Previous Section Contents Page Next Section




Pictures don’t tell stories;    
viewers tell stories.


For Jane


Theoria sed non pro gratia theoriae...




The period covered by this thesis might justly be seen as the zenith of narrative art in northern Europe. This was a time of frenetic growth in the use of narrative imagery. The great Gothic cathedrals were built, sculpted, glazed and equipped, each offering myriad spaces for the display of visual narratives. Meanwhile, increasing numbers of lay patrons were investing in small-scale devotional and secular objects embellished with narrative scenes. Artists responded to this demand by finding new ways of turning stories into images - and narrative art developed a formal complexity unmatched by later periods.
Many of the narrative techniques artists employed were common to multiple media and contexts. They are therefore obvious candidates for the kind of structuralist analysis developed in the field of textual narratology. Over the past three decades, a few art historians have engaged enthusiastically with such discourses. Generally however, narratology has been treated as an embellishment for existing studies of favoured objects or media, rather than a subject worthy of investigation in its own right. This thesis seeks to reverse that relationship - it starts by exploring the current state of textual narratology and then considers its applicability to medieval narrative art. The first three chapters explore various theories to consider (amongst other things) how medieval images in general ‘work’, what is distinctive about narrative images and what roles frames play in their design and reception. The remaining chapters develop the resulting theoretical models further by means of a series of case studies, whilst also introducing further elements of theory and exploring those aspects of visual narratology which are of particular relevance to medieval art. The penultimate chapter is also a case study - but one which aims specifically to address the question of whether visual narratology as a methodological approach has anything to add to our existing understanding of medieval artistic practices.



The initial spark out of which this thesis grew came whilst I was a student on Professor Paul Crossley’s MA course at the Courtauld Institute on ‘The Gothic Cathedral’. An assessed essay on the archivolts of the Job portal at Chartres gradually expanded into an MA dissertation on narrativity in French Gothic portal sculpture, which in due course led me on to the wider consideration of narratology contained herein. It was therefore a pleasant irony that the week before submitting this thesis I found myself back at Chartres, discussing the Job portal archivolts with Professor Crossley’s latest crop of MA students and various other friends and colleagues. Seeing how readily they took on board my explanations of what have sometimes seemed like obscure points of abstract theory was a great boost to morale during these last few steps of the journey.
It scarcely needs saying that without Professor Crossley’s support and encouragement this thesis could never have been written. His boundless enthusiasm and good humour have always proved the perfect antidote to the kind of existential angst that bedevils any PhD project but which is a particular hazard with unconventional topics such as this one. His trust in me is what made this project possible and I am eternally grateful for it. Anyone who studies at the Courtauld Institute benefits from the nurturing effect of being amongst so many highly-tuned minds but I suspect that with my curious interdisciplinary interests I have benefitted more than most. Professor John Lowden has been like a second supervisor at times, patiently filling in the copious lacunae in my knowledge of medieval manuscripts and reassuring me that the word ‘narratology’ really does exist. I have also benefitted from working and drinking with an outstanding generation of fellow students, including (amongst others) Ed Payne, Federico Botana, Francesco Luccini, Jim Harris and, in particular, Dr Laura Cleaver and Hanna Wimmer (Hamburg University), who were my main collaborators on the ‘Medieval Art in Theory’ research project.
I am indebted to the Arts and Humanities Research Council, who funded the first three years of this research project, and also to the Wingate Foundation, who generously supported my initial photographic trips to Bourges Cathedral. However I would never have been able to complete this thesis had it not been for my former colleagues in the IT department at Rabobank, who offered me the temporary employment contract that paid for the essential extra six months of writing after my AHRC funding dried up. I am also grateful to those same chaps for keeping me supplied with beer and good company through my long years of student penury - and for never letting me buy a round.
Above all though I am indebted to my wife, Dr Jane Hamilton-Whatling - for pretty much everything.


Until now, the relationship between art history and narratology has been somewhat flirtatious. Researchers from either side have dallied with the other’s objects or theories, sometimes with slightly questionable motives. Apart from a natural desire to participate in fashionable discourses, the attraction of narrative theory for art historians has generally been the valorisation of some favoured object or other on the basis that its ‘exceptional narrativity’ makes it worthy of yet another study. Coming in the other direction, various narratologists have treated the world of the visual arts as a laboratory to experiment with their existing theories - although almost without exception they have focused their gazes on the kind of post-Renaissance art which, though better known to the general public and more widely accessible, is generally less amenable to consideration as a narrative ‘text’ than are the kind of images on which this thesis focuses.

From the perspective of a medievalist, Wolfgang Kemp’s ground-breaking study of the narrative structure of stained glass, his 1987 Sermo Corporeus (later published in English as ‘The Narratives of Gothic Stained Glass’, Cambridge, 1997), promised much - yet for some reason failed to spawn the expected new generation of visual narratologists. Whilst Kemp’s work undoubtedly had a profound influence on many students of medieval art (including the present author), few seemed inclined to take up the challenge with any real enthusiasm.1  What we have seen instead has been a tendency to pay lip-service to narratology, reader-response theory and similar paradigms, treating them as fashionable condiments with which to season studies which otherwise taste of traditional art history. As Kant might have said, narratology has become a parergon (see chapter 3.a), used primarily as an ornament with which to embellish perfectly solid but nonetheless conventional historiographic modes. This thesis, like the MA dissertation it grew out of, is an attempt to reverse the relationship - to see what, if anything, could be gained from starting out with a coherent body of narratological theory and then applying it to medieval art in general, across a wide range of media and contexts of viewing. My goal is not to build a ‘general theory of medieval visual narratology’ for its own sake but to see what kinds of questions such an approach raises - to see whether it really does offer novel ways to interrogate familiar objects.

