To represent what a mural work site might have looked like in the 12th century, we need to consider the physical constraints involved in the enterprise: feasibility, cost, materials, time, etc. These aspects were necessarily a subject of discussion or negotiation between the various protagonists, that is to say, at the very least, the patron and the craftsman who gathered a team around him to produce the desired decoration. Any agreement between the two parties would have been based on a binding contract, written or oral. It seems hard to imagine, even in the 12th century, that this commitment would simply rely on the good faith of the parties involved, either in regard to the money required for its realisation, or the ability to execute it. If we accept this as a precondition, the question of the draft drawing inevitably arises. But to produce such a draft, the painter would need to have acquired a certain amount of knowledge. This knowledge was assimilated in the course of his training and the exercise of his profession. He had knowledge of materials and techniques, of drawing and the practice of pictorial representation. The tools required to carry out his work and the preliminary tasks like setting up a scaffold, preparing the wall by coating it prior to painting it (unless these tasks had been already been carried out by stonemasons), the laying out of the pictorial composition by tracing or the use of preparatory drawings, the preparation of pigments and binders and choice of the application technique are all steps that were essential to realising the mural itself.1
In the Liber diversarum artium or LDA (Ms H277, inter-university medical library of Montpellier), the first skill to be acquired by the painter is the mastery of drawing.2 Moreover, this ability is presented as the foundation of all painting and the first stage of artistic training. Aside from the LDA, the only other manuscript that refers to the importance of drawing skill in the artist’s training is that of Cennino Cennini, in his Libro dell’arte written in 1437: ‘The foundation of art and the beginning of all manual work is based on drawing and colour’.3 In the other treatises, drawing is only mentioned in passing, as, for example, in the 12th century, by Alexandrer Neckam (1157–1217), who considered it part of the artist’s apprenticeship without giving details of the process.4 The LDA is therefore the only treatise on painting prior to that of Cennino Cennini to measure the importance of learning how to draw and master proportions, as well as the ability to fabricate and use drawing and painting implements.
In the LDA, the apprentice is first taught how to draw on what might be called a rough copy, which was a wooden tablet bleached with soap and bone white, on which the shapes were drawn using a stylus (probably made out of wood, bone, copper or silver).5 This practice is also described by Cennino Cennini, with the young apprentice painter starting by drawing on such tablets and continuing to do so for a year.6 The English monk, Alexander Neckam also advises apprentice goldsmiths to use tablets to sketch their first drawings of jewels. It is easy to see the connection between these wooden tablets, bleached with soap and bone white, and the wax tablets known since Antiquity and widely used in the Middle Ages7 (Figure 24.1). Another parallel can be drawn with the glaziers’ templates or tables referred to already in the 12th century by the monk Theophilus.8 He describes the drawing of a pane of stained glass on a wooden board coated with chalk, whose other side is left free to lead out the window. The board was used not only to mark out the main drawing, which was identical to the lead lines, but probably also the interior drawing and perhaps even the shading. The colours of the various parts were marked with letters. The collaboration between glaziers and painters tends to be well documented, even if it is complex. Documentation about making stained glass often appears to refer to the work of painters, including the making of cartoons.9 These templates were evidently considered to be valuable work instruments, to the point of being subject to theft or bequest. In 1398, one of Jean de Berry’s painters, Jacquemart de Hesdin, was accused by Jean de Olanda of having broken the lock on one of his trunks preserved in the palace of the Duke of Poitiers, in order to steal some of his colours and templates.10
In the first stage of apprenticeship described in the LDA, the apprentice painters learned how to do things correctly: drawing a straight line, a true curve, a square, etc. They also had to learn to draw human figures, known as ymagines, flowers, foliage, vine tendrils, twisted strands, long, straight lines, square and rectangular thrones, various kinds of birds, animals, fish – in short, everything that can be seen and touched on the earth. We are not far from the concept of drawing developed by Cennini who also considered drawing from nature as the first step, even before studying the great masters.11 In the fifth part of his treatise he summarises the entire cycle of apprenticeship for painters: drawing on a tablet for a year, learning how to prepare the materials and supports for six years, studying colours, making gold draperies, learning how to work on walls for another six years, while never abandoning drawing. Talent or a predisposition for painting did not exist as far as Cennini was concerned. The science was considered to be the result of long, hard work that was always accompanied by a master. So anyone who claimed to have learnt by themselves and without discipline was considered to be lying.
