Artwatch UK

Posts tagged “Jean-Pierre Cuzin

Rocking the Louvre: the Bergeon Langle Disclosures on a Leonardo da Vinci restoration

28 April 2012

ArtWatch has been haunted for two decades by a nearly-but-not-made restoration disclosure. In the 1993 Beck/Daley account of the Nippon TV sponsored Sistine Chapel restoration (Art Restoration: The Culture, the Business and the Scandal), we reported that in the late 1980s Leonetto Tintori, the restorer of Masaccio’s “Trinity” in the Santa Maria Novella, Florence, and a member of the international committee that investigated the controversial cleaning, had “urged the Sistine team privately to preserve what he termed ‘Michelangelo’s auxiliary techniques’ which in his view included oil painting as well as glue-based secco” (p. 111). What we had not been able to say was that Tintori (who died in 2000, aged 92) had prepared a dissenting minority report expressly opposing the radical and experimental cleaning method.

Shortly before the press conference called to announce the committee’s findings, Tintori was persuaded by a (now-deceased) member of the Vatican not to go public with his views. He was assured that his judgement had been accepted and that what remained on the Sistine Chapel ceiling of Michelangelo’s finishing auxiliary secco painting would be protected during the cleaning. With a catastrophically embarrassing professional schism averted, the restoration continued and the rest of what Tintori judged to be Michelangelo’s own auxiliary and finishing stages of painting was eliminated. Without knowledge of Tintori’s highly expert dissenting professional testimony, the public was assured that despite intense and widespread opposition the cleaning had received unanimous expert endorsement. Critics of the restoration were left prey to disparagement and even vilification.

On January 4th, we noted that in the widely reported schism that emerged at the Louvre with the resignations of Ségolène Bergeon Langle, the former director of conservation for the Louvre and France’s national museums, and, and Jean-Pierre Cuzin, the former director of paintings at the Louvre, from the Louvre’s international advisory committee on the restoration of Leonardo’s “Virgin and Child with St. Anne”, it had been recognised that the resulting crisis of confidence was of a magnitude not seen since the Sistine Chapel controversy. Restoration advisory committees are not imposed on museums and customarily they serve as political/professional fig leaves. In the wake of the Louvre committee resignations, embarrassed and perhaps panicky members of the museum’s staff offered self-contradictory and unfounded assurances (see below). In January, the Louvre’s head of painting, Vincent Pomarède reportedly claimed that “The recent cleaning was absolutely necessary for both conservation and aesthetical reasons.” This assurance proved unfounded on both grounds. Pomarède added that no member of the committee “has ever said that the cleaning was not prudent and had gone too far technically.” One has now done so – publicly – and left museum restorations under an unprecedented spotlight.

During an earlier cleaning controversy at the Louvre, Edgar Degas threatened to produce an anti-restoration pamphlet that would be what he termed a “bomb” – but he never did so, so far as we know. Now, as Dalya Alberge reports in the Guardian, the French Le Journal des Arts yesterday published an interview with Ségolène Bergeon Langle of truly momentous if not incendiary consequence (see below). We learn that her resignation came after no fewer than twelve letters requesting information on the restoration’s course went unanswered; that it was made in specific and pointed protest against the use of retouching pigments whose safety had not been proven; and, that the Louvre’s public claims of some pressing conservation need to remove the varnish were false, having been made despite it being known within the museum that any potential threat to the paint came not from the varnish but from a single faulty board in the picture’s panel which was reacting to the museum’s insufficiently stable environmental conditions. Perhaps most disturbingly serious for art lovers are Bergeon Langle’s disclosures that along with old (but nonetheless still protective) varnishes, original material of Leonardo’s was removed – against her advice – from the painting; and, aesthetically, that it is confirmed that the modelling of the Virgin’s face was weakened (see Figs. 1 and 2; and, for weakening to the modelling of St. Anne’s face, Figs. 12 and 13).

That the Louvre authorities would not inform even so distinguished a member of its own advisory committee might suggest either that the restorers had not known in advance what they would be doing to the painting; or, they feared that disclosure of their intentions would provoke opposition within the advisory committee. Either way, this was clearly an unacceptable (if not improper) way for a museum to execute irreversible alterations to one of Leonardo’s most advanced sophisticated, complex and problematic works. To Bergeon Langle’s now public “insider” criticisms, additional detailed material to highlight further Louvre procedural shortcomings and misrepresentations to the press and to the public will shortly be presented by Michel Favre-Felix, the president of the Association Internationale pour le Respect de l’Intégrité du Patrimoine Artistique (ARIPA). Favre-Felix is also to call formally for the establishment of a national scientific ethics committee that would be independent of museums and their restoration teams and be charged with re-examining the conservation file on the challenged St. Anne restoration.

A second member of Louvre’s advisory committee, Jacques Franck, the world authority on Leonardo’s painting technique, has said to the Guardian that a restoration likely to generate such disapproval from leading figures should never have been undertaken in the first place and, given that Ségolène Bergeon Langle is unquestionably France’s highest authority on restoration matters, her alarmed protest is therefore one that should mean a lot to both Leonardo scholars and art lovers the world over.

Unfortunately, the restoration-induced changes on the St Anne are not unprecedented. It is Art’s general tragedy that while scholars have quietly enlarged the oeuvre of Leonardo over the last century and a half, restorers have repeatedly swabbed and scritched away at the surviving fabric of those precious works – sometimes to an astounding degree, as with the “Last Supper” in Milan. With the National Gallery’s substitute version of the “Virgin of the Rocks” we have seen how the distinctive Leonardesque expression on the angel’s mouth was altered (without any acknowledgement) despite the fact that a distinguished scholar and former director of the Gallery, Kenneth Clark, had seen the angel’s face as being “the one part of our Virgin of the Rocks where the evidence of Leonardo’s hand seems undeniable, not only in the full, simple modelling, but in the drawing of the hair”. It is a matter of note that four of the most enthusiastically supportive members of the Louvre advisory committee were drawn from the curators and restorers who were directly responsible for the London and Milan Leonardo restorations.

Of Leonardo’s accepted earlier paintings, in 1939 Kenneth Clark lavished especial praise on the treatment of modelling found on two portrait heads – and in his enthusiasm, he awarded the palm of best preservation to both of them. The “Ginevra Benci”, then in the Liechtenstein Collection but now in the National Gallery of Art, Washington, was judged “the best preserved of all Leonardo’s early pictures”; one that “shows most clearly his intentions at this period”; and, one where “within the light oval of the face there is very little shadow, and the modelling is suggested by delicate gradations of tone, especially in the reflected lights.” Clark thrilled to the great refinement of execution: “We see a similar treatment of form in Desiderio’s low reliefs, controlled by the same sensibility to minute variations of surface. There are passages, such as the modelling of the eyelids, which Leonardo never surpassed in delicacy, and here for once he seems to have had none of that distaste for the medium which we can deduce from his later paintings, no less than from contemporary descriptions of his practice.” Ever aesthetically alert and deft, Clark saw all of these ultra-refined technical devices as being entirely “subordinate to the feeling of individual character with which Leonardo had been able to charge his portrait, so that this pale young woman has become one of the most memorable personalities of the Renaissance.” (We are grateful to Carroll Janis for drawing attention to this passage.)

