257. It must be noted here that this claim that the narrative element normally has no place in icons arises from the main premise of Belting’s book, namely that Christian images before the age of Renaissance were not treated as ‘art’ but as objects of veneration which possessed the tangible presence of the Holy. It should not, therefore, be taken as a statement that the author would expect us to consider a product of careful consideration of the interplay between the iconic and the narrative elements in Christian art. In fact, it can hardly be supported in the light of the many examples that demonstrate precisely the opposite properties of Christian imagery. One needs only to think of any icon that contains anything but or besides a portrait (see, for instance, Figures 4 and 5), or of the numerous crosses where the arms of the cross bear images narrating the life of Christ, not to mention the iconostasis itself, especially with its representations of the twelve great feasts of the Orthodox Church, to see to what extent Christian art is indeed inherently narrative. On the issue of possibility of offering supplication to figures in action, see an interesting, though at times somewhat speculative, article by M. J. Anderson, Narrative Icons of the Nativity: The Representational Problematic of the Nativity Feast and Two Solutions, published on 21 Dec 2010 on

14 Interestingly enough, very similar thought process on a very similar image type can be found in B. Schweitzer, ʻDea Nemesis Regina,’ JdI 46 (1931), pp. 175-246.

15 In fact, in 1983, she expressly stated that ʻ[t]he format of these icons is not an innovation,’ and that it was ʻpossible that the form was rediscovered by the antiquarian artists of the 10th century Macedonian Renaissance, or may even have been reimported from the West.’ See N. P. Ševčenko, The Life of St. Nicholas in Byzantine Art (Turin, 1983), pp. 163-4. It was only in 1992 that she disregarded the eleventh-century date of the earliest hagiographical icon of St. Nicholas, advanced the claim that the genre appeared only in the late twelfth century, and that its formal origin was in the so-called κεκοσμημέναι είκόνες. See N. P. Ševčenko ʻVita Icons and “Decorated” Icons of the Komnenian Period’ (as in n. 1), pp. 57-69.

16 On these fascinating Late Antique objects, see: K. Weitzmann, ʻA Tabula Odysseaca,’ AJA, Vol. 45, No. 2 (April-June, 1941), pp. 164-81; A. Sadurska, Les tables illiaques (Warsaw, 1964); R. Brilliant, Visual Narratives (as in n. 1), pp. 53-9; and, most recently, M. Squire, The Illiad in a Nutshell: Visualising Epic on the Tabulae Iliacae (Oxford, 2011).

17 I am using this term in the sense defined by Thierry Groensteen’s seminal work on the composition of the narrative page in the medium of comics. See T. Groensteen, Systéme de la bande dessinée (Paris, 1999), passim, but esp. Ch. I, pp. 31-119.

18 We find such a formal arrangement, for instance, in our Figure 5, as well as in innumerable examples of manuscript illumination, from which it undoubtedly derives.

19 Other comparable devices include: careful positioning of the key figures so that they provide strong visual links between registers, as on the column of Trajan; variation in the colour of the background, as in the Ashburnham Pentateuch (Paris, Bibliothèque Nationale de France, MS nouv. acq. lat. 2334); and playful layout of figural masses, as in the Joshua Roll (Vatican Cod. Palat. gr. 431).

20 This mosaic, now kept in the Musee National du Bardo, Tunis, is a particularly good example of my thesis. Its centre is occupied by nothing but a luxurious mansion, which is, in turn, separated from the rest of the composition only by virtue of its own rectangularity. Were the house absent or placed anywhere else in the composition, we would read this as a rather dull sequence of country life scenes, divided into three superimposed and undifferentiated registers. With this simple addition, it fulfills its role of conveying, as Katherin M. D. Dubabin explained, ‘a picture of the great country estate, watched over by its master and mistress, while season after season its lavish produce pours in. Thereby it communicates the ideal self-image of the aristocratic landowner.’ See K.M.D. Dunbabin, Mosaics of the Greek and Roman World (Cambridge, 1999), pp. 118-9 and fig. 122, p. 120.

21 See K. Weitzmann (ed.), Age of Spirituality: Late Antique and Early Christian Art, Third to Seventh Century (New York, 1978), n. 136, p. 159-60. A very comparable example, regardless of the fact that it is in round, rather than rectangular, form is to be found in the Syrian textile fragment with Joseph story, dating from the fifth or sixth century A. D. See K. Weitzmann (ed.), Age of Spirituality, cat. no. 413, pp. 461-2.

