Review: Stone-washed Renoirs and the Shock of the Undone
We knew at a glance that something was amiss. On June 16th, a newspaper photograph trailed an imminent auction sale of Renoir’s “Baigneuse” of 1888. Even on the evidence of that single de-saturated newsprint reproduction (right, Fig. 1) it seemed clear that the privately owned masterpiece had gone through the picture restoration wash cycle a time (or two) too often. A comparison of Christie’s pre-sale zoom-able online photograph with historic photographs of the painting further suggested that picture conservation’s would-be beauticians had been at work with swab and solvent: Renoir’s bather had been left (Fig. 2) a paler sugar-smooth pictorially and plastically enfeebled version of her original self. (For the picture’s appearance and condition in 1944, see Figs. 7 and 9.)
Just as museum curators who organise splashy temporary exhibitions rarely broadcast the “conservation” injuries borne by works loaned from sister institutions, so auction houses, which of necessity must act primarily as agents for owners, can seem no less reticent on this fraught subject. In practice, we find that in of both of these art spheres, the “now” is often implicitly presented as the “originally-was” and “always-has-been”, thereby thwarting what would be the greatest inducement to halt needless adulterations of unique historically-rooted artefacts: a full public disclosure of “conservation” treatments and frank art-critical discussion of their material and artistic consequences. By coincidence, recent museum and saleroom activities have brought to London a slew of little-seen examples of Renoir’s oeuvre. As cases in point of Renoir’s vulnerability, we examine here treatments of his “Baigneuse” of 1888 and the Washington National Gallery’s “The Dancer” of 1874.
Renoir’s “Baigneuse” was given star billing (on a £12/18m estimate) at Christie’s June 20th Impressionist/Modern sale, for the catalogue of which it provided the cover illustration (Fig. 2). While much was made in the eight pages long catalogue entry of an impeccable and unbroken provenance through ten successive owners, not a word was said about any restorations of the painting, and although many early photographs are identified in the picture’s literature, none is reproduced. It is disclosed that this Renoir is to be included in the forthcoming “catalogue critique” of the artist’s work being prepared by the Wildenstein Institute from the Archives of François Daulte, Durand-Ruel, Venturi, Vollard and Wildenstein. (Perhaps the present condition of the picture will be discussed in that publication?)
On the night of the sale, an announcement that the picture had been withdrawn drew gasps of surprise. Artinfo reported that the vendor had accepted a private offer from an unidentified buyer for an undisclosed sum somewhere within the estimate. Trade and press eyebrows have been raised at such secretive, pre-auction sales and the withdrawal was the more confounding because expectations of a big auction house “event” had been raised by extensive (and quite stunningly fetching) pre-sale press coverage with photographs of the painting enlivened by the seemingly routine inclusion of beautiful young female staff members.
With modern paintings, the starting point for any appraisal should be the earliest known photograph. Old photographs are historic records. Historic records should never be ignored. Old photographs of pictures assembled in homes or exhibition galleries are especially precious and instructive. The photograph of Renoir’s 1905 exhibition at the Grafton Gallery (Fig. 3) testifies not only to the then generally more vivacious relative values within individual works but to the striking variety of pictorial effects and painterly means deployed within Renoir’s oeuvre.
With regard to the photographic testimony of the original appearances of individual pictures, consider first the large, near-central painting in the 1905 Grafton Gallery photograph – Renoir’s “The Dancer”. This picture, now at the National Gallery, Washington, is 138 years old but was then only 31 years old and unrestored. Then, the background was disposed in distinct but linked quadrants (top-left; top-right; bottom right; bottom left). These were not so much naturalistic renderings of an actual space as subservient pictorial devices spotlighting the central bow-tight figure of a child trainee who, through balletic discipline and artistry, had assumed a commanding Velazquez-worthy sideways-on viewer-confronting presence.
To that expressive end Renoir had welded the dramatically contra-directional lower legs into unity by a pronounced dark shadow in the vertical triangular space they bounded. That shadow sprang also from the heel of the (right) weight-supporting foot backwards and upwards in space, thereby throwing the bottom edges of the trailed skirts into relief. This dark zone in the lower-right counterbalanced another in the upper-left, which had in turn emphasised and thrown into relief the front edge of the costume, withdrawing only to leave a lighter, again relieving, tone at the dancer’s dark hair. The progressive loss through restorations of those artful dispositions (as seen in Figs. 4 & 5) and the picture’s general descent towards an inchoate, arbitrary pictorial froth that increasingly resembles the underlying condition seen today in its own infra-red imaging (see Fig. 6), is heart-breaking. Renoir had here been a sculptor before he became a sculptor, playing off forms that asserted his picture plane with others that ran sharply away from or towards it (rather as Michelangelo had famously done in his crucifixion of Haman). Degas, who spoke of Renoir’s “sharpness of tones”, had chided himself for constructing his own drawings of standing dancers from the head down instead of from the feet up. Renoir had here given a masterclass in how to project a standing figure upwards from the floor. These things artists know and appreciate.