Clark’s alertness to the physical/aesthetic characteristics of Leonardo’s hand was to the fore in his reflections on the “Portrait of a Musician” at the Ambrosiana in Milan. In the “subtle luminous modelling” of its head and its “delicate observation of light as it passes across the convex forms”, this work could only be “by Leonardo’s own hand alone and unaided” and it was “very similar to that of the angel in the Virgin of the Rocks”. As it stood before 1939, this too was “perhaps the best preserved of Leonardo’s paintings”, and in it we were then able to “learn something of his actual use of pigment, elsewhere obscured by dirty varnish, and we see that it was less smooth and ‘licked’ than that of his followers.”

Ironically, Clark, with his pathological aversion to “dirty” varnish (which is to say, old varnish on an old painting on an old support), was more responsible than anyone for the subsequent museum restoration/stripping mania. Looking around today’s museums, it is hard not to conclude that Clark might have been more careful in his wishes. Bergeon Langle’s warning against the modern addiction to penetrative imaging systems is particularly apt and timely: the hyper-active restoration changes (see right) made to the modelling and to the expression of those precious living Renaissance faces have cumulatively thinned and abraded pictures surfaces and material components and thereby remorselessly pushed great paintings into sad resemblances of their own infra-red under-states (see particularly, Figs. 4-11 and 19 & 20). Technical curiosity kills more than cats. In the case of Leonardo it has contrived to pull that artist back from his own increasingly lush highly-wrought subtly atmospheric shading towards the brilliant but thinner decorous linearity of Botticelli, when any comparison of the “Mona Lisa’s” hands with those of Leonardo’s “Annunciation” would have warned precisely against such perverse and regressive adulterations.

The interview given to Le Journal des Arts of 27 April, by Ségolène Bergeon Langle read as follows:

Why resign from the Louvre’s scientific advisory committee for the St Anne? “In January 2011 the committee had agreed on a gentle cleaning of late varnishes and the removal of the stains on the Virgin’s cloak. Yet, between July and October 2011 a more pronounced cleaning was done and presented as ‘necessary’, which I objected to. I was then faced with people who would oppose my position, which is technical and not based on aesthetics. My 12 letters [to the Louvre] asking for precisions on some aspects of the cleaning process and on the materials to be used for retouching, remained unanswered. I had to resign (on December 20th, 2011) to be heard just on one specific point: the Gamblin retouching pigments were not to be used since their innocuousness is not proven. Right from the beginning, false ideas have been put forth, like calling ‘repaints’ original retouches by Leonardo in the work’s early stages, or to attribute flaking in the paint layer to the ‘contracting varnish’, a [consequence that was] actually due to the sawing up of the wood [panel]…”

What do you think of the work done? “In my opinion, the precautionary principle hasn’t been respected. We must face the fact that the Virgin’s face is less modelled now. The cleaning should never have gone so far. However, I was happy that the grove [of trees] be preserved and, also, the ground’s material constituents that some ‘felt’ not original (though between January and April 2011 a brown-greenish section of the ground, located below St Anne’s elbow had been removed already). Besides, another matter of much controversy, the whitened layer on Christ Child’s body, has mistakenly been understood as a late varnish [that has] gone mouldy. I’m inclined to believe it was an irreversibly altered [original] glaze and, therefore, I have recommended that it be preserved, but nobody would hear me.”

The current Leonardo exhibition implies that his other paintings in the Louvre should be cleaned also. How do you feel about that? “Just not to do it, by all means! The original flesh paint in the St John-the-Baptist, being rich in oil, displays a significant network of drying cracks and might be fragile in the event of cleaning. For sure, scientific methods are essential but they need sound interpretation and wisdom dually… To date, there is too much boldness originating mistakes and an alarming fascination for infra-red investigation whereby are revealed under-layers that were never meant to be seen.”

Michael Daley

Comments may be left at:

Printer-friendly PDF version of this article

Above, Fig. 1: The Virgin (detail) from Leonardo’s “Virgin and Child with St. Anne”, before restoration.
Above, Fig. 2: The Virgin (detail) after restoration.
Above, Fig. 3: Left, the short Louvre catalogue, published in 2012 after the restoration; right, a plate of the same heads published in 1992.
Above, left, Fig 4: Leonardo’s “The Musician” as published in 1945. Above, right, Fig. 5: an infra-red photograph of the musician published in 2011.
Above, Fig. 6: The musician, as in 1945. Above, right, Fig. 7: the musician, as published in 2011. By any optical appraisal, it can be seen that Leonardo’s painting presently stands somewhere between its 1945 self and an infra-red photograph of itself.
Above, Fig. 8: The musician, detail, as recorded in 1945.
Above, Fig. 9: The musician, detail, as found in 2011.
Above, Fig. 10: The musician, detail, as recorded in 1945.
Above, Fig. 11: The musician, as found in 2011.
Above, Fig. 12: The eyes of St. Anne, in Leonardo’s “Virgin and St. Anne”, before its cleaning at the Louvre.
Above, Fig. 13: The eyes of St. Anne, as found after the picture’s cleaning at the Louvre.
Above, Fig. 14: The eyes in Leonardo’s “Ginevra Benci”, as seen in Bode’s 1921 Studien über Leonardo da Vinci.
Above, Fig. 15: The eyes of “Ginevra Benci”, as found in 2011.
Above, Fig. 16: Andrea del Verrocchio’s “Flora”.
Above, Fig. 17: “Ginevra Benci”, detail, as seen in 1921.
Above, Fig. 18: “Ginevra Benci”, detail, as seen in 2011.
Above, Fig. 19: The musician, detail, as found in 2011.
Above, Fig. 20: The musician, detail, as found in 2011.
Can all the photographs in the world be wrong?
Might anything ever count as a fair demonstration of a restoration-induced injury?
Can no curator or trustee appreciate the inherent physical dangers when allowing restorers, who work with sharp instruments and highly penetrating solvents from the top down, to act upon pictures which artists have built from the bottom up in order to leave their finest and most considered effects exposed at the picture’s surface? Can no one in authority appreciate that every authorised restoration is an accident waiting to happen?
Does no curator ever wonder what has happened to eyebrows and the shading around eyes – and mouths, and nostrils – when pictures are “cleaned” or “restored”? Does no curator appreciate the vital function that shading serves for artists who are attempting to capture from nature, or to evoke imaginatively, a precise and specific personality, state of mind, engagement with the world? Does no curator recognise the tell-tale signs when restorers subvert artistically conjured forms and change the expression on subjects faces?
Would Kenneth Clark, if he were alive today, still consider “Ginevra Benci” and “The Musician” to be Leonardo’s best preserved works – and if not, why not? In the art trade it is recognised that the best preserved works are those that have been preserved least often by “conservators” and “restorers”. Why do people who are charged with protecting art in within the museum service so often take a contrary view? What supports their apparent belief that a much or a radically restored work may count as a “best preserved” specimen?
They all use the words freely, but do any Leonardo scholars, or Leonardo exhibition organisers, truly comprehend the vital conceptual connection between an artist’s system of illusionistic shading and the forms that sculptors literally build? Are any scholars prepared to discuss the manifest changes to Leonardo’s works that emerge in each successive monograph? The elephant in the art restoration room is this: while photography and book reproduction methods improve ceaselessly (see in particular the excellent and instructively enlarged photographs in Giovanni Villa’s Leonardo da Vinci – Painter, The Complete Works), authors themselves habitually refrain from discussing the nature of the often profoundly altered states to which their photographs testify. Ségolène Bergeon Langle, a conservation scientist, has bravely lifted the lid. Will others now discuss what lies below?
Click on the images above for larger versions. NOTE: zooming requires the Adobe Flash Plug-in.