The author of the Montpellier text makes explicit reference to models that had already been drawn elsewhere, which one kept after having traced them out. His description of making a tracing is quite succinct. He says to take a sheet (of parchment) that has been finely abraded, and to prepare it with linseed oil and chicken fat. It must then be coated in a circular motion or in such a way that it dries without cracking. To make a model to use as an example, a sheet must be placed over it, and the work appears like the shadow of the original. The directives provided by the monk from Mount Athos Dionysius in the 1st chapter of The Painter’s Manual entitled ‘How to make copies’ are much more detailed, like those provided by Cennino Cennini.12 The latter devotes four chapters of his Treatise to the manufacture of transparent paper. It is hard to imagine that he would dwell on this subject had it not been in widespread use in the late 14th century and probably even before that, as indicated by the LDA. He describes how to trace onto parchment or transparent paper, to obtain a copy of a head, a figure or a half-figure according to what is produced by the hand of the great masters. He adds more information that is worthy of interest, describing how to make copies on tracing paper of panel paintings, frescoes and drawings on parchment.13 These models on parchment or paper are easily transported from one site or workshop to another, and can be reused, reduced or enlarged elsewhere.
The widespread use of transparent paper has also been highlighted by sources from Westminster Abbey, informing us that the templates were models made of parchment or paper impregnated with wax. Indeed, since classical Antiquity, models have been used in the preparatory drawings for painting, especially with the mysterious ‘catagraphs’ whose invention was attributed by Pliny to Cimon of Cleonae.14 It is also noteworthy that in Book II of the LDA, devoted to painting on wood, there is also a description of cutting a stencil out of a sheet of parchment. Archaeological investigations at Meaux Abbey (England) have brought one of these to light, probably dating back to the 14th century.15
Learning to draw is accompanied in the LDA, by knowledge of the proportions of the human body, directly inspired by Vitruvius, i.e. that in the relationship between the size of the head and that of the body, the face represents one tenth of the body and the whole head, one eighth. It is clear that in the text of the LDA, as it has been handed down to us, these proportions have been modified, the face representing one eighth of the body. It is difficult to comment on how this change came about. After at least two successive copies, it may have been the result of a misinterpretation, but one could also envisage a deliberate reinterpretation according to the context in which the treatise was written. The Byzantine proportions described by the monk Dionysius of Fournar also differ from those described by Vitruvius: ‘Learn, my student, that the body is nine heads high, that is to say nine measures from the forehead to the heels’.16 The few experiments on some of the Catalonian ensembles of paintings show a great deal of irregularity in the relationship of ideal human proportions. However, these proportions are maintained throughout the work for all the figures of identical importance, generally corresponding to each of the registers of the paintings, which are often separated by a decorative frieze. As human bodies present a wide variety of shapes and proportions, we can assume that painters adapted these proportions to the subject of the work and the hierarchy of the figures.
Geometry also makes it possible to draw pictures that are proportional at much bigger scales than the human scale. Juliette Rollier-Hanselmann has begun work on this subject through the study of the monks’ chapel in Berzé-la-Ville. She emphasises that the painter created an extremely dense composition, with over forty figures in a small space (L. 2.94 x W. 3.26 x H 6.62 m.), which demanded strict internal organisation and meticulous preparation of the semi-circular surface. String was used to draw the boundaries of the register as well as the mandorla of the Christ figure. For the latter, which is located on a virtually flat surface, the length of the nose corresponds to the space between the two pupils and the radius of the first circle. The same principle was applied for the heads of the apostles, producing great uniformity of composition. The use of a three-circle system is even more apparent on the female head (in the scene of the martyrdom of St. Blaise) where the preparatory drawing shows through the worn paint layer. The middle circle, previously hidden under the veil, is visible in the gaps in the painted layer, which verifies her hypothesis.17 A study by Hjalmar Torp recalls the Byzantine principles that dictated how artists should paint sacred images. The use of models had several functions: to teach young artists the basic principles for building an image, to respect the proportions of the sacred archetypes and to give the work hidden structure, based on simple geometrical shapes.18 The realisation of an apse decoration presented a number of difficulties for the medieval painter. Curved surfaces and the height of the figures produced deformations and optical effects that had to be rectified. This made the use of geometry and suitable tools even more indispensable.