22 The painting was found in Cave 17 in Dunhuang and the lateral images depict one of the many jataka stories, recounting events from Buddha’s past lives. Cf. R. Whitfield and A. Farrer, Caves of the Thousand Buddhas – Chinese Art from the Silk Route (London, 1990), pp. 23-6, fig. 2.

23 R. Arnheim, Entropy and Art: An Essay on Order and Disorder (Berkely, 1971), passim but esp. pp. 48-51.

24 It, in fact, serves the same purpose as do the various devices for dividing up the narrative into distinct boxes or scenes.

25 N. P. Ševčenko, The Life of St. Nicholas in Byzantine Art (Turin, 1983), pp. 163.

26 Other characters whose lives and works have been honoured with picture cycles include Theseus, Achilles, Aeneas, Romulus, and Mithras. For Theseus, see: F. Brommer, Heracles: The Twelve Labors of the Hero in Ancient Art and Literature (New York, 1986), p. 58. It is interesting to note that several deeds of this hero remarkably resemble those of Heracles, and that they are frequently found together. For instance, Hyginius (Fabulae 30) tells us that Theseus killed the Cretan bull at Marathon, which Hercules had brought to Eurystheus. Brommer (loc. cit.) has, therefore, conjectured that they may have been introduced at the same time for political reasons. For Achilles, see: G.A.S. Snyder, ʻThe So-called Puteal in the Capitoline Museum at Rome’, Journal of Roman Studies 13 (London, 1923), pp. 56-68; K. Weitzmann: ʻObservations on the Cotton Genesis Fragments,’ in K. Weitzmann (ed.), Late classical and mediaeval studies in honor of Albert Mathias Friend, Jr. (Princeton, 1955), pp. 55-6; L. Guerrini, ʻInfanzia di Achille e sua educazione presso Chirone,’ Studi Miscellanei 1 (Rome, 1958-1959), pp. 43-53; J.W. Salomonson, ʻLate Roman Earthenware with Relief Decoration Found in Nothern Africa and Egypt,’ Oudheidkundige Mededelingen 43 (X, 1962), p. 80; D.E. Strong, Greek and Roman Gold and Silver Plate (New York, 1966), p. 197; K. Weitzmann (ed.), Age of Spirituality (as in. n. 21), no. 207, pp. 230-31, and no. 208, pp. 231-4. For Aeneas, see: M. Borda, ʻIl fregio pittorico delle origini di Roma,’ Capitolim xxxiv no. 5 (Milan, 1959), pp. 3-7; C. Dulière, Lupa Romana: recherches d'iconographie et essai d'interprétation (Brussels, 1979), vol. 1, pp. 93-4; J. Evans, The Art of Persuasion: Political Propaganda from Aeneas to Brutus (Ann Arbor, 1992), pp. 35-57; K. Weitzmann (ed.), Age of Spirituality (as in n. 21 ), fig. 25, p. 201. For Romulus, see: A. Bartoli, ʻIl fregio figurato della Basilica Emilia,’ in BdA 25 (Rome, 1950), pp. 289-94; G. Carettoni, ʻIl fregio figurato della Basilica Emilia,’ in RivistArch n.s. 10 (Rome, 1960), pp. 5-78; H. Furuhagen, ʻSome Remarks on the Sculptured Frieze of the Basilica Aemilia in Rome,’ in OpRom (Rome, 1961), pp. 139-55; E. Simon in W. Helbig, Führer 4th edn. II (Tübingen, 1966), pp. 834-43, no. 2062; F.C. Albertson, ʻThe Basilica Aemilia Frieze: Religion and Politics in Late Republican Rome,’ Latomus T. 49, Fasc. 4 (Brussels, 1990), pp. 801-815. For Mithras, see: E. Will, Le relief cultuel greco-romain (Paris, 1955), fig. 390; L.A. Campbell, Mithraic Iconography and Ideology (Leiden, 1968), passim, but esp. pls. 25, 26, and 45.

