HOW TO TELL A FAIRY TALE WITH IMAGES : Narrative Theories and French Paintings from the Early Nineteenth Century

This article first discusses theoretical approaches to the question of pictorial narrative, and argues that images can generate a narrative, but do so by different means than texts. Consequently, visual narratives should not be analysed using the same criteria as developed for textual narratives. Based on this idea, the article further analyses two French paintings from the early nineteenth century that represent a fairy tale by visual means alone, and which can be considered as paintings that tell a fairy tale: Petit Chaperon rouge (c. 1820) by Fleury François Richard, and Peau d’âne (1819) by Jean-Antoine Laurent.

The printed editions of French fairy tales, fables and contes from the seventeenth and eighteenth centuries were so frequently illustrated with drawings or engravings, that one might even suppose that the texts were considered incomplete without an illustration. Illustrated books must also have appealed to the audiences of this period because the question of the relation between poetry and painting was still at the heart of all philosophical reflections on the arts. 1 Literary authors contributed to these interart comparisons, as did Charles Perrault in his poem La peinture, and Jean de La Fontaine, who included a praise of painting in his Songe de Vaux and who used occasionally ekphrastic techniques in his poetry (Bohnet). It comes as no surprise that in this intellectual and artistic context visualisations of fairy tales, fables and contes (or in modern terminology: their intermedial transfer to an image) were highly sought after, not only as illustrations in a book, but also as autonomous paintings. The fables and contes by La Fontaine were used as the subject for paintings as early as the beginning of the eighteenth century, for instance by Jean-Baptiste Oudry (1686-1755) and by Nicolas Lancret . The contes de fée, however, only appeared in painting from the late eighteenth and early nineteeth century onwards (Van-Eecke).
The engraved illustrations in the printed editions of the contes de fée were in black and white and sometimes quite schematic. However, for paintings that narrated a fairy tale by strictly visual means, outside the direct context of a book, the artists had to use different visual strategies. Because illustrated fairy tale books were widespread, the painters had without any doubt knowledge of their iconographical characteristics, but they could not always simply translate an engraved illustration into a painting, because a painting demands a different approach. Illustration and text interact with each other and there is a certain degree of cohesion between the two (Hoogvliet), while a painting is a 'stand alone' visual retelling of a story where the artist has to be more explicit about the setting, the colours and the appearance of the protagonists, and, even more importantly, the moment of the story has to be chosen strategically and the visual narrative has to be functional on its own.
In this article, I would like to address the question of narrative paintings, or images that visualize a story, or even images that tell a story. This is an age-old discussion where one side claims that paintings can become a visual narrative in their own right, while others insist that images simply cannot tell stories. Consequently, I will start by reflecting upon some theoretical approaches to the question of narrative images, both from the past and from the present, in order to subsequently discuss two French paintings from the early nineteenth century that visualize a fairy tale, or paintings that tell a fairy tale.

Narrative Theories and Images
Before entering upon an overview and a discussion of theoretical approaches to narrative images, it is necessary to make one preliminary remark: in my approach, I do not presuppose the existence of pre-iconic or extra-iconic laws that govern what an image is, what an image can or cannot perform. As in modern linguistics, my approach is purely descriptive: I study the ways in which visual discourses are being constructed, and I do not make judgements about 'good' images, 'wrong' images, or 'impossible' images.