Drawing a mandorla implies not only the use of a compass system, by means of a string turning around a nail for example, but also knowing how to construct an ovoid element. This basic form, called vesica piscis (fish bladder), had been well mastered since the time of the Carolingian manuscripts and appeared to be well known by Romanesque painters (Figure 24.2). The drawing of two flamingos by Villard de Honnecourt (13th century – Paris, BnF, ms fr. 19093, fol. 18v) demonstrates the practical and relatively simple use of this geometrical shape.19 The mnemonic method of the flamingos goes back to the geometry of Euclid (c. 325–265 bc).20 The principles used by this Greek scholar have survived through the centuries and could be found in the 6th-century Latin translations of Boethius and Cassiodorus.21 In medieval decoration, we must not overlook the importance of the use of decorative friezes to structure the composition and clearly separate the various subjects depicted.22
Technical knowledge is often hidden behind the enigmatic figures of horsemen, human faces or animal figures in the lodge-book of Villard de Honnecourt: they are all mnemonic figures, which the historian and architect Roland Bechmann has deciphered and interpreted. Villard de Honnecourt’s sketches are there to remind those working on the site of geometrical constructions, when needed, in much the same way that figures of animals, objects and characters in the sky chart help to recall and recognise the arrangements of the stars in the constellations. This is what Villard underlines when he writes that they are useful for working with. In folio 38 of the lodge-book, the four nude figures serve to construct a square grid; in folio 36, the flamingos help to draw a right angle with a compass and the sheep to form a golden rectangle.23 The system of building on the basis of a square grid to be able to transpose and enlarge the drawing is nothing new. An Egyptian papyrus from the 2nd century bc, conserved in Berlin, already depicts different patterns drawn on a grid.24 Another demonstration of this can be seen in the church of San Vincenzo Galliano (11th century, Italy).25 Later, among the many technical details that appear in De Pictura by Leon Battista Alberti (1436) there is one concerning the scaled enlargement of a draft drawing using a grid system.26
As the painter acquires his skills, from learning to draw on a tablet, to constituting his own stock of traced copies, painting on parchment and then on wooden panels, he gradually prepares the way for wall painting, and the ability to change scale. The transition from producing a ‘small’ drawing to laying out a ‘very large’ decoration therefore requires specific knowledge and tools.
When we talk about drawing, particularly large-scale wall drawing, the question of tools is fundamental and warrants closer attention. The author of the LDA specifies that the objects used for wall drawing include an iron or steel tip (Figure 24.3), a paintbrush (documented in Book I of the LDA), a compass, a ruler (or rather, a rod), a string or a condermenia fectam de carta. The fabrication of a paintbrush is described and documented in the LDA, with a sketch illustrating the following text: ‘Fabricate a paintbrush with a small tuft of hairs from the tail (of a squirrel?), bind them together, cut off the base of a feather and insert them into it to give them the shape’27 (Figure 24.4). A brush might be made from the hairs of a squirrel’s tail as described, or from pig bristles as seen in Cennino Cennini.28 Several types of compasses were used at the time of Villard de Honnecourt. In his lodge-book we find a sector in which the quadrant, fixed to one of the legs, slides across the other leg, to block the compass in certain open positions and use the graduations engraved on the curved part to find the angles and proportions (Figure 24.5). Compasses with leg joints were made in variable sizes, usually on the site itself, with dividers, as a precision factor29 (Figure 24.6). The ruler (or measuring rod, virga) was a simple measuring tool by definition. It was the standard used to measure lengths on the site, it met the need for economy of materials, ease of handling, it took up little space in the workshop and on the site, and was easy to use as it did not require the ability to read. It also made it possible to draw straight lines30 (Figure 24.7). The string line was a simple, practical tool found on any site. When weighted, it served as a plumb line and defined the vertical. It was also used to draw radii converging to the centre, the joints of an arc from the centre, or circles of any size31 (Figure 24.8). We do not know the precise meaning of the word condermenia referred to previously. If used in the sense of cartoon, it could apply to the craft of stonemasonry (in which case it would be called a mole) but also to glassmaking and painting, as has already been pointed out.