Another Restored Leonardo, Another Sponsored Celebration – Ferragamo at the Louvre…

8 April 2012

The penny dropped last week in Paris: museum picture restoration is becoming a money-making machine in which the artistic sums may not necessarily add up. The Louvre’s restored Leonardo “Virgin and Child with St. Anne” re-appeared in a series of openings for a swish exhibition-of-celebration, “La Sainte Anne l’ultime chef-d’oeuvre de Léonardo de Vinci”, sponsored by the Italian fashion house Salvatore Ferragamo (whose 2012 fashion show is to be held within the Louvre). The official defences of the restoration are found in the exhibition literature and in a DVD film (“Leonardo de Vinci The Restoration of the Century”) celebrating “The spectacular operation, the likes of which occurs only once a century”. Although there may be a touch of “The Official” in the film, acknowledgement is certainly made of opposition to the course of the restoration that came from within the advisory committee itself. Jean-Pierre Cuzin, the former director of paintings at the Louvre, is seen to speak with great eloquence on the option of, essentially, leaving well enough alone. Reference is also made to wider opposition that was reported in what is described as “a virulent press campaign”. The organisation of the exhibition itself is seen to testify to the rapid growth of mutual support systems within the international museum community. At the same time, we can now better gauge the restoration’s artistic consequences and better appreciate why two eminent art authorities, Ségolène Bergeon Langle, the former director of conservation for the Louvre and France’s national museums, and Jean-Pierre Cuzin, resigned from the restoration’s international advisory committee.

As with the Credit Suisse sponsored exhibition “Leonardo da Vinci – Painter at the Court of Milan” which gathered £1.5bn worth of Leonardos in celebration of the National Gallery’s restored “Virgin of the Rocks” (see Figs. 17 & 18), this Ferragamo/Louvre exhibition has drawn masterpieces by Raphael, Michelangelo, Pontormo, Andrea del Sarto and others. The National Gallery (two of whose staff members served on the St. Anne restoration’s international advisory committee) loaned its hitherto unloanable Leonardo “Burlington House Cartoon” in a straight swap for the Louvre’s earlier loan of Leonardo’s original “Virgin of the Rocks” to the London Leonardo-fest. It would seem that in this international bonanza, one restored Leonardo begets another and each begets a plumply lucrative exhibition and catalogue. The current escalation in travel and restoration risks for works of art is terrifying.

The Louvre’s current exhibition is said by the curator, Vincent Delieuvin, to comprise “a science workshop”. But this “workshop” could not inform the treatment of the painting because it followed, not preceded, the restoration. Moreover, the exhibition itself imposed a guillotine on the restoration. Members of the international advisory committee who wished for more tests, for more consideration of vexing issues, felt thwarted by the Louvre’s need to finish the restoration in time for the arrival of the stupendous borrowed treasures. The cumulative assembled testimony of the exhibition’s many borrowed copies and derivatives of the “St. Anne” might well have been instructive, but, not having been seen, it finds no reflection in the restored Leonardo which artistically has pulled away from its own off-spring (see Figs. 13-16). Delieuvin’s reported twin claims that the restoration “is a true resurrection of the ‘St. Anne’” and that “The painting has recovered a depth and a relief almost like sculpture, with an intense palette of lapis lazuli blue, lacquer red, grays and vibrant browns”, seem both rather tastelessly hyperbolic and at variance with visual evidence (- see right).

Certain structural stresses in the over-heating art economy have become visible. At the exhibition’s epicentre the “Burlington House Cartoon” and the “St. Anne” (for which picture the drawing was a study), have been brought together side by side in a spectacular but counter-productive coup de théâtre (see Figs. 13 & 14). The drawn study, now discoloured but sombrely potent in a magnificently worthy black frame, conjures a breath-taking orchestration of monumentally poetic forms, forms that rightly have been seen to rival the pedimental female groupings of the Parthenon sculptures. Since the Second World War there has been no drawing in existence to rival this fragile and brittle manifestation of the grandeur of Leonardo’s thoughts. (If lost – and in recent years lorries have been burnt-out in the Channel Tunnel and a ferry and its lorry cargo was lost in the Channel – no insurance money or state indemnity could acquire another of its kind.) In contrast, the restoration-weakened “St. Anne”, with its now arbitrarily floating, obtrusively abstract and glitzy lapis lazuli blue drapery, has departed from its formerly-realised self, as the adjacent cartoon and the exhibition’s many derivative pictures mutely testify. To see strong colours subsumed within tight sculpturally integrated groups, we must now look to derivatives of Leonardo rather than to their progenitor (see Figs. 15 & 16).

As if to inoculate the exhibition viewer against this back-firing juxtaposition, the wall immediately opposite the cartoon and the “St. Anne” carries a portentous notice headed “A fundamental restoration”:

A fundamental restoration of Leonardo da Vinci’s St Anne was initially envisaged in the 1990s when a few quite conclusive cleaning tests were carried out.

The picture’s dull appearance, its hues discoloured and distorted by numerous repaintings of the sky and the Virgin Mary’s blue mantle, demanded the intervention that finally began in 2009. Minute bulges, very probably caused by the stress exerted on the picture layer by the hardening of old restoration varnishes rendered the restoration inevitable.

Preceded by an exceptional series of preliminary examinations and scientific imagery analyses carried out by the laboratories at the Centre de recherche et de restauration des musées de France (C2RMF) and generously financed by Mr Barry Lam, the restoration itself began late in 2010. The restorer, Cinzia Pasquali, was chosen following an invitation to tender and worked for more than a year at the C2RMF in the painting workshop in the Pavillon de Flore.

The restoration comprised two principle problems: the removal of the discordant repaintings, some resulting from very ancient and thick accumulations of retouches, the thinning of restoration varnishes oxidised and deteriorated by too many partial cleanings, the moving of excess varnish from one area to another using solvents, retouches and refixings down the centuries throughout its long history, the picture had obviously been devarnished and revarnished many times but fortunately the picture layer had been sufficient robust to resist this. The extremely irregular and oxidised state of the surviving varnishes distorted all the tints and, by a well-known physical effect, ‘decolourised’ and yellowed the original hues.