Let us consider the first part of this quotation. Here the author indicates that the Renaissance was the period in history where painters started to make images that represent a punctum temporis, also known as the 'pregnant moment': that one split second in a narrative where everything changes; the point in time that marks the cleavage between 'before' and 'after'. This is a historical commonplace that is still often repeated in modern text books, although it has been dismissed by Lew Andrews, who has pointed out in his study Story and Space in Renaissance Art (1995) 2 , that precisely because of the rebirth of pictorial space during the Renaissance, it became possible to produce narrative paintings where several episodes from a narrative are being represented in the unified space suggested by linear perspective. Andrews calls this type of visual representations "continuous narrative", but others call it "multiphase pictures". This visual strategy was employed several times by one of the heroes of Renaissance painting, the Italian painter Titian, for instance in his Diana and Actaeon (1556-1559), where Actaeon and Diana are seen twice within the frame of a single pictorial space: in the foreground Acteaon accidentally surprises the goddess when she takes her bath, and in the background we see Diana another time, now hunting Actaeon who has been transformed into a deer. For us it is difficult to accept images such as these; we are inclined to qualify them as 'wrong' images, because in our modern culture we have different pictorial conventions: for us, pictures should suggest a visual experience, as if we were looking through a window to the real world during a single moment, and in the real world a person cannot be in two places at the same time. In spite of our modern ideas, this type of visual narrative was often used by the Romans (Penn Small) -we only have to think of famous examples such as Trajan's column -and this explains in part the importance of this mode of visual narrative for Renaissance painters. Continuous narrative employed in images has often been qualified as 'wrong' and 'medieval', but Andrews argues that because the painters of the medieval period did not employ linear perspective to suggest pictorial space, they could simply not use continuous narrative. This bold statement has been criticised recently by Nathalie Crohn Schmitt (2004), who argues convincingly that continuous narrative was used by the medieval painters of the miniatures in the Holkham Bible Picture Book and in Queen Mary's Psalter.
It was only from the seventeenth century onwards that the punctum temporis, or 'pregnant moment' became an imperative for paintings evoking an action, a history, or a narrative, due to the influence of Aristotle's statement that art should imitate nature. Theoreticians of art saw the unity of time and space as a compelling rule for paintings (Blanc,(161)(162)(163)(164)(165)(166)(167)(168)(169)(170)(171)(172)(173)(174)(175)(176), and I need not recall the discussions provoked by Poussin's painting Les Israélites recueillant la manne dans le désert, painted between 1637 and 1639 (Unglaub). Later, during the eighteenth century, Gotthold Ephraim Lessing in his Laokoon, oder über die Grenzen der Mahlerey und Poesie (1766) sharply demarcated the boundaries between poetry and painting: poetry is a temporal art, while painting is a spatial art that lacks a temporal axis. During the beginning of the twentieth century, modernism inspired a quest for pure images (Elkins, Mitchell 1994, 5, Timms & Schulz), in combination with a rejection of figurative and narrative art. As a result of all this, our modern eyes and minds do not easily accept narrative paintings.
Our assumptions about what images can do and what images should do have been criticized sharply by W.J.T. Mitchell in his book Iconology: Image, Text, Ideology (1986) in which he critically undermines our modern cultural conventions about images. In the chapter relevant for us here, entitled "Space and Time" (95-115), he discusses in detail the arguments advanced by Lessing, and especially the ideological prejudices that underlie his argument that poetry is a temporal art and painting a spatial art. Mitchell argues convincingly that it is especially Lessing's abhorrence of 'impure' paintings that informs the importance he attached to clear and distinct limits of the arts, because for Lessing impure genres implied decadence and cultural decay. This indicates that the absence of the dimension of time in a painting is not a natural given; it is a cultural convention and subject to change.
If we return now to the quotation of Fresnault-Deruelle reproduced above, especially its second part: "...., mille exemples montrent qu'une certaine « orientation des faire » (selon les axes gauche/droite et avant/arrière) continue, plus ou moins subrepticement, d'informer la dispositio générale des oeuvres", we can say that the author is right in claiming that paintings with a historical or narrative subject almost inevitably represent an action which is oriented in a certain direction. We also might qualify the direction of this action as a temporal axis. But is this temporal axis the same as a narrative? Fresnault-Deruelle himself, in his book L'éloquence des images (1993b, 186), introduces the term "narrativité", which he defines as the simple possibility of a narration, which is present in all figurative paintings. But many specialists in literary narratology are very doubtful about the possibilities of visual narrative, and often they conclude that images cannot generate a genuine narrative. For instance, Áron Kibédi Varga ( , 1989Varga ( , 1990) is critical about the possibilities of autonomous visual narratives, that is, images generating a narrative independently from a pre-existing textual narrative. In his view, the visual and the verbal are inseparable and the paratactic positioning of events in an image is not the same as a narrative. Images can only suggest a narrative, and when it comes to images visualizing a story, he concludes: "The image is not a second way of telling the tale, but a way of evoking it" (1988,204).
