While science has been of great service to archaeology, art history has not benefited in the same way. The technological tools that science has placed in the hands of museum curators and historians are plainly more modest than those enjoyed by the archaeologist, which range from stereoscopic aerial photography to the electromagnetic detector. Yet this is not the only reason why art history has lagged behind archaeology in terms of scientific progress. Archaeology is a field whose scientific character is jealously guarded by its practitioners, for whom it might be termed an article of faith. By contrast, the boundaries of art history as a historical discipline are poorly defined, as are those of literature and philosophy. While it is natural for philosophical thought to draw on physical data assembled by scientists, this ought to occur in a scientific, that is to say, an objective manner. For fifty years now, art history has overlapped with art criticism, which is essentially subjective and gives rise to intellectual or philosophical speculation, in which the work of art is merely the pretext rather than the object.

The notion of the ‘connoisseur’ – close neighbour to the ‘amateur’ – whose expertise is based on intuitive perception, has encouraged the rise in art history of the imagination: ‘that mistress of error and falsity’, as Pascal put it, ‘the more deceptive that she is not always so’. A form of expertise, inaugurated forty years ago by Bernard Berenson (1865–1959), is employed today – especially in Italy – in a manner suggestive of a full-blown ‘mannerist’ crisis (going by the name of filologia) in art criticism, as the latter sees its scope diminished and its object exhausted. This expertise, based on a number of comparative and historical factors, consists of bold extrapolation with a view to presenting daring hypotheses capable of surprising ‘connoisseurs’ and winning their applause for the magician’s brilliant conjuring trick. This according to the principle that the further apart the poles, the more spectacular the spark. So it is, for instance, that Dosso Dossi’s (1490–1542) hand is detected in Raphael’s (1483–1520) Transfiguration. Art criticism construed in this way becomes a kind of intellectual game played by a few initiates, far removed from the crowd which, for its part, wants it to be a form of ‘literature’. The art essay, made famous by great writers such as André Gide (1869–1951), Paul Claudel (1868–1955) and André Malraux (1901–1976), has indeed given rise to a ‘literary genre’ that permits lesser minds to dispense with all scientific rigour, so as to furnish the public with abstruse issues that substitute for depth, or alternatively romanticised tales that can be grasped by the greatest number; who, by the way, barely read them, instead requiring the image to speak for itself.

This ‘grasping’ of art through vision alone is based on an intuitive perception of the work of art which, we must confess, is more developed among 15-year-olds (in France at least) than it was thirty years ago among their adult parents, even those who were art historians. While this will help refine the intellectual faculties, it will also lead, if we are not careful, to a considerable degree of intellectual laziness. It will extend culture and lower its level, with the viewer no longer making the slightest effort to understand the historical and intellectual circumstances on which the creation of works of art was contingent. These works, no longer situated in space or time, will be perceived as a kaleidoscope of forms and colours with a meaning of their own. It is already the case in the United States that the crowds who pour into museums do not know which paintings they admire, to the extent that curators have had to label the works with the artist’s name.

It would undoubtedly be an exaggeration to argue that the immense success of art in the modern world has paralysed the development of science or that the art book – especially in France – is in the process of killing art history. Any publisher worth its salt should be able to combine the collection of reproductions required by contemporary tastes with a noteworthy scientific synthesis – a useful summary of current scholarship, which will at least vouch to the casual purchaser for the value of the illustrations. It is not unreasonable, however, to state that it is harder nowadays than it was twenty years ago to publish a strictly scholarly analysis on a subject lacking in ‘broad appeal’. The commercial success enjoyed by the art book has subjected it to the laws of supply and demand, and hence to fashion, which has placed publishers under greater restraint than before. While it is possible to publish four books a year on Van Gogh (1853–1890), three on Gauguin (1848–1903) and as many as you please on Picasso (1881–1973), I defy any art historian with a manuscript on an ‘unpopular’ painter such as Poussin (1594–1665) or Claude Lorrain (1600–1682) to find any takers (in France, at least). In 1925, by contrast, he would have had little difficulty. The efforts of young people wishing to become art historians are singularly constrained by this. Purely scholarly publications therefore have to be subsidised by private benefactors or official bodies. Archaeology is more favoured than art history in this respect, as it benefits from institutes, foundations and specialist publishers which have a tradition of facing down ‘fashion’ and are rightly proud of their limited circulation.

In preparing the publication of this book, I was especially pleased to re-establish contact with the dynamic publishers A. & J. Picard who, twenty-five years ago, produced the substantial volume on Mont Saint-Michel that was to be the cornerstone of my own career. Nor did they hesitate to support the magazine L’Amour de l’Art, which René Huyghe (1906–1997) and I produced in the ten years prior to the last war as a kind of laboratory devoted to the study of artistic forms and civilisations.

Jacqueline Marette’s book – supported by the Centre National de la Recherche Scientifique and judiciously edited by Jacques Picard – which we present to the public today, deals with a previously unpublished subject that may justly be described as ‘scientific’. Having grown out of a dissertation prepared for the École du Louvre, it aims to study the painting of the ‘primitives’ from a very specific angle, namely that of the wooden supports they used for their paintings. Her analysis of wood types and craft techniques draws on the same statistical and cartographic techniques that have proved so successful among archaeologists, and which art historians are also slowly beginning to employ. These techniques can yield surprising results with the potential to overturn quite a few accepted ideas. Colette Caubisens-Lasfargues used this method in a dissertation for the École du Louvre, in which she demonstrated that, contrary to widespread belief, history painting was very much in the minority in French salons during the revolutionary period, compared to genre paintings and works on realistic themes. Similarly, by publishing the estate inventory of Mathieu Le Nain (1607–1677), Georges Wildenstein (1892–1963) has been able to conclude that the artist chiefly painted portraits and religious scenes and only rarely the genre works with which he has come to be associated. 1‘L’Inventaire après décès de Mathieu Le Nain’, 1677. Gazette des Beaux-Arts, April 1955, pp. 198–210.