27 F. Brommer: Heracles (as in n. 26), p. 55. Brommer also draws our attention to the fact that this Dodekathlos is well attested in the literary sources. It was definitely known already to Theocritus and Apollonius in the third century B.C, while later authors such as Diodorus Siclus (Bibliotheca historica, iv. 1-27; I have used the I. Bekker, L. Dindorf, F. Vogel, and I. Bekker’s translation, available at, Apollodorus (Bibliotheke, ii. 74-126; I have used Sir J.G. Frazer’s translation available at and Hyginus (Fab. 30) report on the stories of the Dodekathlos in minute detail. See F. Brommer: Heracles (as in n. 21), p. 64. However, the author seems to be a bit undecided as to how to interpret this fact. He underlines that ʻthe word “Dodekathlos” did not exist at all in antiquity’ and that ʻsuch words turn up for the first time in the Hellenistic period, but they are rare and do not refer to a cycle of twelve combats but to the victor of twelve fights.’ He concludes, contradicting his previous claim, that ‘it seems likely that the concept also was unknown.’ - loc. cit., p. 63.

28 I have decided to leave out of the main discussion the very interesting case of the funerary monument at Igel, Germany, on the grounds that it does not contain narrative panels. It should, however, be noted that the figures surrounding the relief of the apotheosis of Hercules differ from those surrounding the other three main reliefs of this monument. They are not putti in random poses, but are quite clearly Giants borrowed from a gigantomachy scene. In a somewhat wider sense then, and if we remember the prominent role Hercules played in that mythical battle, these figures represent a contracted form of the hero's greatest deed. This relief could, with such reasoning, be comparable especially to the mosaic found in ancient Cartima (Fig. 21), where the labours have likewise been abbreviated into simple depictions of Hercules’ subdued opponents – for instance, instead of the hero fighting the Nemean lion, we see the lion lying defeated under a tree, and instead of the cleaning of the Augean stables, we see the god of river Alpheus sitting pensively on a rock. On the relief in Igel see: H. Dragendorff and E. Krüger, Das Grabmal von Igel (Trier, 1924), esp. pp. 58-60 and 70-2, fig. 42, and pl. 8; F. Cumont, Recehreches sur le symbolisme funéraire des Romains (Paris, 1942), pp. 174-5, pl. XIV 1. On the mosaic from Cartima: E. Hübner, Die antiken Bildwerken in Madrid (Berlin, 1862) no. 827, p. 310; E. Hübner, ʻMosaico di Cartama,’ Annali dell'Instituto Internazionale di Corrispondenza Archeologica XXXIV (1862), pp. 288-90; A. Ballil Illana, ʻMosaico con representación de los Trabajos de Hércules hallado en Cartama,’ Boletin del Seminario de Estudios de Arte y Arqueologia XLIII (Valladolid, 1977), pp. 371-9.

29 The object has been dated to the second century A.D. and is commonly referred to as the votive relief of Cassia Priscilla. The inscriptions clearly indicate the two main characters, Omphale and Heracles, and inform us that the relief was dedicated to the memory of Cassia Priscilla's mother (Cassia/Mani filia/Priscilla/fecit). It is presently kept in the Museo Nazionale in Naples. To the best of my knowledge, a complete analysis of it has not yet been offered. Ševčenko points to a good reproduction in A. Grabar, The Beginnings of Christian Art, 200-395 (London, 1967), fig. 72, p. 78, but the text of Grabar's book doesn't make any direct mention of this relief at all. It had previously been reproduced and briefly discussed in B. Schweitzer, ʻDea Nemesis Regina,’ JdI 46 (1931), p. 235, fig. 21. A short analysis of it can be found in N. B. Kampen, ʻOmphale and the Instability of Gender’ in eadem and B.A. Bergmann (eds.), Sexuality in Ancient Art: Near East, Egypt, Greece, and Italy (Cambridge, 1996), pp. 233-46. General discussions of the motif of Heracles and Omphale in art can be found in C. Caprino, ʻOnfale,’ Enciclopedia dell'Arte Antica V (1963), pp. 695-8; K. Schauenburg, ʻHerakles und Omphale’, RhM 103 (1960), pp. 57-76; and E.G. Suhr, ʻHerakles and Omphale,’ AJA 57 (1953), pp. 251-67.