Marie Laure Ryan, in her introduction to the articles collected in Narrative Across Media: The Languages of Storytelling (2004, 139) qualified images that evoke a narrative as "the illustrative mode": "Compared to the ability to articulate new stories, illustrative narrativity is admittedly a rather weak and subordinated mode, but this does not mean that it should be dismissed as entirely parasitic". During the past years, narratology has been starting to change its outlook by "disconnecting the concept of narrativity from textual media" (Meister, xi), and Ryan has evolved towards a more nuanced theoretical position: "There are, quite simply, meanings that are better expressed visually or musically than verbally, and these meanings should not be declared a priori irrelevant to the narrative experience" (Ryan 2005, 10).
The progress of the theoretical discussion is not helped by its participants' lack of interest in publications concerning the medieval period, because medieval painters had an original and creative approach to narrative images, and this should inform the modern discussion. Michael Curschmann (1990Curschmann ( , 2007 has underlined in several articles that certain images from the Middle Ages generated a narrative by visual means only, even without the existence of a pre-existing textual narrative. Other scholars have explained that the apparent lack of narrative capacities of paintings is not the result of the medium itself, but rather of our cultural conventions concerning its use, as Wendy Steiner puts it in her seminal book Pictures of Romance (1988, 9) 3 : Many of the traits producing strong literary narratives are the same as those producing strong pictorial ones, but historical developments have made strongly narrative paintings extremely rare. It is not the medium of painting but its conventions that have reduced narrativity to an apparently peripheral concern for art historians.
A different approach, based on linguistic theory and social semiotics, can be found in Reading Images: The Grammar of Visual Design (1996) a ground-breaking and fundamental study by Günter Kress and Theo Van Leeuwen. In chapter relevant for us here, entitled "Narrative representations: designing social action" (43-78) the authors indicate that, in their opinion, the existence of narrative images is a natural given: "What in language is realized by words of the category 'action verbs' is in pictures realized by elements that can be formally defined as vectors" (44). Vectors may be formed by bodies or limbs or tools in action. Based on Halliday's functional semiotic theory Kress and Van Leeuwen argue that images can represent actors and vectors, and that the presence of these two elements can create "narrative patterns" within visual representations.
One of the recent contributions to this discussion is by Werner Wolf in an article with the title 'Narrative and narrativity: a narratological reconceptualization and its applicability to the visual arts' (2003). 4 Wolf rightly criticizes the loose definition of visual narrative by modern scholars, especially art historians: "it is still either the reference by means of a visual representation to some literary narrative, or the representation of any kind of action in a picture" (180). In his argument, Wolfs conceives of narrativity as a gradual notion, where narratives can be "weaker" and "stronger". Consequently he considers "serial pictures" and "multiphase pictures" (continuous narrative) as strong narratives. On the other hand, pictures that represent a "pregnant moment" or a "frozen action" are a weak narrative, because of their derivative nature: these images refer to a story and the viewer has to narrativize the represented moment. These weak narratives include the type of visual vectors that Kress and Van Leeuwen consider as narratives. He concludes: Central elements of narrativity are not easily to be realized in a painterly medium, and in some cases they are even downright impossible.
