On Place Kléber in the heart of Strasbourg stands a Classical building of 1765-78 designed by Jacques-François Blondel - the Aubette. Its facade once hid one of the jewels of Modernism, a series of spaces created using the colours of De Stijl-founder Theo van Doesburg and Dadaists Jean Arp and Sophie Taeuber-Arp, which for many years was thought to be lost. After two successive programmes of work, the main interiors have now been recreated.
The brothers Paul and André Horn, respectively the head of a building society and of an architect's office, took on the lease of the right wing of the Aubette in 1922, and decided to turn it into a place for eating and entertainment. In 1926, they gave the commission for the designs to Jean Arp and Sophie Taeuber-Arp who, faced with the magnitude of the task, linked up with Theo van Doesburg.
In February 1928, the rooms on the ground oor, the mezzanine, and the first floor were opened; in all these spaces straight lines ruled. Several months later, this rectilinear approach was balanced by an 'Arpian' curve enlivening the American bar and the dance space in the basement. But the vision of the three artists, to create an all-encompassing work of art formed from the colours of the spaces, the lighting scheme, the decoration, and even the furniture, was never fully respected. From the beginning, garlands and other trimmings appeared at the Aubette; at the end of the 1930s, Spanish-style decorations by a local painter obscured at least one room. And later followed even more depredations, which appeared to be irreversible.
It was thought that the original scheme had been completely destroyed. But research undertaken between 1977 and 1983, at the behest of Jean-Louis Faure, the head of conservation at Strasbourg's Musée des Beaux-Arts, found some traces remained.
There followed a recognition of the value of the place's heritage, and, in 1985, van Doesburg's Cinema-Dance Hall and the staircase he designed with a colour scheme by the Arps were classed as historic monuments. Van Doesburg's Salle des Fêtes (a party space) and Sophie Taeuber-Arp's foyer-bar followed in 1989.
In 1990, Faure, with the support of businessman Pierre Horn, the son of Paul, launched a programme to restore the Cinema-Dance Hall, which was completed in 1994. A second programme of work, covering the other rooms on the oor, was begun in 2002 by Fabrice Hergott, director of the museums of Strasbourg. The two operations would not have been possible without the support of the city of Strasbourg, and though this was partly financial, what was equally important was the city's firm belief that the Aubette deserved recognition at the highest level of 20th-century European heritage - a bold opinion to hold in a country where historic monuments were having a hard time.
During each of the two programmes of work, the decision was taken to nominate a scientific committee. The first included Faure, Serge Lemoine, a specialist in construction techniques;
Françoise Ducros, a historian specialising in Purism; Evert van Straaten, an expert in the work of van Doesburg; and Carel Blotkamp and Jan Leering, both specialists in Neo-Plasticism.
The size of the second committee was deliberately more limited.
It included van Straaten again, joined by Mariel Polman, an engineer who worked for the national historic monuments organisation in the Netherlands, specialising in the colours of the Modern Movement; and myself, to advise on the Dada elements of the project. Daniel Gaymard, a historic-monuments architect, was involved with both programmes.
Faced with the fact that very few of the original elements remained, both committees rejected the idea of a minimal intervention (removal of additions, revival of colours, consolidation etc. ) or of restoration, instead going for a recreation which they justified both from the historic point of view and from necessity.
It turned out that, on this project, none of the three artists had ever touched a paintbrush. The job of painting was delegated to workmen, who applied the colours in the most uniform and impersonal way possible, usually by roller. Photographs of the construction site are very revealing, showing the workers in their overalls and the artists in city clothes.
During the 1920s, among the Constructivists as well as the Dadaists, the artist's personal 'mark' was no longer considered a measure of authenticity. The concept of the work was as important as its realisation, which could easily be delegated to a third party, as long as they had the requisite technical skills. Arp, for instance, used a carpenter to make his wood reliefs, and although he painted them himself, he had said, according to Louis Aragon, that if they became dirty they could be repainted 'without troubling the artist'.
It was no accident that van Doesburg, in the issue of De Stijl devoted to the Aubette, gave a lot of space to highly technical issues, providing future occupants of the building with a kind of 'user's manual' for keeping it in condition.
Nothing more was needed to convince the two committees' members that recreation of the original would be entirely in keeping with the intentions and spirit of its three creators. This solution also had the advantage of not interfering with the original, which would remain accessible to future generations who might wish to reverse the later work. Two issues were then raised: the first to do with the support system to protect the original and act as a foundation for the new colours, and the second concerning the colours themselves.
The question of colours was, in the case of the CinemaDance Hall, more complex than anticipated, because of an unexpected problem that came to light on the east wall. Although it looked intact, it turned out to have, on two-thirds of its surface, colours that were completely different from those shown on van Doesburg's gouache sketch, which was considered to be definitive.
In addition, they did not match the black and white photographs that were taken just before the Aubette opened (for example, in the centre of the wall, there is an oblique angled shape in dark blue instead of the white drawn by van Doesburg).
On the other hand, microphotograpy of samples made it possible to establish the existence of layers of colour corresponding to those seen in the 1928 photographs beneath the visible colours.
This meant that there had been overpainting, which was subsequently also detected on the north and west walls, and the first committee had to determine whether or not this overpainting had been under van Doesburg's direction.