This is hardly surprising when we picture life on a site in the Middle Ages and the exchanges between people working in the different trades. As we saw in the discussion on proportions, craftsmen in the Middle Ages used measures that were directly related to the morphology of the human body. The problem with such units of measurement (cubit, line, pace, span, palm, foot, etc.) was that they varied according to the place and the person using them32 (Figure 24.9).
There are several ways of making under-drawings or of marking out the surface to be painted according to the technique used. Works painted from the 12th century onwards in Catalan (or neighbouring) territories illustrate several types of under-drawings: sinopia (Figure 24.10), incisions in the lime plaster (Figure 24.11), drawn with a line (Figure 24.12), drawn in diluted ochre (Figure 24.13) but also, more rarely, in blue. Observation of some decorations shows that several techniques can be employed for the same work, depending on the desired drawing.
Clearly, one does not proceed in the same way for a fresco as for tempera. The latter does not involve the same constraints, especially in regard to time, as the surface is dry. One of the characteristics of fresco painting is that it is done in phases of a day, lasting about six hours (depending on climate conditions) that is to say the time during which the intonaco applied to the wall stays wet (intonaco being the name of the lime plaster on which the fresco is painted). Therefore, every morning, one had to start by defining the area to be painted in a day (giornata). Thus, on big sites, it was common to divide up the work, allowing several people to work simultaneously on different giornate. Every morning one had to know exactly which task to give to the men on the scaffolding. One also had to know exactly where to place the bowls of intonaco to coincide as closely as possible with the silhouettes of the subjects represented, the figures, decorative areas, etc., and to best conceal the junctions between the different giornate. The Sinopia is the wall sketch and the division of the spaces executed on the arriccio (lime plaster to which the intonaco is applied). The Sinopia was sometimes applied directly to the wall, even if the pigment used was not the red earth that gave its name to this kind of drawing. The proper function of sinopia was to serve as a guide for the later execution of the painting on the intonaco. Preparatory drawings and scale drawings are also visible on the intonaco or distemper intended to receive the painting.
An initial draft drawing had to precede the composition and execution of a wall painting. This raises two key aspects in regard to the usefulness of this reduced model. The first is to consider that most of the transactions between the patron of a wall painting and the painter himself could not have been simple tacit agreements, whether we are talking about the 12th, 13th or 14th centuries. As noted by Philippe Lorentz, the proliferation of work contracts is linked to the growth of cities in the 12th and 13th centuries, the development of independent crafts and the emergence of what we call ‘artists’ today.33 However we do not see contracts for works of art prior to the 13th century. The examples presented by Lorentz show that the assumption of a contract containing information on the programme of the work is based on an assumption. Indeed it was considered pointless to go into details about questions of iconography in a legal act. It was more a case of determining the wages of the artist and the details of execution and payment. The formal aspects of the work, based on the agreement between the patron and the artist prior to executing the work, were included in a formal document, the ‘devise’ (the division or written draft) that is often mentioned, sometimes summarised and, very rarely, copied out verbatim in the notarial act.34 When Philippe Lorentz examines the relative share of the artist and the patron in the genesis of a work, he admits how difficult this is to assess. However, in the rare cases where the ‘devise’ and the finished work have both been preserved, one can see that the patron is not necessarily the main actor.35 Moreover, the draft model of the wall decoration is not merely to meet the patron’s very legitimate desire for advance knowledge of the work to be paid for. Executing a wall painting implies preparation beforehand, with a scaled-down design of the scheme. This question is not dealt with in the LDA but the preparation of draft drawings intended for the site is confirmed by Cennino Cennini in his treatise when he says: ‘If you want to make houses, make them in your drawing in the size you wish’.36
It would be difficult to improvise painted scenes populated by a large number of figures without first thinking out the precise details of the composition, or taking the time to make a sketch. The technical, i.e. material, execution of the painting could not be improvised and only the presence of a very detailed draft drawing would have addressed the issue of the extremely meticulous organisation of the work before starting on a cycle of wall paintings, if only in terms of provision or the coordination and division of tasks. Unfortunately, although these draft drawings probably existed, they have since disappeared, which is understandable in light of the fact that they had no intrinsic value for their contemporaries, and could be scraped off and reused until they were worn away.