The gradual thinning of these varnishes to a uniform level was therefore the restoration’s major challenge. During this process, resin analyses and measurements of varnish thickness, conducted by new techniques developed at the C2RMF enabled an extremely precise approach to the thinning, which had to be undertaken delicately, both to preserve a degree of patina on the picture and protect the painting itself from any contact with the solvents used. This extremely gentle cleaning process revealed a painting in vivid colours and resuscitated the splendid lapis lazuli blues and refined violet reds and crimson kermes gum lacquer.”

In this classic museum PR conflation of aesthetic and conservation “needs”, we are variously told that aesthetic changes had been necessary on urgent conservation grounds; that the restoration was “envisaged” some time ago and that this aspiration had been reinforced by the picture itself whose dull appearance “demanded” a restoration. Meeting this demand from the inanimate is said to have been made “inevitable” by a mysterious conservation ailment in the form of “minute bulges” which “very probably” were being caused by the varnish itself.

“Very probably” is a distinctly weasel-phrase and seemed the more so because Ségolène Bergeon Langle had very recently pointed out that the minute manifestations were confined to a single board (which had been badly cut when first made) within the panel, and therefore could not have derived from overall varnishes which some were itching to remove. This analysis of the actual cause was accepted on the DVD film where it was claimed that the restoration had had to proceed because of “lifting due to contraction of the wooden panel”. That raised the larger restoration question: if the varnish was not causing the lifting, was there any conservation reason for removing it at all? A frankly negative admission on that point would, of course, have greatly strengthened the position of the “moderates” on the advisory committee who were, under any circumstances, urging restraint and caution. We now hear that not only is it admitted that the liftings of paint were indeed caused by this plank, but also that they had easily been repaired locally. In hope of preventing future liftings, the panel painting is to be encased in framing that will incorporate a suitable micro-climate designed to stabilise the offending wood. On the face of it, this is good news but, in today’s museum practice, a risk removed often seems to make space for another to be incurred. And, sure enough, we also hear it is now thought that, with the provision of its own micro-climate, this great picture can be regarded as safely peripatetic – and that as such it is to be despatched in the first instance to an annexe of the Louvre at Lens, in northern France. But then where next – Tokyo? Dubai? California? And on what tariff? Perhaps in addition to adding this “restoration of the century” to our list of cleanings sold on misleading conservation-necessity prospectuses, the picture should also be put on our Now At Grave Risk of Travel Injuries category? We trust that the Louvre authorities will amend their misleading wall notice on the restoration.

The material on the picture’s surface is said to have been the accumulated product of many and various previous restorations (some with caustic substances), throughout all of which Leonardo’s original paint had suffered no injury. What chance, therefore, could the last restoration’s highly advanced, “extremely precise” techniques have produced anything other than an “extremely gentle” cleaning? Well, first of all, the proof of the pudding is in the appraising of the result – see right. Second, it is never wise to take restorers’ own prognoses at face value. Errors can occur at any point of the restoration process. The suggestion that a uniform layer of varnish had been left in place might surprise members of the international advisory committee who had been under the impression that varying thicknesses of varnish would be left in place according to specific needs for caution (as with the especially vulnerable face of “St. Anne” – see right).

Further, questions arise in terms of conservation methodological practice. In restorations, paintings are stripped down and then reassembled by repainting. Where, then, are the detailed photographic sequences showing the painting before cleaning; after cleaning but before retouching; and, after cleaning and retouching? Without such hard visual documents the path of the restoration cannot be retraced. It was only when the National Gallery kindly provided such photographs that we were able to identify an unacknowledged change that had been made to the angel’s mouth in the London “Virgin of the Rocks” (see comments at Figs. 17 & 18).

In the two versions of the Louvre exhibition catalogue (one of 52 pages at 8 Euros and one of 448 pages at 45 Euros) there are not even any facing images showing the picture before and after restoration. Such a pairing is found in the (excellent) Beaux Arts special “Léonard de Vinci – Les secrets d’un génie” at 6.90 Euros (a similar comparison is shown here in Figs. 14 & 16). There is also a helpful before and after restoration record of the National Gallery’s “Virgin of the Rocks” (see Figs. 17 & 18).

The Louvre and the National Gallery leonardo restorations share a common methodological feature: in both cases it is said that old varnishes were thinned but not completely removed. This claim creates a conundrum because in both cases changes have taken place which seem inexplicable in terms of a mere thinning of varnish. When explanations are sought or when appraisals are attempted, restoration authorities sometimes take fright, retreating behind claims that theirs is a highly specialised technical field whose mysteries are simply unfathomable to outsiders. Restorers themselves often don the proverbial doctors’ white coat and claim to have acted on not aesthetic grounds at all but on (quasi-) medical ones. For the “St. Anne” restorer, Cinzia Pasquali, this restoration was not made for aesthetic reasons. Instead: “This was about caring for a sick patient. From the conservation point of view we had to intervene, primarily to address a cracking of the varnish that could leave the paint exposed to damage.” Well, we now know that in this particular case the patient was not as sick as had been thought. But more importantly, we should remember at all times that works of art are made by people to be looked at by people. They are not created as laboratory specimens. Artists work with materials so as to produce values and relationships between values. No scientific test can analyse a value, let alone an inter-related group of values. To its maker, the professional test of a work of art is how it looks – the painter stops working precisely when the picture looks right.

In the realm of art and away from corporately funded museum politics, the ultimate test of a restoration is also how it looks – but that is to say, not how it strikes the passing viewer (who may or may not be thrilled by solvent-brightened colours) but how it looks now as compared with how it looked previously; how it looked not just immediately prior to restoration but in its successively recorded history; and, most especially, how it looked the last time it was cleaned. If picture cleanings did no harm, if they were as simple and non-destructive as cleaning a window, each restored work would return to its appearance when last cleaned, and there would be no surprises, “discoveries”, “revelations”, “restorations of the century” – or controversies. While no one ever berates a window cleaner for ruining the views, restorations irreversibly change a picture’s “view” on to the world. Restoration is a one-way street that runs away from history, away from the original work.

All cleaning controversies turn on the extent to which pictures suffer during restoration. Even among those who authorise restorations, some concede that there are losses as well as gains and frankly admit to seeking the best trade-off between improved legibility and pictorial injury. Defensive restorers insist that pictures cannot be harmed by their own “advanced”, “gentle” and “scientifically underpinned” methods. Making a fetish of the “safety” and the “science” of restoration methods attempts to shelve restorers responsibility to identify and account for all material and aesthetic changes. Given that all restorers’ methods cannot be superior, none should be held beyond question. With the physical alteration of art, aesthetic appraisal is essential to scholarship and art’s protection. In appraising restorations, the comparison of like with like is of the essence.

In visual arts, appraisals are necessarily made by visual comparisons. Pictures are made by eye, hand and mind, to be viewed by eye and mind. Because each cleaning destroys the earlier state, comparisons can only be made between pre and post-restoration photographs. While straightforward cleaning might always be expected to achieve a greater vivacity of pictorial effect, it should never be made at the expense of the pictorial relationships, patterns, or gradations made in the service of modelling, that can be seen to reside in the uncleaned work. If the relationships can be seen it is because they are there – whatever chemical analyses might suggest to the contrary. The aesthetic production of pictorial values by artists is the proper science of art. Unfortunately, in such terms, the values that were formerly evident in this great picture seem not to have fared well in this last cleaning.