Compendious photographic evidence suggests that restorers (frequently working myopically through head-mounted magnifying eyepieces) have consistently confounded dirt or discoloured varnish with the shiftingly elusive dark grounds used by artists to set off light-toned figures. As seen in our post of June 1st, Klimt’s portrait of Serena Lederer (which was acquired by the Metropolitan Museum of Art, New York, in 1980) has suffered in just such a manner. In the same post we saw also how Renoir’s deployment of a dark background zone in the upper left quadrant of background and a secondary but counter-balancing dark zone in the lower right quadrant of his “Dance in the City” had also been undone by successive restorers.
By courtesy of the 1905 photograph of the dancer we can now see that by 1944 the picture’s decisive tonal orchestration had already been subverted (see Fig. 3 and caption at Fig. 6). By the time of the picture’s appearance at the 2012 Frick show of Renoir’s full-length portraits (which was reviewed in our post of June 1st), the original dark tones in the lower right quadrant had effectively disappeared, leaving two odd arbitrarily truncated dark attachments to the right heel (Fig. 5). Cumulatively, this painting has suffered needless artistic vandalism of which no one speaks. The fact that graphite underdrawing is now visible on the painting has been mentioned but without any hint of alarm or censure.
With Renoir’s “Baigneuse” of 1888, the earliest photograph in our own records (- donations to ArtWatch of old photographs or postcards are always most gratefully received) is that published in 1944 when the painting was 56 years old, as seen here in greyscale at Fig. 7 (left) and at Fig. 9. Six years later, by 1950, the painting had been radically transformed, as seen at Fig. 7 (right, in greyscale) and Fig. 8 (left, in colour). The differences that emerged between 1944 and 1950 were compounded by further changes between the picture’s 1950 state (seen in colour at Fig. 8, left) and its 2012 state (seen in colour at Fig. 8, right). However many times and by whomever this painting might have been “restored”, it is clear that the resulting interventions have profoundly altered its constructional and pictorial rationales. The total extent of the alterations that occurred between 1944 and 2012 are examined right in greyscale details in Figs. 11-18. The differences between the 1950 and 2012 states are examined in colour details at Figs. 19, 20 and 21.
By 1888 Renoir had visited Algiers and Italy, come to admire Cezanne as well as Delacroix, discovered Italian painting and read Cennino Cennini’s Treatise on Painting. He had just completed an intense series of classically inspired, Ingresque female nudes, culminating in that declared trial for decorative painting, the Philadelphia Museum’s great “Bathers” of 1887, by which date he held the nude to be one of the most “essential forms of art”.
Compared with Fantin-Latour’s palpable but fluidly allegorical figure at Fig. 10, Renoir’s “Baigneuse” has, in its 1944 state, a markedly more stolid, out-of-Courbet corporeality. For all its spirited brushwork and sparkling colour, plastically, it constitutes an essay in composure, stability and parallelism. The torso seemingly rests on its own base of compressed and spreading buttocks and thighs. The thighs, knees and lower legs are held together in a manner more primly archaic (Egyptian) than classical. Movement is confined to the bather’s right hand which dries the left side of the waist. This action has enlivening consequences. The upper torso is pulled round by the right arm and the head is turned leftwards and downwards as if to contemplate the drying action of the towel. The left arm is required to be held aloft to free the left side of the figure, and, flexing at the elbow as the left hand draws across the face, it first echoes the thighs but then curls gracefully, weightlessly away in space.
What, then, explains the differences between the picture’s previous and its present condition? In such cases it is always possible to play the “Sistine Chapel Ceiling Restoration Defence” and claim that in 1944 the then 56 years old picture was very dirty and that the removal of this dirt has liberated the forms and the colours of the painting to a hitherto unsuspected degree. But the pattern of relationships that is visible, even under dirt, should not change character during a cleaning. Rather, it should emerge enhanced, with the lights lighter and the darks darker – and all individual values holding their previous positions. This has not happened – the picture has got progressively lighter with successive cleanings instead of returning to its previously cleaned state. If it really had been left by Renoir in today’s state, how could the previous but now lost artistically constructive values ever have arisen? If left untouched for the next 56 years, would anyone expect the painting to return to its 1944 appearance with the stripes on the towel and the shading of the fingers regaining strength? Would a general shading and enhancement of forms once more helpfully tuck the left hand behind the head? How might dirt have drawn more clearly and repositioned the left shoulder? How might it have more emphatically shaded the right, distant side of the face?