In the event, this thesis has turned out to be at once much less and rather more than I had originally intended. What is presented here is roughly one third of the original plan which, for much of its long period of gestation, was referred to (with tongue firmly in cheek) as my Summa Narratologica. At the heart of that scheme was the idea that a field like visual narratology could always be approached from any one of three main directions. Firstly, one could concentrate on a particular type of object, or a particular setting for narrative images (the narratives of enamel reliquary chests or portal sculpture for example), or one could go even further down this line of specialisation and just study the narratives of a single object or building. A second approach would be to start from the story and consider the visualisations of a particular type of narrative, such as Apocalypse images, hagiography, or some group of secular romance epics. Finally, a third way would be to take the more theoretical road and consider the ‘generic structures of narrative’, and how they are expressed in images, regardless of medium or subject. The first of these approaches I like to think of as the syntagmatic axis, the second as the paradigmatic axis and the third as the ‘structural’ axis.

The distinction between syntagmatic and paradigmatic analysis is an important one, which has been summarised thus:

Syntagma is an orderly joining of signs or elements into a syntactic unit. Thus syntagmatic analysis is concerned with the building blocks of ‘language’ and how they are joined together - for example, how words are formed into correctly structured sentences or how architectonic primitives are combined in a [Gothic] portal. Paradigmatic analysis, by contrast, is concerned with the interchangeable forms or meanings that each particular element within the syntagma can take. The syntagmatic axis is considered as orthogonal to the paradigmatic and each element in an assemblage can be considered from either viewpoint. For example, a lintel occupies a particular position within a portal, where it has specific spatial relations to the other components (syntagmatic approach) but it can also carry relief sculpture, whose subjects (paradigmatic axis) could also appear in other contexts and are not specific to ‘lintel-ness’. The overall meaning of a semiotic construct, or assemblage, is therefore the product of these two axes.2

My plan for the current thesis was originally to split it into three sections. The first section was to have been the ‘structural corpus’, bringing together all the various elements of semiotics, cognitive theory and narratology which I regard as important in understanding medieval narrative art. The second section was to have been the ‘paradigmatic corpus’ - a survey of the major types of narrative (Old Testament, New Testament, Hagiography, Romance epics, and all their various sub-divisions) that provided food for the medieval artist, together with a consideration of how subject matter affected the ways in which particular types of stories were expressed in images, regardless of medium. Finally, the third section would have been the ‘syntagmatic corpus’, exploring the many different settings for narrative art within the ‘hypernarrative’ of medieval visual culture as a whole -  and how the specifics of physical context and medium influenced the manner of the visualisation, regardless of the type of narrative story (a small part of that particular problem-space was mapped out in rudimentary fashion in my MA thesis).

It was when the still very-far-from-complete structural corpus reached around ninety thousand words on its own that I realise my tripartite Summa model was no longer viable. I therefore decided to change tack and concentrate on the theoretical aspects of visual narratology but to include a large number of examples and case studies of varying length - both to help explain the theory and also to stimulate further investigation. I was however reluctant to entirely relinquish the triumvirate of syntagmatic, paradigmatic and structural approaches, not least because I find these three orthogonal axes a helpful way of thinking about the various historiographic approaches to cultural history generally. I have therefore retained ‘stubs’ of the three original corpuses in chapter 8, which contains, if not a corpus then at least an ‘excursus’ aligned with each axis.

It will be noticed that nearly all of the examples and case studies presented in this thesis are French. Some of the illuminated manuscripts discussed are English or German - but all of the examples of stained glass and portal sculpture are in France. My reasons for this are partly due to questions of original production and the survival of material - but they are mainly practical. A narratological study of a stained glass window for example requires highly detailed photographs of every single panel, plus as many related windows as possible. One or two isolated images from a cycle are effectively useless. English cathedral administrators generally discourage serious photography and whilst I would have liked to include the remarkable windows of Canterbury Cathedral in my study, my approaches concerning photography remained unanswered. By contrast I found I was able to spend weeks at a time in the cathedrals of Bourges, Chartres, Angers, Poitiers, Tours, Le Mans and Rouen, from dawn to dusk, photographing every window, panel by panel, without any interruption or official hindrance. 3

Even though I have not yet had the opportunity to consider monumental narrative art outside France, I believe that regional variations would make an important additional level of analysis for future study,  as would the issue of how narrative techniques changed over time. A true Summa Narratologica would thus require not three but five axes of study - the structural, the syntagmatic, the paradigmatic, the geographical and the diachronic. This thesis addresses only the first of these axes. The remaining four may have to wait a while.


Previous Section Contents Page Next Section



1 Although I do not always agree with the content of Professor Kemp’s book, there is no escaping the fact that without its stimulus, this thesis would never have been written.

2 S. Whatling, Narrativity in French Gothic Portal Sculpture (MA Thesis), University of London (Courtauld Institute of Art), 2005, p. 8.

3 With a few exceptions, the photographs of stained glass in the accompanying volume of plates were all taken by the author between 2006 and 2009, using a Nikon D2X camera equipped with a 180-500mm zoom lens, mounted on a heavy tripod. Most of the images were then post-processed in Adobe Photoshop CS4 using a range of techniques to correct the problems of perspectival distortion, colour desaturation and extreme contrast range that bedevil any attempt to photograph stained glass. These images and many others can also be found on the author’s website,, which aims (eventually) to present high resolution images of all the major 13th century windows in France.