The wall paintings of Saint-Martin-de-Fenollar (Pyrénées-Orientales, France) illustrate the point in question. In the 12th century, this building was part of the Diocese of Elne and the property of the Viscount of Castellnou. It appears to have been owned by the Abbey of St. Mary of Arles-sur-Tech from 844. It is located on the old Roman road that passed through the fortified post of Clausures, north of the Perthus Pass. Oral tradition identifies the Church of St. Martin-de-Fenollar as the ‘Chapel of Mahut’, associated with Mahaut, daughter of Raymond Berenger III, Count of Barcelona, who inherited the land where the church is located, in 1131.
The church consists of a nave (which was apparently vaulted in the 12th century) and a rectangular chancel. The painting ensemble of Saint-Martin-de-Fenollar remains the largest painted area and the best preserved of the wall paintings of Pyrénées-Orientales. The walls of the chancel are covered with frescoes and a few fragments remain on the walls of the nave.
On the vault of the chancel, at the centre of the composition, stands a Christ in Majesty. He is depicted accompanied by symbols of the evangelists, and carried by angels. They are designated by four verses from the Carmen Paschale by Coelius Sedulius, poet of the mid 4th century (Figure 24.14). On the eastern wall of the chancel, above the axial window, the Virgin Mary appears (Figure 24.15). This head-and-shoulders image of the crowned Mary is midway between the divine space, represented by Christ and the Tetramorph on the vault, and the history of the Incarnation painted beneath it, in which she is the main protagonist. At the springing of the vault, on both sides, against a background of horizontal stripes, are the Elders of the Apocalypse (Figure 24.16). They are arranged on an intermediate register converging both towards the Christ in Majesty and the Virgin in prayer. The elders flank the chancel, walking towards its centre. This is clearly to emphasise the fundamental role of the Virgin in this iconographic cycle. The second register presents a cycle of the Incarnation (Figure 24.17): from left to right the scenes depicted are the Annunciation, the Nativity, the Annunciation to the Shepherds and the Adoration of the Magi (Figure 24.18). The lowest register depicts pale yellow scalloped draperies. On the southern wall of the nave, in the first span, two quadrupeds with clawed feet confront one another on either side of a tree, against a background of drapery. This decoration was made slightly later than the wall paintings previously mentioned.37
The historiography of these paintings has proposed a chronology that runs through the first half of the 12th century, but the research of Pierre Ponsich on the Viscountess of Castellnou has really advanced the dating of these paintings. He dates them in the second quarter of the 12th century, arguing that Mahaut could not have taken possession of her property until after her marriage to Jasper II, Viscount of Castellnou, in around 1135. At the age of about forty, Mahaut probably played an important role in the restoration of the church of Arles-sur-Tech. This is recalled by the act of consecration of the abbey in 1157.38
Parallels have been drawn between the Fenollar paintings and those of the Church of Saint-Sauveur-de-Casesnoves (in the treatment of the faces and clothing). In addition, the inscription accompanying Matthew the Evangelist at Fenollar presents the same variant of the Carmen Paschale as at Casesnoves. Elements such as these would tend to assert the existence of a common culture between Fenollar and Casesnoves, with the same desire to highlight the figure of the Virgin Mary. Casesnoves belonged to the Abbey of Saint-Michel-de-Cuxa in the 12th century, where the cult of Mary was also highly developed. This finding brings us to the idea that the influence of the culture of the prestigious monastery on the painter of Fenollar should to be taken into account.39 From this historiographical overview, we can deduce that common models circulated for some time, enriched with local cultures and that, in all likelihood, a simple vector was needed for this to spread: an illuminated book of religious inspiration could have fulfilled this function. The question of the sources from which the Fenollar painter could have drawn the elements of his wall painting is difficult to resolve. Nevertheless, the hypothesis that the Gospels of Saint-Michel-de-Cuxa could have been included in the sources is quite plausible.40 Indeed, these Gospel verses include the Carmen Paschale which has already been referred to and which existed in several copies in the inventory of the Ripoll manuscripts.