The statistical results Jacqueline Marette presents here lead us to conclude that locally sourced wood was used systematically during the studied period. This was not the case later, when the use of wood became a matter of choice and ceased to be determined by a tradition born from the fact that easel painting was originally intended as a substitute for precious metalwork. The newfound availability of exotic woods provided these later artists with a new material, the harmonious tones of which were used to great advantage by the Dutch in particular, by means of ground layers and simple glazes – a technique that Monticelli (1824–1886) would revive to brilliant effect in the 19th century.

The data Jacqueline Marette has assembled is such that historians of ‘primitive’ painting will in future be obliged to perform a careful analysis of the support, its material, and its construction. Analysis of this nature has the potential to settle issues of attribution or at the very least to provide useful clues in cases of uncertainty. Any historian studying the enigmatic Man with Wine Glass in the Louvre, for instance, cannot now ignore the fact that it is painted on a walnut panel – a type of wood of which Jacqueline Marette did not find a single example north of the Loire.

It is my wish that this book will draw curators’ attention to the importance of the support to the painting as a historical document. It has to be admitted, sadly, that until now only the painted surface bearing the image has been considered interesting in most cases. To conserve that ‘skin’, one has not hesitated – even in certain European countries – to subject the support to surgical procedures that alter it fundamentally or even cause it to disappear. It is standard practice in the American art trade to cradle a panel or, better still, to transfer it onto a new panel – preferably of a precious type, such as mahogany, itself cradled – to satisfy the desire of a clientele for whom an art object should appear tidy, both front and back. Signs of age or wear in a painting are interpreted by this clientele as indicating a lack of care and hence a lesser value. Even paintings on canvas have been remounted on cradled panels to give them this supposedly attractive look. 2This can cause the painting to deteriorate, as I have seen in the case of a canvas from Van Gogh’s dark period, which the Louvre purchased in New York, and which had been remounted and cradled. In similar fashion, the State Hermitage Museum (St. Petersburg) systematically transferred its panels to canvas in the 19th century. Radical measures of this kind, which affect the integrity of the object and deprive historians of precious data, ought only to be used as a last resort, when there is no other means of saving the paint layer. Where such action is genuinely unavoidable, the altered or destroyed support should be given a precise descriptive analysis. Panels that have retained their original supports ought to receive specialised care and, above all, to be spared having to travel for exhibitions, other than with the expectation of valuable scientific results.

Some years ago, while studying the work of Petrus Christus (ca. 1425–1475/76), 3See Germain Bazin, ‘Petrus Christus et les rapports entre l’Italie et la Flandre au milieu du XVe siècle’, Revue des Arts, 1952, no. 4, pp. 194–210.several clues led me to believe that this Flemish painter, influenced by Antonello da Messina (1430–1479), must have spent time in southern Italy, having possibly come from Spain. The Death of the Virgin – recently purchased at the time by the San Diego Museum of Art and thoroughly infused with Antonello’s spirit – does indeed have a 16th-century Sicilian provenance, and so knowing the type of wood on which it was painted could have provided me with useful information. The painting had, alas, been remounted on canvas! Mr. Hentschel, head of Knoedler art dealers, who had sold the painting to the museum in San Diego, was nevertheless able to inform me that the support had been made of softwood. He also furnished me with another, extremely valuable detail, namely that there were signs of gluing on the back of the panel, consisting of fibres and adhesive: a procedure I then believed to have been used solely by the southern European schools. Jacqueline Marette, by contrast, has only detected this technique in Germany and in Spain (which was strongly influenced by Germany). Sicily belonged to the Spanish sphere of influence at the time, so the San Diego Death of the Virgin might have been painted on a panel prepared in Spain or Italy. However, to take this line of reasoning any further would require careful analysis of the lost panel (to determine, for instance, whether fibres were used on the front, painted side, and how the panel was assembled). We would also have to analyse 15th-century Sicilian panels and surviving Petrus Christus supports. 4Especially the eight which, out of around a total of thirty panels, originated in Italy or Spain. This substantial and purely material research effort would enhance our knowledge of this artist more than any critical speculation. The stakes are high, since the question touches on the mystery of when Van Eyck’s style and technique – in other words, oil painting itself – were introduced to Italy. Sadly the massacre of original material that has been inflicted on paintings over two centuries means that such research can now only be performed in a very fragmentary manner.

Jacqueline Marette’s work will already be a noteworthy achievement if it succeeds in drawing attention to the importance of these issues. Let us hope it encourages similar research, which will increase the number of panels studied and further confirm her statistics. All panels that have retained their original supports and are housed in museums ought now to be inventoried with a view to their study. The manner in which museum curators in all European countries (with one predictable exception 5This has, unfortunately, affected an entire school.) and in the United States responded to Jacqueline Marette’s requests for analysis is positive proof of the interest her research has aroused and a guarantor of its scientific value.

Germain Bazin

Curator in Chief of Paintings and Drawings du Louvre