Having confirmed that the later layers of colour and their method of application were of a very average quality, the committee deduced that these modifications could not have been done under the aegis of van Doesburg. This argument received additional weight from a study of the style of the modified version, which seemed to the committee members to lack coherence and not to match the principles and ideas that van Doesburg espoused at the time that the project was undertaken. As a result, in its 1991 report the committee proclaimed 'unanimously the view to take as the guideline for the recreation of the Aubette, the definitive project as it had been executed in February 1928 and as it appears to exist, underlying the existing overpainting'.
Test samples were taken from the walls and sent to the Dutch paint manufacturer Sikkens which, after analysis, prepared samples of paint that were subsequently selected and approved by the committee. Each sample reflected the final colour - no glazes or undercoats were recreated. A support paper was chosen for the application of the colours. It was agreed, completely logically in line with the previous historic demonstration, to return to the original colours; only a misunderstanding, which occurred during the painting process and without the committee being informed, can explain why the final reconstitution was done with the colours being made to appear aged. For this reason visitors typically experience a moment of uncertainty - is this a dirty white or a yellow? - which prevents the space having a completely successful immediate impact.
The second committee, having taken into account the results of the recreation of the Cinema-Dance Hall, decided that paper was not an adequate support medium for the paint.
This was because its uniformly matte appearance prevented the accurate recreation of the brilliance of the original effects, and also because of the joints, which prevented a homogenous reading of the space.
Olivier Dietsch, engineer of the city of Strasbourg, and an adviser to the second committee, came up with the solution that was used. A non-acid paper was chosen to protect the original. In the party room, a combination of galvanised mesh and cardboard was placed on this, on to which a traditional plaster was applied everywhere except for the relief bands and the ceiling, which were recreated in fibrous plasterwork.
A similar solution was used for the foyer-bar, but a different approach was taken on the staircase. This had previously been covered in a thick layer of synthetic plaster, and so was already perfectly protected. It was therefore only necessary to apply a thin skim of plaster, which was made perfectly smooth.
These technical solutions made it possible to create the same kind of surface that served as a base for the original work.
In terms of colours, the choice had been made to recreate the original ones, using the very high-level research done by Polman in collaboration with the Strasbourg-based paint company Eschlimann, which specialise in the restoration of historic monuments and had been appointed for this job. In general, a similar strategy was followed for each of the three spaces: stratigraphic research, chemical tests and laboratory analysis made it possible to determine exactly the sequence of paint layers as well as the composition of the paints and the binders that were used.
Samples were prepared, entirely in line with the process used in 1928. This meant that all the same layers of paint were applied, using the same pigments and binders. Once evaluated and validated in situ (where important comparison surfaces had been opened up to this end), these samples served as the basis for recreating the feel of the original with a smaller number of layers and colours, this time of present-day manufacture.
The matching samples and the present-day recreations were compared and evaluated in situ by the committee, based on the scientific analyses, but also with regard to documentation (photographs, sketches, descriptions, knowledge of the work and theories of the artists), and through an aesthetic evaluation both of individual elements and the overall effect.
This process made it possible to establish that the three artists had very different ways of using colours and of superimposing layers of paint. For instance, van Doesburg applied two or three layers of undercoat and two or three layers of colour to arrive at the final result, showing that the definition of the colours was a laborious process and that it was necessary to make several adjustments - it was not at all a matter of simply using standard colours. On the other hand, trying to give the impression of a material that was 'entirely colour' throughout 'its mass', van Doesburg covered the red, yellow, blue and black surfaces with a very shiny rosin varnish which gave a depth and a brilliance to the colour - an effect that can clearly be seen in the photographs of the time.
Since rosin varnish has the disadvantage of yellowing severely with time, it was decided instead to apply a damar varnish which has the same qualities, if less pronounced. In contrast, according to the Arps themselves, they used far fewer layers of paint and did not apply a varnish.
The stratigraphic investigation was especially enlightening in the case of the foyer-bar, because it showed that one of the two compositions that were known through sketches had been carried out, but had then been covered over with a different final version, for which no preparatory drawings existed.
So in order to recreate this original work, the only documentary evidence came from the photographs of 1928 which were, unfortunately, inadequate. Laboratory analyses were decisive in determining the colours of almost all of the foyer-bar. Only the north-east corner, for which no paint samples had been found, posed a problem. But the committee decided that it was possible to determine the colours of this area, based on a sensible hypothesis.
On the staircase, Arp's stained-glass window was recreated exactly as it had been, following the original maquette, which was in the Strasbourg Museum of Modern and Contemporary Art; the original framing, which had had to be taken down for reasons of safety; and photographic evidence. The enamelled lights and the ventilation grilles, of which the originals had also been saved, were remade identically. This was also the case for the wooden plinths, of which a fragment had been found in very poor condition. In paying such attention to details, the committee hoped to be true to van Doesburg's concept: 'I have always tried to finish the whole as I envisaged it, in an appropriate style and to the last detail.'
The two programmes of restoration of the rooms on the first oor of the Aubette, carried out more than 10 years apart, each give the visitor a very different impression. If one takes an interest in the history of restoration, one must conclude that the unity of the original project has been destroyed and that it is now very difficult to see it as a homogeneous whole.
It would be nice to redo the Cinema-Dance Hall, but the question of finance is an insuperable barrier, unless some private donors come forward. But even in its present condition this oor does allow one to experience a space created entirely through colour, where the visitor is placed, as van Doesburg wished, 'in the painting instead of in front of it' - one of his fundamental desires.