Michael Daley

Comments may be left at:

Printer-friendly PDF version of this article

Above, Fig. 1: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail, before cleaning. Note the artist’s very selective and artfully focussed disposition of his brightest lights, and, at the same time, the extremely subtle but sculpturally effective modelling of the Child’s right shoulder and arm. Note too, the careful placement of tonal values throughout this grouping and how successfully these values contribute to a general sense of sculptural placement in space – for example, how the Child’s left forearm recedes behind the bright wooly top of the lamb’s head, and how appreciably it recedes from the Child’s nearer right arm.
Above, Fig. 2: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail, after cleaning.
How to explain the differences between these two greyscale versions above, of Figs. 3 and 4 below? It has been said that the restorer left a thin layer of varnish over the paint throughout the picture. It is hard, on the face of it and given the scale and nature of the changes that took place in this single restoration, to see how they might have resulted from a reduction rather than an elimination of the varnish (but see below). And it is hard, too, to see how a restorer, cleaning freely by hand with nothing more precise than a cotton wool swab (evidently dragged not rolled in this instance) might differentiate perfectly between the lowest level of a varnish and the upper level of an old glaze of similar colour and tonality at every point. When previous restorers had applied their varnishes, often, presumably, after harsh cleanings, would those then-new, solvent-saturated varnishes not have integrated themselves to any degree within whatever material was to hand? One question that might always be borne in mind when evaluating pre and post-cleaning states, is whether or not the cleaned (altered) state looks more or less characteristic of the artist’s known traits. (This might be held a perilously subjective notion to conservators of a certain “scientific” bent, but without such an artistic navigational system, how might any restorer proceed?) Does this after-treatment Fig. 2 detail look more Leonardesque than the before treatment detail at Fig. 1? It is hard to see how this question could be answered in the affirmative. The melting of the Child’s limbs into and out of the artist’s light in Fig. 1 seems quintessentially Leonardesque, while the after-cleaning state of Fig. 2 might be thought rather more Michelangesque by comparison. A key difference between these two great Renaissance figures (and sometime rivals) was that Michelangelo was not averse to autonomously forceful contours. Leonardo, of course, wished to out-sculpt the great sculptor with shaded simulations of form on the picture plane; with forms disposed within an envelope of space and light that was entirely of his own shaping and in no way dependent on the contingencies of the real world in which sculptors’ productions must always take their chances. The existence of the surface upon which paint was applied was a fact to be denied or concealed by the sheer force of artistry. One consequence of this cleaning is that the painting’s picture surface comes further to the fore. The restorer, Cinzia Pasquali, attributes this to the fact that although the work was known to be unfinished, “now we can actually see it”, as if that might be considered some kind of gain, but anything that causes the picture surface to compete for attention with the intended illusion upon it can hardly be thought characteristic of Leonardo’s wishes or intentions, and anything that causes a picture to seem unfinished and less resolved than previously was the case, might more realistically be taken as an aesthetic alarm call than as a vindication of raw method.
Above, Fig. 3: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail, before cleaning. Work on this section of the painting produced one of the strongest and most keenly contested controversies within the international advisory committee. A “varnish” on the Child’s body was taken by the restorer (and others) to be an earlier restorer’s decayed varnish. This reading was challenged by the student of Leonardo’s painting technique, Jacques Franck, who noted that this material had contributed to the modelling of the Child to a substantial degree and in a manner that went beyond any straightforward varnish layer. He felt, therefore, that it should be preserved until no doubt existed about its precise function and date of making. He and others of this opinion called for the disputed coating to be revived rather than removed, but they were over-ruled on the committee. Analytical tests were made on the material and these were said to have proved on chemical grounds that the material taken to be constructive orginal glazing by Leonardo was in fact only a later varnish. But if this chemical analysis is held to have provided an indisputable basis for excluding the possibility that the material was original, then it is incumbent upon those who removed it to explain how the various apparently artistic effects that it had contributed, had been achieved. In a nutshell the problem is: How might an overall “varnish” as opposed to a glazed layer, contribute differently in local areas that happen to coincide with discrete parts within an artist’s design? In effect, this is the same challenge that we mounted over two decades ago to the restorers at the Sistine Chapel who held that sculpturally-enhancing shadows on Michelangelo’s frescoes were the happy consequence of soot from candle smoke that had accumulated on the ceiling over many centuries.
Above, Fig. 4: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail, after cleaning. Visually, it would seem clear that the ground around the lamb’s tail has been in effect “scoured”; that darkened passages which threw the lamb into relief and prominence have been in effect “abraded”. Is it possible that a mere thinning of an overall varnish could have been responsible for such a transition?
Above, Fig. 5: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail (St. Anne), before cleaning.
Above, Fig. 6: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail (St. Anne), after cleaning. The changes in this face can only thought alarming and deleterious. Vincent Delieuvin’s claim that cleaning has enhanced the sculptural effects of the picture seems plain wrong. For any draughtsman or sculptor, the image at Fig. 5 would have to be considered to hold more “information” than that at Fig. 6. The shadows contribute to a far greater sense of sculptural relief and surface relationships. One might say that the cleaned state now appears to be modelled in shallower relief than that found before treatment. Before the cleaning, the plaited braid of hair running over the top of the head, partook of a general system of shading. After cleaning it has emerged generally lighter and, on the viewer’s left side, no longer tucks into the general ensemble that comprises the more shaded side of the head. Moreover, of the shading on the face, it can immediately be noted that a certain transparency has been introduced – it is now possible to see under St. Anne’s (true) right eye, an earlier positioning of the iris by Leonardo. It is a commonly encountered consequence of picture cleanings that they take works further towards the condition of transparency that is seen in infra-red photographs, where the light penetrates the surface of the paint. While that kind of “imaging” is very useful in terms of identifying earlier stages in a work’s genesis and, specifically, in identifying an artist’s own under-drawing, it cannot be a good thing for works of art themselves to be rendered transparent.
Above, Fig. 7: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail (St. Anne), before cleaning.
Above, Fig. 8: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail (St. Anne), after cleaning. This issue and the artistic dangers of increased transparency had been raised on the international advisory committee by Jacques Franck who had urged that a sufficiency of varnish be left on the face of St. Anne to prevent the inevitable consequence of age-induced transparency in Leonardo’s paintwork from emerging. He has described the peculiarly heightened dangers that were to be expected because of Leonardo’s method on flesh passages at that late stage in his career (when the “Mona lisa” was being produced). He paraphrases his submissions to the committee on the constructive use of “velatura” in the St. Anne head in the following terms: “In Leonardo’s time, those ‘velature’ were meant to interplay optically with the undermodelling and a more roughly worked state of the image. It resulted in opalescent flesh tones linked to the shadows very gradually, thus producing the typical smoky effect called “sfumato” in these sections. The ageing of the binding agent through time has made the opalescent micro-layers become increasingly transparent: details like the eyebrows, some sharp accents in the mouth, in the nose’s end seem to have been executed in the final state of the Saint’s head but have not. They are parts of the underdrawing that are emerging in the visible light due to increased transparency now.The same with the undermodelling. To date, the soft transitions having lessened markedly, the contrasts between light and shade are much stronger, inevitably so. More microns of old brown varnish left [in place] would have compensated for the now missing opalescent subtleties of Leonardo’s “sfumato”. Hence the difference to be observed between before and after cleaning.The Louvre was advised by me not to thin too much for that very reason. Leonardo’s subtleties need a substantial ‘veil’ of old varnish left over them, a situation clearly respected by Alfio del Serra in cleaning [Leonardo’s] Annunciation in the Uffizi, for the picture’s atmospheric effect is beautifully preserved.”
Above, Fig. 9: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail (St. Anne), before cleaning.
Above, Fig. 10: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail (St. Anne), after cleaning. One sees in this comparison of the eyes, not only the emerged iris but also a substantial degree of lost shading around the eyes. This is very commonly encountered in restored faces. The eye is best understood anatomically as a ball set in a hollow (a hollow that is formed by the brow/nose/cheekbone configuration), and its surface is only made visible through lids that close for protection and during sleep. The “eye” that is defined by the aperture of the opened lids is properly to be understood as a part of the surface of the larger eyeball. The relationships between these component parts are very distinctive to individual heads (or to the idealised “types” of heads devised by artists). Anything that reduces the original artist’s construction of those relationships (made by shading essentially) is extremely harmful to the “plastic” properties of the head as well as to its characteristic expression. A commonly encountered feature of restored faces is that the shading around the eye, which “sets” it properly in its recessed protective zone, is so diminished that the more precisely delineated parts – the shapes of the eyelid apertures and the iris/pupil – of the eye become more apparent, become over emphatically drawn. With regard to the level of cleaning that is said to occurred on the St. Anne, Franck had been assured that varnish would be left in place to a thickness of 18 to 20 microns. A micron is only one-thousandth of a millimetre or 0.001 mm. It might be wondered how a restorer working with solvent-laden cotton wool swabs (as seen in use on the DVD film, for example) might ever be able reliably and predictably to operate evenly to such ultra-fine tolerances. In the event, Franck was told that on the St. Anne the level of varnish that had been retained was of only 12 – not 18 to 20 – microns depth, or in other words of 0.012mm. This raises the question: Was this, as had been promised, the area of greatest varnish thickness that was left in place, or was this, in fact, the “uniform level” to which the painting had been cleaned throughout, as is described in the exhibition on the wall notice?
Above, Fig. 11: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail (St. Anne), before cleaning.
Above, Fig. 12: Leonardo da Vinci, “Virgin and Child with St. Anne”, detail (St. Anne), after cleaning. It would seem inconceivable, to any sculptor’s trained-eye that this section of the face, after cleaning, might be considered to enjoy enhanced sculptural values. In purely formal terms, what is seen here is an advance and an expansion of the lights at the expense of the darks – which darks had comprised in this working method the “constructive” component of Leonardo’s “modelling” on the light ground of his picture. The lights had not been painted as values, they were merely the sections of ground left unmodified by Leonardo’s meltingly applied shadows.
Above, left, Fig. 13: “The Virgin and Child with Saint Anne and the Infant St John the Baptist”, (“The Burlington House Cartoon”). Above, right, Fig. 14: “Virgin and Child with St. Anne”, before cleaning.
Above, left, Fig. 15: “The Virgin and Child with St. Anne”, copy, c. 1508-13, Armand Hammer Museum of Art, Calif., US. Above, right, Fig. 16: The “Virgin and Child with St. Anne”, after cleaning.
Above, left, Fig. 17: “The Virgin of the Rocks”, the National Gallery, before cleaning. Above, right, Fig 18: “The Virgin of the Rocks”, the National Gallery, after cleaning. As with the Louvre’s St. Anne, the varnish (which had been described as having come to constitute a threat to, as well as a disfigurement of, the paint) was said to have been thinned not removed. Indeed, when shown the painting part-cleaned, and when it was lit with an ultra-violet lamp, remains of (patchy) varnish were to be seen on the picture. Against that evidence, we face the problem of how the changes that manifestly occurred (as seen above) could have arisen. For example, how was the angel’s mouth changed if it remained under a film of varnish? What accounts for the fact that after the last cleaning the picture did not return to anything like its condition when previously cleaned sixty years before? Specifically, what accounts for the great lightening of the sky seen in the top right, as opposed to the sky seen on the left? What accounts for the great change in the Virgin’s blue robes?
Above, Fig. 19: The short Louvre catalogue, left; right, an illustration published in 1992 of the principal heads in the “St. Anne.” The emerging chasm between such photographic records of the same painting has yet to be addressed by scholars and curators.
Click on the images above for larger versions. NOTE: zooming requires the Adobe Flash Plug-in.