[…] On the other hand, one could argue that the visual arts are superior in narrativity in at least one detail: they can represent parallel scenes of a multi-strand narrative simultaneously (192). This last remark by Wolf is very important: images can be used to represent a narrative in a way that is not obvious for a text. Narrativity simply operates differently in a visual narrative. The problem of most of the approaches to visual narrative presented here, is that they tend to take textual narrative as the given against which visual narratives are being measured. These are consequently dismissed because they do not perform narrativity in the same way as texts do. But in fact, texts cannot perform a narrative; texts can only refer to actions by means of verbs. Images can also refer to actions, but by different means than a text. Moreover, a textual narrative is usually composed of sequential phrases. We have seen that 'strong' visual narratives can be composed of sequential images. Therefore it is more correct to consider monoscenic images of the punctum temporis type as a single phrase cut out from a larger narrative, as something like a film still, rather than to insist on the incompleteness of the image. When discussing pictorial narrative, it is better to speak of short narratives instead of 'weak narratives', and of long narratives instead of 'strong narratives'. With this in mind, I will discuss in the following paragraphs two narrative paintings that tell a fairy tale, and in doing so, I will focus on the specificity of their strategies for generating a visual narrative.

Paintings that Tell a Fairy Tale
In the past, painters often chose a historical, mythological or fictional narrative as a subject for their paintings. It would not be correct to consider these paintings as still lives, because something is happening there. So let us turn now to paintings that represent narrative scenes from fairy tales. Usually, these paintings are associated with German romanticism and it is true that some of the most splendid examples were made in that country, such as the very rich and complicated painting of Aschenbrodel (Cinderella) made by the German painter Moritz von Schwind in the years 1852-1854, now in the Alte Pinakotek, Munich. This painting marked the beginning of a vogue of fairy tale paintings in Germany during the second half of the nineteenth century.
As Regina Freyberger has recently showed, this painting by von Schwind reaches amazing levels of narrative complexity by combining scenes from the Cinderella story with scenes from the stories of Sleeping Beauty and of Cupid and Psyche.
It is less known that in France painters had already started decades earlier to use fairy tales for the composition of their paintings. In 1992, Corinne Van-Eecke published an article entitled 'Contes et fables dans les livrets de salon', for which she searched a sample of the catalogues of the Salons between 1700 and 1914. The first example of a painting of a fairy tale that she found is Le Petit Poucet retrouvant son chemin by the painter Jean-Louis Demarne, presented in the Salon of 1798. Unfortunately, the present location of this painting is not known. Several other fairy tale paintings were exposed in the Salons during the first two decades of the nineteenth century. These paintings are all associated with the style troubadour of that period (Baudson, Chaudonneret 1980, Petit, Tscherny & Stair Sainty): 5 Of these paintings, only the location of Jean-Antoine Laurent's Peau d'âne is known (Bourg-en-Bresse, Musée de Brou) (fig. 2). The list is not complete though, and painters must have made other paintings of fairy tales that were not exhibited at the Salons. For instance, the Louvre owns another fairy tale painting from the same period, Le Petit Chaperon rouge (c. 1820) ( fig. 1) by the Lyon-based painter Fleury François Richard.
In France, from the late seventeenth century onwards, a cultural and intellectual current gained force that aimed to give new life to the national past (Petit). During the eighteenth century, paintings with historical subjects from France's past, especially form its medieval and Renaissance past, were used for the propagation of patriotic sentiments, preferably by highlighting the emotional and sentimental aspects of the events represented. During the early nineteenth century, the contes de fée were also considered as a fundamental part of France's heritage from the Middle Ages, and, consequently, of its national identity. This incited painters to turn to fairy tales as a subject for their paintings (Petit, 96-96, Tscherny & Stair Sainty, 65-66, Chaudonneret 2002, 73).
Nationalistic and nostalgic sentiments are most notably present in the paintings by Fleury François Richard (1777-1852) (Baudson, 48-51, Chaudonneret 1980, 47-109, Chaudonneret 2002, and his painting of Le Petit Chaperon rouge is no exception. Richard placed the story in the half-dark interior of an old chapel, or a shed. There is a strong contrast with the sunny summer's day outside, visible through the small Gothic arched windows. The light falls dramatically on the wolf in the bed and on Petit Chaperon rouge, who is just a young and vulnerable child. She greets her 'grandmother' by lifting her hand slightly, and she does not seem to recognize the wolf, although the animal is obviously not disguised. The narrative in this painting is of the classical punctum temporis: it shows us only a very brief moment from the narrative, just before the violent outcome of the story. Using the terminology of Kress and Van Leeuwen, the only vectors visible are the greeting gesture of the child and the luring (or menacing) movements of the wolf's paws; there is no vector indicating the actual attack of the wolf. However, the skirt and shoes lingering on the ground make clear to the spectator what just has happened to the grandmother and what is going to happen to Petit Chaperon rouge in the next moment. Consequently, we might say that grandmother's clothes introduce a temporal reference to the painting, because these objects enlarge the very short visual narrative of one single moment into a longer visual narrative of several moments.