30 Since several scenes are missing from the mosaic found in Liria, and some scenes on the medallion of Gordianus are rather poorly preserved, we cannot be certain that these objects contained scenes only from the 'canonical' Dodekathlos. Based primarily on the fact that the lowermost panel in the Liria mosaic depicts Hercules’ drinking contest with Dionysius, A. Ballil, ʻMosaico con representación de los Trabajos de Hércules hallado en Cartama,’ (as in n. 28), p. 37 has speculated that the missing panels may have contained scenes from the hero’s exploits in the Hades. Three of the scenes from the textile panel kept in New York (Fig. 16) have also been lost, and an additional one ruined to the point where we could only conjecture as to what labour it depicted, but their contents can be safely reconstructed from those of an identical panel in Saint Petersburg (Fig. 23). Instead of the canonical scene of the battle with Geryon, one of the scenes in the textile panels seems to be that of the battle with Kyknos, which does not appear in the Dodekathlos. Cf. F. Brommer, Denkmälerlisten zur griechischen Heldensage I (Marburg, 1971), p. 12, cat. no. 2. With this one certain exception, our five objects otherwise adhere to the canonical cycle established in the Olympian frieze.

31 P. Jongste, The Twelve Labours of Hercules on Roman Sarcophagi (Rome, 1992), pp. 11 and 32.

32 The other objects all display some significant difference from the vita icons – the Egyptian textiles do not employ any device for the separation of the narrative scenes; the floor mosaic from Liria does not have a fixed viewpoint for the labours; the mosaic from Cartama has an additional panel competing for prominence with that of the central representation of Hercules; the medallion of Gordianus, finally, is of circular rather than rectangular design.

33 While the presence of Omphale might, in an earlier artwork, suggest mockery rather than praise of the deeds of Hercules, the relief’s second-century date and Roman origin indicate that she was probably not meant to be a symbol either of the decadence of the East or of transgressive sexuality. There is no exchange of clothes between her and Hercules, the inscription shows no uneasiness about the possible ambivalence towards the male hero, and only the reversed attributes – quiver and bow beneath the feet of Omphale and the wool-basket and spindle beneath those of Hercules – remind us of the story’s former connotations. Cf. N.B. Kampen, ʻOmphale and the Instability of Gender’ (as in n. 29), pp. 239-42. The relief, thus, unquestionably belongs to the genre of artistic praises of the deeds of illustrious individuals. Bernhard Schweitzer has speculated that this mythological couple may even have been understood as a type of hierogamy in the tradition of Zeus and Hera. See B. Schweitzer, ʻDea Nemesis Regina,’ JdI 46 (1931), p. 238.

34 We must be careful not to push this too far, though, as the relief seems to display the same decline in quality which P. Jongste has noted as happening on the sarcophagi in the years between 170 and 180 A.D. ʻThe iconography of the twelve labours of Hercules changed. It was forced into a straitjacket of the frieze shaped chest with every labour plotted next to the other as close as possible in an attempt to have eight or ten labours depicted on the front and four or two on the short sides. The result was a rather dull composition of vertical lines. The actions became the static reminiscence of what used to be a glorious endeavour by a muscle-man. The real hero fighting a huge monster and performing the impossible became the athlete taming young deer and killing puppy-like creatures. This was sadly enough the price that had to be paid to satisfy a larger group of customers, who wanted their sarcophagi cheap and fast. The loss of quality could be concealed by way of paint and gilding. The customer became the servant of his own needs: not a specially made or uniquely composed quality product, but a marble funerary container in Greek style that could be afforded by the man next door as well.’ – See P. Jongste: The Twelve Labours of Hercules on Roman Sarcophagi (Rome, 1992), p. 33. While there are no perfect matches between this relief and any of the sarcophagi, the six labours in the top row of the relief are cramped together in much the same manner as described by Jongste. Close iconographic parallels for most of the individual panels of the relief can, moreover, be found in the sarcophagi. Compare, for instance, the Cerberus panel of this relief to the depiction of that labour on a sarcophagus in the Galleria degli Ufizzi, Florenze (inv. 110; reproduced in Jongste (op. cit.), fig 7, cat. nr. B1, pp. 43-5).

35 Pausanias, Description of Greece v.10.9

36 Pseudo-Apollodorus, Bibliotheke ii.5.1-ii.5.12. The order given in the Bibliotheke is only slightly altered, though, and seems to have a much stronger connection to this artwork (compare Tables 1 and 2).