What Price a Smile? The Louvre Leonardo Mouths that are Now at Risk

26th January 2012

There is an air of confusion, and even a whiff of crisis, in today’s international art restoration community. At the National Gallery’s CHARISMA symposium on Leonardo’s technical practices and influence (January 13-14), many fascinating and highly informative demonstrations were made of non-invasive imaging techniques. A ground-breaking paper by Ana González Mozo revealed how the Prado’s copy of the Mona Lisa had faithfully tracked Leonardo’s own revisions on the original work. Imagine – a copy not so much of the finished masterpiece but of its very genesis. The proceedings, happily, are to be published. Unfortunately, at this richly rewarding and even-handedly conducted event, two critically serious weaknesses emerged during the proceedings and discussions that seemed linked and together to constitute a wider international art conservation faultline.

That is, firstly, there are manifest deficiencies of understanding on the crucial relationship between the discoveries that are being made through advances of technical analysis, and the original painterly/artistic means by which the art-objects-under-investigation were produced by artists in the first place. This single shortcoming carries profound cultural and professional consequences and is, we believe, a root cause of many of the controversies which arise in the field. Secondly, and concerning these controversies, a number of speakers (particularly Italian speakers, as it happens) used their papers to dismiss critics of their (sometimes long-past) restorations. Such combative unrepentance suggests how hard the task remains to establish some proper and effective systems and habits of disinterested critical appraisal in what is becoming an alarmingly expanionist and self-authorising and validating field. Much might be gained if restorer/conservationists themselves would devote more energies to those fruitful knowledge-advancing studies that leave vulnerable and uniquely precious historical works of art intact and free from what too often seem injudicious or debilitating interventions.

