Madonna and Child with St Anne and a Lamb was bought before 1985 by the Sackler family or one of its trusts. Its earlier provenance is unclear, but it apparently belonged to a Mario Sassi in the 1950s and was later in a Venetian collection. By 1979 the Leonardo was in Geneva, and it presumably had required an export licence to leave Italy.
The drawing was first published in 1979, in a monograph by the Italian Leonardo specialist Carlo Pedretti. Between 1979 and 1982 he showed it in exhibitions in Vinci, Rome and Milan.
Pedretti, the leading expert of Leonardo drawings studies, continues to accept Madonna and Child with St Anne and a Lamb. In April 2008 he wrote to Dickinson that he was “well familiar” with the work, which he regarded as authentic. Pedretti’s judgement gives the drawing great credibility, although some scholars feel that he is too “inclusive” in his attributions.
The other leading authority is Martin Kemp, emeritus professor at the University of Oxford. In 1992 he exhibited the work at the National Gallery of Scotland in Edinburgh, labelled as by Leonardo. However, in his catalogue, Kemp admitted its status was unclear: it could have been authentic, from the artist’s studio or a skilled forgery. Kemp himself plumped for a fourth theory: the underlying drawing is by Leonardo, dating to around 1501, but the composition has been reinforced by a later hand.
After having had the chance to study the work more closely in Edinburgh alongside fully accepted drawings, Kemp began to waver. When he saw the drawing again in spring 2006 he expressed his concerns.
In June 2008, Kemp and his colleague Juliana Barone (Birkbeck College, University of London) were invited to inspect the work at Dickinson’s gallery. On 3 July 2008 Kemp revised his opinion, concluding that rather than dating from around 1501, he would “now assign the drawing to the French period, 1516-9”, as a Leonardo.
Although no longer the owner, Gheri Sackler commented after learning of Kemp’s verdict that “definitely this increases the value of the drawing considerably”. Kemp is only too aware that academics are put in a difficult position when their attribution has a major impact on the financial value of a work. He himself has a strict policy of not accepting payment for attributions.
Kemp is now reluctant to discuss Madonna and Child with St Anne and a Lamb, because of the extraneous legal disputes which have developed. However, since 2008 he is believed to have had growing doubts about the drawing.
Conservator Peter Bower examined the drawing in 2008 at the request of Dickinson. Bower believes the paper is French, rather than Italian. “The paper evidence suggests that it is likely that the drawing was executed during Leonardo’s time at Amboise between 1517-19,” he concluded.
Italian specialist Carlo Vecce (University of Naples L’Orientale) told us that although he has not inspected the drawing, based on photographs he believes it is “authentic”. It was also accepted by Patricia Trutty-Coohill (Siena College, Loudonville, New York) in her 1993 study of Leonardo drawings in American collections (at this point its location was given as New York).
However, other scholars are doubtful about the attribution. Carmen Bambach, a specialist at New York’s Metropolitan Museum (and curator of the 2003 exhibition “Leonardo da Vinci: Master Craftsman”), saw the drawing in 2006. She rejects it and told us that it is “of a later date, on paper of 1540-90”. This makes it at least 20 years after Leonardo’s death and possibly considerably later.
Hugo Chapman, a British Museum specialist and curator of last year’s “Fra Angelico to Leonardo” exhibition, saw Madonna and Child with St Anne and a Lamb some years ago. He told us that he “does not think it is by Leonardo”.
Martin Clayton, a Leonardo specialist at Windsor Castle, has not seen the drawing and as a Royal Collection curator he feels it would be inappropriate to comment. He does not refer to the drawing in any of his published writings.
Frank Zöllner, of Leipzig University, does not include the work in his 2003 monograph, Leonardo da Vinci: the Complete Paintings and Drawings.
Specialists from both Christie’s and Sotheby’s have seen the drawing, and neither auction house apparently accepts it (although they never comment on works they are not selling). John Martin QC, for Accidia, said in the High Court that “the whole complaint by Mr Kazeminy was on the basis that the drawing was tainted because Sotheby’s and Christie’s had seen it and had said ‘not interested’.”
Dickinson director Emma Ward is convinced of the Leonardo attribution and is undertaking what will be the most comprehensive research on the drawing. She intends to invite all the experts to inspect it out of the frame. Once this work has been completed, Madonna and Child with St Anne and a Lamb will be offered for sale—and the market will decide.
The most expensive Leonardo drawing to sell at auction is Horse and Rider, around 1480-81, which went for £8.1m at Christie’s in 2001, since when prices of the greatest old masters have risen.