37 These versions are: Sophocles, Trachiniae 1091ff; Euripides, Heracles 359-435, 1270-5; Diodorus Siculus, Bibliotheca Historica iv.10.7-iv.13.4, iv.15.3-iv.17.2, iv.25.1, iv.26.1-iv.27.5; Pausanias, Description of Greece v.10.9, v.26.7; Quintus Smyrnaeus, Posthomerica vi.208ff.; Virgil, Aeneid viii.287ff; Ovid, Metamporphoses ix.182ff (I have used the translation by B. More, available at:; Hyginus, Fabulae 30 (I have used the translation by M. Grant available at Out of these, Sophocles only mentions five labours (the Nemean lion, the Lernean hydra, the Erymanthian boar, Cerberus, and the golden apples); Euripides has ten labours, but one of them is the ‘non-canonical’ battle with Kyknos (missing are the boar, the Augean stables, the Stymphalian birds, and the Cretan bull); Diodorus Siclus has all the twelve labours, in order only slightly deviating from that of Pseudo-Apollodorus’ Bibliotheke; Virgil only has three labours – the bull, the lion, and Cerberus; Virgil does not name or describe the Dodekathlos but sings a rather generalised encomium to Hercules’ deeds; Ovid has seven deeds, and he mentions them in reversed order – starting from Cerberus and progressing towards the Nemean lion; and Hyginius gives basically the same account as Diodorus, disagreeing only on the last labour – in Hyginius it is the capture of Cerberus, rather than the theft of the golden apples.

38 One must, of course, always be careful not to push such arguments too far, as there are numerous examples of narrative art from all cultures and periods that, although canonical texts did exist, simply did not follow their chronology. This will be further explored later.

39 Cf., for instance, W.A.P. Childs, ‘Textile panel with labors of Heracles, and Dionysios’ in K. Weitzmann (ed.), Age of Spirituality (as in n. 21), cat. no. 136, p. 160.

40 In addition to the vast majority of the manuscript illustrations that narrate a story, good examples of this can be found in two plates and one marble disc adorned with scenes from the life of Achilles. See K. Weitzmann (ed.), Age of Spirituality (as in n. 21), nos. 207, 208, and 210, pp. 230-6.

41 As has been indicated above, the order that we now consider canonical is based primarily on the Bibliotheke, but if we accept that the Cerberus episode was in fact inadvertently omitted from Pausanias’ account (which only mentions five metopes on the eastern porch and thus probably does have a lacuna), then both sources would agree on the first and the last deed. An article on the Olympian metopes published in Perseus Digital Library explains that the viewer would have proceeded from the western porch to the East, reading the metopes left to right, and that the first and the last deed would thus be the Nemean lion and Cerberus, respectively. The Augean stables may have been put at the prominent place at the end of the sequence because of the scene’s local importance, given that Augeas was king of Elis. See the anonymous article ‘Olympia metopes (Sculpture)’ at All the other sources also agree on the Nemeian lion being the first in the series, and on either Cerberus or the golden apples being the last. Cf. n. 37 supra.

42 Cf. N. P. Ševčenko, The Vita Icon and the Painter as Hagiographer’, DOP 53 (1999), p. 151; P. Chatterjee, Narrating Sanctity (as in n. 1), p. 76.

43 The very consistent iconography of the vita icons certainly does point to the possibility that their artists have used pictorial models at least in addition to, if not in fact in place of, textual ones. And it is certainly possible that the individual scenes that would eventually be used to assemble a vita icon were reproduced and disseminated on separate and unnumbered sheets of writing material. Since we only know of a handful of Hercules cycles that resemble the vita icon form, and since these few objects have such different provenance, the limitedness of iconographic parallels between them is not particularly telling. In general, though, Hercules cycles have not only a canonised set of deeds, but also display some clear iconographic trajectories that could allow us to suppose the existence and use of at least some model books. I have not yet had the opportunity, though, to personally study this subject in enough depth to form any definite opinion on it. And, generally speaking, visual narration in Byzantium has attracted so little interest from my predecessors that this study can only begin to fill that gap.