For some speakers, “virtual” indications of changes within pictures, seemed to be taken in themselves as invitations to swabs-on interventions. Certainly it seemed tactless and provocative when, at a time of great disputation, a Louvre curator, Vincent Delieuvin, ended a talk on non-invasive examinations of the museum’s Leonardo “Virgin of the Rocks” with the bald declaration “Restoration of the painting is possible”. Possible? No doubt it is politically so today at the Louvre, but desirable even under the present controversial circumstances? And what exactly has made this hitherto untreatable picture (see caption at Figs. 2 and 3) treatable? Must every restored masterpiece immediately trigger another? Will curators and their restorers never allow decent intervals for the fumes to evaporate and for their handiwork to be evaluated with due attention and consideration after the inevitable PR media barrages that nowadays accompany all major restorations? For that matter, is no one in authority concerned by the general and wider risks that are being incurred in the current Leonardo binge in which the oeuvre itself is being speculatively “grown” even as its most secure works are being recast by today’s more adventurous generation of technicians?

Concerning the Louvre fracas, Didier Rykner, surprised and disappointed us in his Art Tribune post of January 7th by disparaging the international debate on the restoration of Leonardo’s “The Virgin and Child with Saint Anne”, that followed a Guardian report. Headed “A follow-up to the Debate…” that somewhat churlish post afforded a proxy defence of the Louvre’s current policy. Of the momentous resignations of Sègoléne Bergeon Langle, the former director for conservation at the Louvre and France’s national museums, and Jean-Pierre Cuzin, a former director of paintings at the Louvre, from the Louvre’s own international advisory committee, Rykner complains that the restoration has sparked a debate “when in fact nothing new has happened since since the month of October (see our article in French).” One thing that is new is that everyone now knows (thanks to the diligence and reach of the international press) what very few previously appreciated: a profound schism exists over picture restoration methods at one of the world’s greatest museums – in truth, the world’s greatest single museum – and, that the restoration of a major work of the world’s most famous artist is the centre of that dispute.

Rykner characterised the “Virgin and Child with St Anne” restoration as one in which only “three points had to be solved”, when, far from being some problem-solving exercise, this is a debate about the very nature and legitimacy of actions that might be made on an immensely complex, unfinished, ancient and precious masterpiece. Restorations never take place in antiseptic clinical spaces. They always reflect philosophies, interests and professional inclinations or dis-inclinations towards radical interventions. Conflicts in this arena are far from trivial and it is desirable that they should not take place in the dark.

There is great irony in the present situation. For many years after the Second World War, the Louvre occupied the virtuous conceptually elegant and restrained end of the restoration spectrum. And it did so in explicit opposition to the brutal simplistic campaign of “total cleaning” launched at the National Gallery by the restorer Helmut Ruhemann, who insisted that “important paintings should be thoroughly cleaned before we theorise about them”. Ruhemann’s influence at the National Gallery was under-written by the long-serving post-war director Philip Hendy, who wrote in the catalogue to the didactic 1947 exhibition of pictures cleaned (in secret) during the war: “The purpose of this exhibition is not only to do justice to the cleaned pictures, but, by extending the knowledge of their condition, to bring about a higher standard of criticism in this all-important subject.” By 1958 Bernard Berenson had been persuaded that “the recent restorations of pictures in the National Gallery are defended by most English people against the almost universal disapproval of continental craftsmen and critics.”

Berenson was right about the continental opposition but, perhaps being overly-reliant on the reports of Kenneth Clark (who had launched the National Gallery’s cleaning blitzkrieg in the late 1930s when he had hired “the brilliant Ruhemann”), he had quite missed the scale and the ferocity of opposition within England. In 1947 Hendy had made the self-defeating claim that: “There would not in all probability have been so much criticism of the appearance of the pictures cleaned since the war began if the public had been kept waiting until the cleaned pictures could be shown as they are shown [all together] now. Instead, they had to be scattered among the dirty pictures upon walls which there had been neither time no labour even to brush down after six years of increasing exposure to the elements.” If Hendy could not see the force of a widespread preference for “dirty” over “cleaned” pictures, by 1977 his predessor director, Kenneth Clark had quite forgotten the controversy the returned pictures had provoked: “A cleaned picture often looks very bright when first it is finished, but it loses its shine after a year or two, and I was fortunate in that the pictures I brought back for the reopening of the Gallery, all of which had been cleaned, had had time to ‘settle down’. No one complained about them.”

For a concrete example of the war-time consequences of Ruhemann-esque preludes to theorising, see Figs. 9 and 10. For a discussion of the role of the German restorer, Johan Hell, whose highly restrained work in Britain earned much support among artists and private patrons, see this author in the June, July/August and September 2006 issues of the Jackdaw. Hell’s great influence in Britain (he restored the collection of the Dulwich Picture Gallery to acclaim) was carried to the United States, albeit in dilute form, by his sometime student, John Brealey. In 1951, René Huyghe, the Louvre’s chief curator of paintings and drawings advised restorers and curators “If there is the slightest doubt, stop; for in this matter it is inadmissible to make a mistake. Excessive prudence leaves the future open without compromising it. But mistakes made by presumption are irreparable.” In the Sunday Times (15 January 2012) Bergeon Langle said “We prefer the most moderate cleaning possible. But my views weren’t taken into consideration.”

Not heeded, perhaps, but widely noted nonetheless. The debate that has been opened by the two Louvre resignations shows no sign of abating. In a recent article Bergeon Langle has further explained: “I deemed that the restoration was not being carried out in line with what I imagined was necessary for this Louvre painting. That is my firm belief.” We now learn through such increasingly frequent press reports and the assorted leaks and briefings on this schismatic dispute within the museum’s twenty members strong advisory committee, that the Louvre’s present curators, with the encouragement of certain English and Italian curator and conservator members of that committee, are resolved to pursue more radical, less precautionary treatments. What makes this dispute-among-experts so alarming is that evidence of the consequences of the shift from caution and physical restraint within the Louvre already exists – we don’t have to read the book (and the official accounts, should they ever be made available), we can simply look at the pictures: as previously shown, restorers at the museum may now change and then re-change the expressions on old master faces without even being required to record their adulterations in the museum’s dossiers. This, by any standards is an already extraordinary and indefensible position.

News of the Louvre resignations comes just seven months after the shocking disclosure of its cosmetic facial exercises and at the point where the cleaning of the “St. Anne” is completed and the perilous stage of retouching begins. It so happens that the expression of St. Anne’s mouth is interrupted by a panel join (see Fig. 1) which might provide another temptation for a little cosmetic surgery with the retouching brushes. We were not reassured when the Louvre’s present head of painting, Vincent Pomarede said recently that the retouchings would be “reversible”. “So what?”, we would ask. As shown right (Fig. 5), the Louvre recently made two egregious errors on a single Veronese face. That the second falsification – which is still in place – is theoretically as removable as the one it superseded does not mitigate the presently persisting offence and lapse of curatorial vigilance. The two members of the Louvre advisory committee from the National Gallery who have been enthusiastic supporters of the current restoration (the then curator Luke Syson, who has moved to the Metropolitan Museum, New York, and the head of conservation, Larry Keith), happen themselves to have been responsible on their own patch for another incomprehensible, initially unacknowledged (- even in the Gallery’s own Technical Bulletin report) but again “reversible” alteration to a (Leonardo) mouth – see Figs. 6, 7 and 8.