Both the decorative profusion of the painter of Fenollar and the intensity of the associated colours used in this setting are extraordinary: the angels have bicoloured wings, the nativity scene is displayed on a checkered red and white background, the Christ Child is supported on capitals formed like giant palms; the modelling of the flesh is done with thick dabs of pure tones, curvilinear red triangles denote the cheeks and corners of the nose, a red dot is placed between the eyes, there is hatching on the wrists, neck, palms and feet. A number of these features and certain ornamental recurrences can also be found in the paintings of the church of Saint-Nazaire of Hautes-Clauzes.
Technical investigations in the Factura research programme now allow us to relativise this perspective.41 The murals of Saint-Martin-de-Fenollar and Santa Maria des Hautes-Clauzes were frescoed with dry highlights but were executed using very different techniques. Although the composition of the original plaster tends to be the same in both churches, the technique used in Fenollar is very simple. Except for the highlights, it only contains two layers of preparation and one paint layer. On the contrary, the paintings of Hautes-Clauzes have two preparatory layers, covered with three paint layers. Moreover, even if they reveal certain thematic similarities, their stylistic treatments are also very different.42
The palette of pigments used has a common base but there are also some notable differences. At Fenollar, the identified pigments are green earths, cinnabar, an ochre mixture for the orange colour, the white is based on lime, the blues are lazurite and indigo, the grey is made with a yellow ochre mixed with carbon black and lime. However, in the Hautes-Clauzes, the pigments comprise a mixture of ochre and cinnabar for the red colouring, a red ochre and lazurite overlay, a mixture of lime and red ochre for the skin tones, a superposition in various places of red ochre, lime white and minium and green with a copper base.43
In regard to drawing and representation, the wall painting of Fenollar is an exemplary source of information. Markings for the lay out of the decoration were made with incisions (for the separation of registers), the protruding edge of the wall was also used in the composition, drawings were made using a string line for the hanging and the diluted ochre used for the under-drawings for some of the decoration is visible. The composition of this wall-painting cycle demonstrates the importance of the Virgin. The elders of the Apocalypse are arranged on an intermediate register converging both towards the Christ in Majesty and towards the Virgin in prayer. They are placed around the chancel, walking towards the centre of it. Their size decreases as they advance. The bands encircling the register are spaced further apart as they move away from the figures of Christ and the Virgin. The purpose underlying this form of representation is undoubtedly that of a commission calling for Christ and the Virgin to be strongly featured. The intercession of the Virgin thus becomes one of the essential elements of this iconography. While it is easy to imagine this as the desire of a literate patron, who was as product of his or her time, it would seem that the substance given to this desire, through this particular pictorial representation of the Elders and the visual effect it produces, comes under the role of the painter (Figure 24.19). By placing the scene of the Adoration of the Magi in a right-to-left direction, unlike the rest of the cycle of paintings, the painter was able to direct the viewer’s gaze to the Virgin and Child, doubly represented in the centre of the chancel, on one hand by the throne in the register of the Adoration of the Magi, and on the other, by the glory at the edge of the theophanical space.44 This dual representation and the central position it occupies reinforces the idea of the essential role of the Marian figure in the Fenollar cycle of paintings.
Painting in the context of a monumental site implies specific knowledge and skills that make the painter a privileged partner of the patron. His role in the medieval process of creating painted ensembles was probably more significant than we are prepared to admit today. The question of composition, and thus of drawing, is a fundamental aspect of the skills acquired by the painter during his training, and the testimony left by painted cycles such as that of Saint-Martin Fenollar or sources of artistic technology such as the LDA, are quite remarkable in this way. Each of the tools used by the painter in the process of learning (wax tablets), reproducing (copies, templates, grids) or laying out the decoration (tools for drawing, working out proportions, geometry, draft drawings) made him the custodian of a craft, and thus of specific knowledge that was essential in the design and execution of a pictorial programme.