44 Many elements of the classical orders are often thought to be tectonic representations in stone of the wooden architecture of the primitive hut. The alternating sequence of triglyphs and metopes would, thus, derive from the post-and-beam constructions in traditional timber framing.

45  As John Lowden explains, all these ivories ‘have survived in secondary use, and their relation to such imperial objects as the famous “Barberini diptych” in the Louvre has clouded their original function, so that they are sometimes treated as plaques rather than as book covers.’ See J. Lowden, ‘The Word Made Visible: The Exterior of the Early Christian Book as Visual Argument’ in W.E. Klingshirn and L. Safran (eds.), The Early Christian Book (Washington DC, 2007), p. 35. For earlier treatments of these ivories see W.F. Volbach, Elfenbeinarbeiten der Spatantike und des fruhen Mittelalters, 3rd edition (Mainz, 1976), nos. 119, 126-9, 130-3, 145, and 156; F. Steenbock, Der kirchliche Prachteinband im frühen Mittelalter: von den Anfängen bis zum Beginn der Gotik (Berlin, 1965), esp. nos. 5, 8, 10; D. Gaborit-Chopin, Ivoires du Moyen Age (Fribourg, 1978), pp. 26-7 and nos. 24-5; E. Kitzinger, ‘A Pair of Silver Book Covers in the Sion Treasure,’ in U.E. McCracken, L.M.C. Randall, and R.H. Randall, Jr. (eds.), Gatherings in Honor of Dorothy E. Miner (Baltimore, 1974), pp. 3–17; K. Weitzmann (ed.), Age of Spirituality (as in n. 21), cat. nos. 458-61. Hypotheses on the existence of imperial diptychs in Late Antiquity can be found in, for instance, J.D. Breckenridge, ‘Diptych leaf with Ariadne’ and (idem) ‘Diptych leaf with Justinian as defender of faith’ in K. Weitzmann (ed.), Age of Spirituality (as in n. 21), no. 25, pp. 31-2, and no. 28, pp. 33-5, respectively; A. Cutler ‘Barberini ivory’ and (idem) ‘Diptych’ in A. Kazhdan (ed.), The Oxford Dictionary of Byzantium, vol. 1 (Oxford, 1991), p. 254 and 636-7, respectively; A. Cutler, ‘Barberiniana. Notes on the Making, Content, and Provenance of Louvre OA. 9063’ in J. Engemann and E. Dassmann, Tesserae: Festschrift für Josef Engemann, Jahrbuch für Antike und Christentum, Ergänzungsband 18 (Münster Westfallen, 1991), p. 329-39. This theoretical type of ivory diptych would have leaves made up of five parts, with the central one representing the emperor. In contrast to the well-attested consular diptychs, however, no complete example of such an imperial counterpart exists, and even the most well-known example, often called the Barberini diptych, was, in all probability, not a diptych. 

46 The very few instances when these lateral panels do not depict episodes from the life of Christ do not, however, undermine our case. For instance, on the back cover in Milan, on the right of the bejewelled cross, we find an image of Christ seated on a globe, flanked by two figures, whose garlands He blesses. While this cannot be directly counted among Christ’s deeds on Earth, the meaning of the scene, whomever the flanking figures may represent, is certainly His institution of the Church. In a somewhat wider sense, then, even this panel does not exclusively belong to the iconic, as opposed to the narrative, mode. In a symbolic manner, it in fact depicts one of Christ’s most important and historically momentous deeds. We should also not be overly concerned with the fact that the images of the Virgin are predominantly surrounded by episodes that cannot strictly be called ‘deeds’ either. If we keep in mind the Virgin’s uniquely self-effacing role in the economy of salvation, scenes such as the Annunciation and the Nativity can indeed be understood as Her exceptional, even heroic, deeds. Finally, scenes from the story of Jonah (in the bottom panel on the back of the Murano diptych) were probably understood in the sense that this Old Testament prophet was mentioned in the Ordo commendationis animae – as prefigurations of the salvific actions of Christ, most especially that of His own resurrection. Thus, although the protagonist is, at first glance, different, Christian art proves uniquely capable of using episodes from more than one life to offer encomium to a single, divine, person.