One of the more monumentally controversial restorations defended at the National Gallery symposium was that of Leonardo’s “Last Supper” by Pinin Brambilla Barcilon. Sponsored by the Olivetti corporation, the restoration had been directed by the distinguished Leonardo scholar Pietro Marani, who is a member of the Louvre’s “St. Anne” advisory committee, and like Syson and Keith an enthusiastic supporter of the present restoration. When we pointed out at the conference that the previous restoration of the “Last Supper” in the 1950s had been praised at the time by Bernard Berenson as an ultimate recovery of all the surviving authentic Leonardo paintwork, Pietro Marani countered by claiming a right for each generation’s restorers and curators to impose their own values and interpretations on historic works of art. The consequences for Leonardo’s “Last Supper” of the claimed (“Buggins’ Turn”?) right to undo and redo every great work of art, will be examined in our next post.

Michael Daley

Comments may be left at:

Printer-friendly PDF version of this article

Above, Fig. 1: St. Anne, detail from Leonardo’s “Virgin and Child with St. Anne”. The vertical line that runs through the right-hand end of the mouth marks a split in the panel.
Above, Fig. 2: Detail of the face of the angel in the Louvre’s version of Leonardo da Vinci’s “Virgin of the Rocks”. The heavy craquelure is, in part, a likely exacerbating consequence of the paint film having been (badly) transferred from panel to canvas by restorers at the Louvre in 1806. In any event, the persisting fragility and vulnerability of this surface has hitherto protected the painting from “restorations”, of which the National Gallery version of this painting (see Figs. 6, 7 and 8) has had two since the Second World War.

It seems to be being claimed now that scientific tests have shown that the painting can safely be restored. Given a) the extremely cracked and irregular surface; and, b) the extraordinary subtlety of the modelling that attaches to the individual fragments of the paint but works artistically across them and despite their physical and optical disruptions, it is hard to imagine what assurances “scientific analyses” might offer in support of taking the sheer risk of working upon such a painting with solvents which are notoriously invasive and pentrating and love nothing better than a good crack through which to advance themselves. Would a restorer attempt to clean each item of paintwork individually or work across several at a time while attempting to preserve the immeasurably subtle artistic relationships that would be out of sight underneath the solvent-laden and abrasive swab?

Above, Fig. 3: Detail of the face of the Virgin in the Louvre’s version of Leonardo da Vinci’s “Virgin of the Rocks”.
Above, Fig. 4: The mouth of the “Mona Lisa”. Those who would restore this painting, even in its present parlous state, might first recall Leonardo’s account of painting flesh values:

Put on the carnations with silk brushes, and while they are fresh you can make the shadows as vaporous as you will. The carnations should be painted with white, lake, and massicot; the shadow should be of black and matoric (massicot), with a little lake and black. When the picture has been sketched in lightly, let it dry and then retouch it with lake soaked in gum-water, and this should have been left some time in gum-water, because it is thus better, and will not have any lustre when it is used…”

The scholar of Leonardo’s painting techniques, Jacques Franck, has adduced that the “vaporous” subtlety of expression encountered here could only have been produced by minute touches of brushwork applied in multiple glaze layers, and has warned that with such a blur of superimposed micro-layers of vulnerable glazes, one could not be sure that the solvent used, even in the course of a very gentle cleaning, would not cause inevitable damage, whether in the ultra delicate zone of the mouth or elsewhere. As it is, any gum-bound pinks have long since faded or perished at restorers’ hands, along with the Mona Lisa’s eyebrows – which have survived on the Prado’s copy.

Above, Fig. 5: The top two images show changes made to the mouth and nose of a Titian at the Prado. Among the many solecisms introduced in this single “restoration”, the tip of the nose now terminates not in the centre of the philtrum (the groove that runs from the upper lip), as human anatomy decrees, but behind its far edge, in a quasi-cubist manner; the lower lip has been padded-out so as to swing entirely in a convex curve to the left terminus of the mouth; that terminus now sports two little creases, not one, thereby suggesting that the mouth turns both up and down. The bottom three images show the recent changes made to a Veronese face at the Louvre that were detected and reported by Michel Favre-Felix of ARIPA.

Above, Fig. 6: Detail of the face of the angel in the National Gallery “Virgin of the Rocks”, above, after its recent cleaning but before retouching. Below, Fig. 7, below, the angel after its recent cleaning and retouching, in which the far side of the mouth was turned down against art historical and technical (including x-ray) evidence.
Above, Fig. 8: Changes made to the mouth and nose tip of the angel in the National Gallery’s “Virgin of the Rocks” in the course of two restorations.
Above, Fig. 9, and below, Fig. 10: Details from the National Gallery’s Pontormo “Joseph with Jacob in Egypt”, before (left, and top), and after two cleanings and restorations (right, and bottom). These images are taken from the 1938 and 1990 editions of Kenneth Clark’s book “One Hundred Details from Pictures in the National Gallery”.

In between the two photographs (the post-cleanings photographs are grey-scale conversions from the large colour plate in the 1990 edition) this painting was restored twice by the Gallery, in 1940 and in 1981-82. It was said of the first cleaning that it had been undertaken because the picture was “much disfigured by dark spots and discoloured varnish”.

The second cleaning was undertaken because “The picture is still blistering. There are numerous old small retouchings, obvious repairs along the cracks and a large hole in the green robe lower left…[and because there was a] discoloured layer of old varnish with some blanching.”

The treatment report after the second restoration read: “The discoloured varnish and retouchings were removed with propan-2-ol and white spirit leaving a thick greyish layer of surface dirt and varnish remnants which was removed with a potassium oleate soap. The blisters were laid with gelatine and the painting restored with pigments in Paraloid B-72 and varnished with Ketone-N.”

It should be said that we are greatly indebted to the National Gallery for making its conservation records and photographs available to us. In this particular case, in setting out the post, we queried our own scanned post-cleaning photographs and re-scanned them as a precaution. The second scan was identical to the first. Thus, in so far as the (Gallery’s own published) photographs are reliable, the viewer may fairly make here an appraisal of the artistic cost (in terms of lost tonal differences; lost modelling of forms; lost details – as in curls of hair, for example) that was paid for the removal of the disfigurement represented by spots of darkened retouching. Moreover, such trade-offs, which are surprisingly common in museum restorations, are rarely strictly necessary – or necessary at all. Jean-Pierre Cuzin, the former director of paintings at the Louvre who resigned from the St Anne advisory committee, was of the opinion that the discoloured touches on the St Anne could simply have been repainted. (The varnish had been thinned, replenished and regenerated on a number of occasions since the war.) Such retouching practices are common: aside from its two major cleanings and restorations, this Pontormo had simply been retouched in 1955 – and again in 1979, just two years before its second cleaning.

Perhaps the last word should be given to Sègoléne Bergeon Langle who has recently commented: “As a general rule, I prefer old works to be cleaned only very lightly. You can never retrieve the original colours, what you get instead is the current state of the painting materials. Leaving a slight veil over the work makes it more harmonious.”

Click on the images above for larger versions. NOTE: zooming requires the Adobe Flash Plug-in.