If we consider the difference between the 1950 and 2012 colour plates (shown at Figs. 8, 19, 20 & 21), what might account for the loss of orange coloured modelling in the left cheek, and of individual brushstrokes depicting the hair? Is it possible to claim on the evidence of these photographs that there has been a build-up of dirt on the picture over the last 62 years?
When examining the bather’s face in close-up today, as shown at Fig. 21, can we have any confidence that the paint presently surviving in that section is just as it was when left by Renoir in 1888? What kind of brush or paint might he have used that would have resulted in the present fragmentary, seemingly abraded, scattering of orange paint that lies over the blue background between the hand, the face and the shoulder?
In the next post we examine the conservation fate of more than a score of Renoirs that have been loaned from the Sterling and Francine Clark Art Institute in Williamstown, Massachusetts to the Royal Academy. We shall see how Sterling Clark learned the hard way not to trust art experts on matters of condition in paintings when, having been assured that Domenico Ghirlandaio’s “Portrait of a Lady” had never been repainted, he bought it, only to discover, very shortly afterwards, a postcard of the painting showing it in an earlier and quite different state.
Printable PDF version of this article:
Comments may be left at: email@example.com
How the National Gallery belatedly vindicated the restoration criticisms of Sir Ernst Gombrich
History has repeatedly shown that scholars and art-lovers (no matter how distinguished and mild-mannered) who put themselves between museum picture restorers and their professional ambitions, run high risks.
In 1950 Ernst Gombrich drew attention, in a Burlington Magazine letter, to Pliny’s description of wondrous effects achieved by Apelles when finishing off his paintings with a thinly spread dark coating or “varnish”. How could we be sure when stripping off “varnishes” today, he asked, that no Renaissance masters had applied toned varnishes to their own works in emulation of antiquity’s fabled painter? He received silence.
When he repeated the question in his seminal 1960 book Art and Illusion, his scholarly reputation and position as director of the Warburg Institute at London University commanded an answer. One came from Helmut Ruhemann, the National Gallery’s consultant restorer and author of its notorious “total cleaning” policy. Ruhemann insisted in the British Journal of Aesthetics that there was no evidence whatsoever “for anything so improbable as that a great old master should cover his picture with a ‘toning-down layer’.”
Gombrich returned play in a 1962 Burlington Magazine article (“Dark varnishes: Variations on a Theme from Pliny”). The discovery of a single instance of a tinted overall varnish, he suggested, would undermine the dogmatic philosophy of the National Gallery’s restorers. A dual reply came from the gallery’s “heavy mob” – its head of science, Joyce Plesters (who was married to the restorer Norman Brommelle), and the pugnacious former trustee and collector, Denis Mahon, in two further Burlington articles.
Plesters herself dismissed Gombrich on two fronts: for lacking “technical knowledge” and for displaying incomplete and misinterpreted scholarship. The entire documented technical history of art, she claimed, showed that “no convincing case” could be made for a single artist ever having emulated Apelles’ legendary dark varnishes. The passage from Pliny, she sniffed, was merely a matter of “academic rather than practical importance”. She offered to “sift” and “throw light upon” any future historical material that Professor Gombrich might uncover – should he but present it directly to the National Gallery. Her technical rank-pulling was underwritten (as perhaps was her article in part) by the director, Sir Philip Hendy, who disparaged technically ignorant “university art historians” in the gallery’s annual report.
In reality Plesters was a technical incompetent. It was she who claimed that the Raphael cartoons at the Victoria and Albert Museum were stuck onto “backing sheets” when there are none. It was she who described the large (150 cms wide) panel The Entombment, which is attributed to Michelangelo, as a single massive plank when it is comprised of three boards held by butterfly keys. It was she who counted six boards on the large panel Samson and Delilah, which is attributed to Rubens, when there are seven.
Her errors were products of a then unchecked institutional culture of technical adventurism and gross aesthetic recklessness. Great Renaissance paintings were ironed onto boards of compressed paper (Sundeala board) which today are too unstable to be moved. One such was Sebastiano del Piombo’s The Raising of Lazarus. That painting, originally on panel, had been transferred to canvas. When decision was made to re-attach the canvas to a Sundeala “panel”, technical examination identified three further “backing” canvases. When these three “backings” were duly removed it was discovered that no fourth and “original” canvas existed and that the surviving paint was attached only to a layer of disintegrating paper. But that crisis-of-their-own-making provided the gallery’s restorers with opportunity to play what Professor Thomas Molnar here called “demiurge” and improve upon the artistic content of the painting. In order to stabilise the paint layer which they had left loose and unprotected, the restorers embedded it from behind with terylene fabric attached by lashings of warm, dilute wax-resin cement. Because Sebastiano had painted his picture on a warm-coloured ground and because paint becomes more translucent with age and allows the tone of the ground greater influence on the picture’s values, the restorers decided to brighten things up and give the picture a brilliant white ground (like that of a Pre-Raphaelite painting) by adding highly reflective pigments to their own remedial wax-resin cement applications.