47 See F. Wormald, The miniatures in the Gospels of St. Augustine (Cambridge, 1954), pl. II. The manuscript is traditionally considered to be directly connected to St. Augustine, either as having been brought by him to England personally in 597, or as having being sent to him by Pope Gregory the Great in 601. It is, therefore, most often referred to as the St Augustine Gospels. Cf. K. Weitzmann, Late Antique and Early Christian Book Illumination (New York, 1977), p. 22. Its alternative name, the Canterbury Gospels, can be rather confusing as it often refers to an eighth-century Anglo-Saxon gospel book that was actually produced at Canterbury. The illumination in question (f. 129v) depicts the evangelist Luke and appears as a frontispiece to his Gospel. All four Gospels in this manuscript were, in all likelihood, once preceded by similar portraits of their authors. The only other illumination that survives (f. 125), though, is a sequence of twelve narrative scenes, divided into four rows of three scenes each, detailing the events between Christ’s entry into Jerusalem and the way to Calvary.

48 Cf. G. Schiller, Iconography of Christian Art, Vol. II (London, 1971), p. 14. Schiller (op. cit., vol. I, p. 155) has also hypothesised that the narrative scenes might derive from a fresco cycle depicting the life of Christ in the Santa Maria Antiqua in Rome.

49 This oratory or, more precisely, John VII’s funerary chapel, was situated at the east end of old St. Peter’s nave and was destroyed with that last standing part of the Early Christian edifice in 1605. The cycle in question is known to us thanks to the fact that Pope Paul V commissioned G. Grimaldi to document the structure, before its demolition, ‘in picture et scriptura.’ Cf. M. Schapiro, Late Antique, Early Christian and Medieval Art – Selected Papers (London, 1980), p. 98, fig. 20; P.J. Nordhagen, Studies in Byzantine and Early Medieval Painting (London, 1990), pp. 58-104; H Belting, Likeness and Presence (Chicago, 1994), pp. 126-7 and fig. 76, p. 128; A. van Dijk, ‘Jerusalem, Antioch, Rome, and Constantinople: The Peter Cycle in the Oratory of Pope John VII (705-707),’ DOP 55 (2001), pp. 305-7, fig. 3.

50 For this we do not need to rely on Grimaldi’s drawings and descriptions, as this part of the mosaic was preserved and can still be seen in the church of San Marco in Florenze. On the origins and connotations of this imperial, Byzantine iconography in the papal Rome, cf. H Belting, Likeness and Presence (Chicago, 1994), pp. 126-7. It is also worth reminding ourselves here that John VII’s episcopate not only belongs to the period of the so-called Byzantine papacy (537-752 A.D.), but that he was also of Greek nationality and a son of a Byzantine official. Cf J.N.D. Kelly, The Oxford Dictionary of Popes (Oxford, 1986), p. 84.

51 That this imperial iconography was very much a conscious choice of Pope John VII’s is confirmed by the fact that we find the Virgin wearing the same purple garments and the same crown in an encaustic icon usually called Santa Maria della Clemenza, commissioned by the same Pope. Cf. H Belting, Likeness and Presence (Chicago, 1994), p. 126.

52 If I believed that a direct formal precedent for the appearance of the vita icons was indeed needed, I would be most tempted to seek it either in book art, where it was so immensely and lastingly popular, or precisely here – in a cycle commissioned by a Byzantine pope, executed most probably by Byzantine artists, and housed in a church that was always in position to exert its influence all over the Christian world.

53 The ivory is now kept in the Bodleian Library at Oxford (MS. Douce 176), inset in the eighteenth-century leather cover of a Latin Gospel lectionary that was also made around the year 800 A.D., for the nunnery of Chelles, near Paris. On the ivory itself and the very likely possibility that it originally belonged to this lectionary, see: L. Nees, ‘Carolingian Art and Politics,’ in R. E. Sullivan (ed.): The Gentle Voices of Teachers: Aspects of Learning in the Carolingian Age (Columbus, 1995), pp. 195-202.

54 See P. Lasko, Ars Sacra, 800-1200, 2nd edition (London, 1994), p. 24. Lasko (loc. cit.) also notes that three more scenes in this ivory also derived from some similar Late Antique source ‘because another small Carolingian Court school fragment of three scenes from a five-part diptych, now in the British Museum, must have used this same Late Antique source.’ Cf., however, also L. Ness, ‘Carolingian Art and Politics,’ (as in n. 53), p. 198, for a valuable caveat that the Carolingian artist was by no means a slavish copyist, but that he ‘has in fact “corrected” his late antique pictorial model by recourse to the appropriate textual source.’