But Richard's painting of Petit Chaperon rouge does more than narrating the story visually. It also contains a play with binary opposites: light-dark, innocence-evil, before-after. Another important aspect is the attention to detail, both in the number of objects and in the depiction of surfaces. In many of the troubadour paintings the historical décor was as least as important as the historical event or literary theme represented. Richard studied in detail the medieval religious architecture of the Lyon area and he often used this as the background for his paintings. Likewise, the setting he chose for his Petit Chaperon rouge is a small medieval building with narrow gothic arched windows. The plaster is coming off the walls and these details, together with the spider webs, add to the impression of an age-old architecture. There is hay on the floor, we see some agricultural tools and there is a wooden construction in the background. These picturesque details evoke romantic poverty and simple country life. The message of this painting is an unambiguous cultural claim: since medieval times, the story of Petit Chaperon rouge has been a part of France's cultural heritage, especially of that of honest and simple country folk. 7 Jean-Antoine Laurent (1763-1832) was originally a miniaturist and a porcelain painter (Chaudonneret 28,(41)(42)Tscherny & Stair Sainty,(74)(75). From the early nineteenth century onwards, he turned to easel painting and his works were highly appreciated, especially by Empress Joséphine who bought several of his paintings. As we can see from his presentations at the Salons, he seems to have had a particular interest in depicting fairy tales but, unfortunately, only Peau d'âne from the Salon of 1819 survives. Here the architecture is not from France's national past, but its Mauresque arches and arabesque decorations evoke an opulent and dreamlike Orient, that fits well with the story of Peau d'âne. The richly inlaid furniture and the burning incense-pot add to this orientalizing effect. The costumes, on the other hand, seem rather medievalizing and European. It seems as if we see here a classical example of the punctum temporis: Peau d'âne has just put the ring on her finger and thrown off the ugly donkey skin, and she appears in all her beauty in her splendid dress.
The prince kneels in front of her and asks her to marry him, her new mother-in-law warmly receives her into the family, while the king admires the ring on her finger. But the narrativity in this painting is more complicated. Above the head of Peau d'âne an inscription in semiarabesques reads "Aimez la vertu", 8 and it is as if the older woman says these words to her, as a kind of advice. Maybe we can read this older woman also as the fairy godmother who guides Peau d'âne and the old king on the left as her father who wants to marry her. Consequently, we are also looking at a painting of Peau d'âne who has to choose between an incestuous marriage with her father and a morally acceptable marriage with the young prince. Laurent's painting very sophisticatedly combines two narrative scenes in one painting, and in this manner his painting has a narrative complexity that texts can hardly achieve.

Conclusion
Pictorial narrative has been the subject of many theoretical reflections and its possibility or impossibility has been hotly debated in the past. The discussion still continues in our days, although modern art seems to have lost its interest in single pictures narrating a history since its rejection of figurative painting and the widespread availability of moving images. The question if pictorial narrative is theoretically acceptable seems to depend much on the willingness of the scholar in question to accept it. In all cases, judging pictorial narrative by measuring it to the characteristics of textual narrative does not produce an interesting, nor an effective theoretical argument. The two examples of French fairy tale paintings from the early nineteenth century by Fleury François Richard and by Jean-Antoine Laurent discussed here, are a perfect illustration of the infinite number of choices that artists can make while painting a narrative. This is also the reason why artists, from painters to film directors, have always been highly interested in the transfer of a narrative from a textual to a visual medium, because it enables the exploration of new perspectives on the story and of new narrative possibilities.