Plesters died in August 1996. Earlier that year, the National Gallery had published a report in its Technical Bulletin on the cleaning of two paintings by a Leonardo follower, Giampietrino. One, his Salome, had clearly suffered the Gallery’s trademark restoration losses of modelled form (see right and below), but his Christ Carrying the Cross was miraculously unscathed. Moreover, that picture was found simultaneously to display an “intensity of colour” and a restrained “overall effect” – precisely the paradoxical combination attributed by Pliny to Apelles but that had been pronounced technically preposterous by Ruhemann, Plesters, Mahon, Hendy et al.
It further emerged that Giampietrino, having first built up an “illusion of relief” with “dark translucent glazes”, had, again just as Pliny had said of Apelles, deliberately “restricted his own range of values” with a “final extremely thin overall toning layer consisting of warm dark pigments and black in a medium essentially of walnut oil, with a little resin”. Sir Ernst, nearly half a century on, had finally been vindicated but the report, inexplicably, made no reference to the dispute of the 1960s – to the very dispute which in 1985 had been described by the Burlington Magazine’s then editor, Neil MacGregor, as “one of the most celebrated jousts” ever. Had the National Gallery, having ridiculed Gombrich in the 1960s, not told him of its own remarkable technical/art historical discovery and of his own vindication? It had not. When we reported the findings in June 1996, Sir Ernst was approaching his 87th birthday. He replied:
“I could hardly have a nicer present than the information you sent me. I don’t see the National Gallery’s Technical Bulletin, and would have missed their final conversion to an obvious truth…”
Gombrich’s vindication proved a double one. Not only had the gallery discovered a technical/physical corroboration of the scholar’s astute original supposition, but the survival of a Renaissance artist’s final toned coating served further to corroborate Gombrich’s general criticisms of the gallery’s over-zealous picture cleanings. Because the two Giampietrino works were restored at the same time in the same gallery, but with the surface of the one being protected from solvent action by an ancient oil-film, while that of the other was unprotected, an unwitting laboratory experiment had been conducted on the gallery’s own “cleanings”. We can now compare the appearance of the restored but protected painting, with that of the restored but unprotected one (see right and Michael Daley, “The Lost Art of Picture Conservation”, The Art Review, September, 1999). As can be seen here, the unprotected painting (the Salome) suffered clear and dramatic losses of modelling and weakening of forms.
For a number of years after the twin Giampietrino restorations, it was possible to examine the two cleaned specimens side by side and to demonstrate the unequal effects of the treatments they had received. That is no longer possible. One of the pair has been relegated to the ill-lit basement of the reserve collection which is accessible to the public for only a few hours a week on Wednesday afternoons.
The relegated work is not the restoration-injured Salome, but the miraculously preserved Christ, the very picture which now arguably constitutes the best-preserved example of a Renaissance artist’s technique in the entire collection. This picture, which might be expected to enjoy pride of place in the main galleries, shares its new dungeon exile with another recent National Gallery Embarrassment – the Beccafumi panel painting Marcia which was dropped and smashed at the Gallery when being “de-installed” from a temporary exhibition. We had hoped and suggested that the Christ might make a return to daylight on the occasion of the Gallery’s forthcoming Leonardo blockbuster exhibition, but it seems that it will not do so – not even to join Giampietrino’s full-sized faithful copy of Leonardo’s Last Supper. (For many years, that Giampietrino mouldered in the Royal Academy’s basement as embarrassing relic of the institution’s former artistic interests.) When the last restoration of Leonardo’s Last Supper got into difficulties, the copy was taken to Milan so that full-size tracings of Leonardo’s figures might establish the limits of the restorer’s own substantial watercolour in-painting.
It seems fitting that last word be given to Sir Ernst, who died on November 3rd 2001. In another letter in 1988 he had recalled:
“I believe it was Francis Bacon who said ‘knowledge is power’. I had to learn the hard way that power can also masquerade as knowledge, and since there are very few people able to judge these issues, they very easily get away with it.”
Printable PDF version of this article:
Comments may be left at: firstname.lastname@example.org