55 The book was first published in 1621 and the frontispiece was accompanied by an explanation set in verse, which not only describes the presence and meaning of ‘Jealousie’, ‘Inamorato’, or ‘Hypochondriacus’, but also of the author’s portrait: 
‘Ten distinct Squares here seen apart, 
 Are joyn’d in one by cutters art. 
Now last of all to fill a place, 
 Presented is the Author’s face; 
And in that habit which he wears,
His image to the world appears, 
His mind no art can well express, 
That by his writings you may guess. 
 It was not pride, nor yet vain glory, 
 (Though others do it commonly) 
Made him do this: if you must know, 
The Printer would needs have it so.’

56 The engraving in Fig. 42.1 dates from 1670 and depicts the conclave that ultimately elected Pope Clement X. It is kept in Bibliothèque nationale de France, Paris. The one in Fig. 42.2, dating from 1691, presents the election of Pope Alexander VIII. It is found in Giovanni-Dominico De Rossi`s Album Louis-Philippe Châteaux de Versailles et de Trianon, Versailles.

57 As we have seen in Figure 17, comparable Chinese examples can also be found in much earlier periods. In fact, they are so numerous that it would take no less effort and space to compose a full catalogue of them than it would to catalogue all the Western examples.

58 N.P. Ševčenko, Vita icons (as in n. 1), p. 57 and fig. 42, p. 59.. She was referring to an illustration of Michael Dukakis from 1988, the year he ran for President of the United States of America.

59 Even from before the fall of Constantinople, we have a number of examples where the centre is not occupied by a portrait. In addition to the book covers of the fourteenth-century lectionary already examined (Fig. 9.1-2), particularly good examples may be found in icons of St. George, where the central panel is in fact a narrative image, depicting the saint on horseback, in the action of slaying the dragon, with additional background characters such as the king and the princess. If we are also willing to look at what Ševčenko would consider the vita icon’s afterlife in the Slavic world, the number of examples grows exponentially and includes all sorts of narrative scenes (based on both the Old and the New Testament as well as on various Christian apocrypha) and what is often termed ‘theological icons’ (see, for instance, a famous mid-sixteenth century quadripartite icon, produced by the Pskov masters and representing the divine economy, now in the Cathedral of the Annunciation, Moscow; or the late-sixteenth to early-seventeenth century work of the Stroganov school, now in the Solvychegodsk Museum of History and Art, that presents the renovation of the Temple in Jerusalem).

60 That the Byzantine viewers were very familiar with both verbal and visual encomium has been conclusively demonstrated by Henry Maguire. As he has shown, those who received a higher education in Byzantium were completely conscious not only of the practice, but also the theory, of what had already been standardised in the third century by the rhetor Menander. – Cf. H. Maguire, ‘The Art of Comparing in Byzantium,’ Art Bulletin 70 (1988), pp. 88-103. It is particularly interesting to note how the vita icons conform to but also reinterpret the standard encomiastic sequence, namely that of narrating and praising the person’s birth, upbringing, deeds, and the life’s acme. For a Christian saint, the conversion to the one true faith or the sacrament of baptism is sometimes substituted for the scene of his physical birth. His upbringing does not always refer to his childhood and youth, but may also happen in mature age – as his instruction and growth in the Christian faith, and perhaps even as his diaconal, priestly, and/or episcopal consecration. And the acme of his illustrious life is, in a vast majority of the vita icons, in fact his martyrdom.

61 See Y. Ustinova, Caves and the Ancient Greek Mind: Descending Underground in the Search for Ultimate Truth (Oxford, 2009), pp. 47-52.

62 N. P. Ševčenko, Vita icons (as in n. 1), pp. 57, 61, and 67; eadem, ʻThe Vita Icon and the Painter as Hagiographer’ (as in n. 1), p. 152.

63 W. Burkert, Creation of the Sacred: Tracks of Biology in Early Religions (Cambridge, MA and London, 1996)

64 R. Beck, The Religion of the Mithras Cult in the Roman Empire: Mysteries of the Unconquered Sun (Oxford, 2006)