Artwatch UK

Posts tagged “Martin Kemp

Problems with “La Bella Principessa” – Part III: Dr. Pisarek responds to Prof. Kemp

In June 2015 Kasia Pisarek, an independent scholar (and a member of ArtWatch UK) published an article, “La Bella Principessa – Arguments against the Attribution to Leonardo”, in the Polish scholarly journal Artibus et Historiae.

In her article Dr Pisarek presented a number of interlocking historical, aesthetic and technical criticisms of the attribution to Leonardo of the drawing “La Bella Principessa”, as it has been made and advanced by Professor Martin Kemp. In response, Prof. Kemp produced an article (“Leonardo da Vinci La Bella Principessa: Errors, Misconceptions, and Allegations of Forgery”) which challenges Dr Pisarek’s account on grounds of what he claims and alleges to be: “mistakes, misconceptions and a series of false allegations”.


In his response Kemp says “I do not run an authentication service, but research items of special interest regardless of ownership.” More recently ( May 16) Kemp announced on his website that “After speaking at the Art in Authentication Congress in The Hague, I confirm that I am withdrawing the [unpaid – Ed.] ‘advice service’ I have been providing.”


Kemp discloses that in responding to Pisarek’s article he also sought by “extension” to counter other un-identified challenges to his Leonardo attribution. When this multi-targeted professional defence was submitted to Artibus et Historiae, it was rejected, as Kemp acknowledges, and as the Art Newspaper reports (“La Bella Principessa: Still an Enigma”, Features, May 2016), because of its resemblance to “an errata list”. The article was subsequently carried on the Authentication in Art website to accompany a paper given by Kemp at the AiA’s May 2016 Congress. This non-profit organisation, on which Kemp serves as an advisor, was founded in 2012 at The Hague. On May 8th we made a formal request to the AiA for Kasia Pisarek’s Artibus et Historiae article also to be posted so that the congress speakers and attendees might see both of what Dr Pisarek’s compilation of evidence consisted and of what Prof. Kemp complained. We have yet to receive a reply. For Kasia Pisarek’s Artibus et Historiae article, “La Bella Principessa – Arguments against the Attribution to Leonardo” click here. For Martin Kemp’s response to it, see: “Leonardo da Vinci La Bella Principessa: Errors, Misconceptions, and Allegations of Forgery”.


In December 2015 Kasia Pisarek delivered a paper based on her Artibus essay at the ArtWatch UK/LSE Law Department/Center for Art Law conference “Art, Law and Crises of Connoisseurship”. At this conference, an arts journalist, Simon Hewitt, delivered an attack from the floor on her article. He did so as a proxy for Kemp, by whom he had been briefed, and for the collector/dealer owner of “La Bella Principessa”, Peter Silverman, with whom he is co-writing a book on what Silverman describes to us as “various aspects of the art market, sometimes highlighted by others’ and my own discoveries”. A research assistant of Kemp’s, Kasia Wozniak, who had spent four years attempting to show that “La Bella Principessa” had once been part of a late 15th century book known as the Warsaw Sforziad, had also written to us in denigration of Pisarek’s article which we had forwarded to her at her request.

Earlier, both Professor Kemp, who had declined an invitation to speak at this conference, and Mr Silverman, had suggested that we invite Hewitt to speak at the conference. Silverman expressly requested that this be done so that Hewitt might “present his discoveries as a counterweight to Ms. Pisarek”. We carry Pisarek’s full reply to Kemp’s listed objections below but comment first on a crucial new aspect of this disputed attribution that has emerged in Kemp’s response to Pisarek.


Above, Fig. 1: A facsimile of the “La Bella Principessa” drawing being inserted into the Warsaw Sforziad (but showing only the relationship of the top of the facsimile to the top of the book and not the accompanying relationship of the facsimile and book at the bottom). Photo by courtesy of Lumiere Technology.

For many reasons, it is essential to Prof. Kemp’s Leonardo ascription that it be accepted that “La Bella Principessa” had originally been incorporated within the Warsaw Sforziad now held in the National Library, Poland. It has not been so accepted because, contrary to press releases claims and media coverage thereafter, nothing material or documentary has established such a relationship. No record of any connection between the drawing and the book has been found despite extensive searches by researchers such as Kasia Wozniak. Moreover, despite extensive searches of Berenson’s archives by Kemp and Silverman, no record of the supposed 15th century drawing in any context predates its only-recently acknowledged ownership by the late painter/restorer Giannino Marchig. This pictorially and graphically mongrel work remains without a history and, prior to September 2011, without any claim – even by its only known previous owner – that it might have been a work by Leonardo da Vinci.

Martin Kemp has challenged Kasia Pisarek’s measurements between the stitches in the book’s binding. He and Pascal Cotte (of Lumiere Technology) claim that three stitch holes are present on the left hand edge of the “La Bella Principessa” sheet and that these match three of the book’s five stitches. No confirming visual evidence has been produced in support of this claim. In addition to fresh evidence on dimensions provided below by Pisarek, it might be noted that the Kemp/Cotte claims of a match have been variously and only rarely unequivocally phrased. All emphases below are added:

The stitch holes in the vellum of the portrait match those in the book – “The Original Source of the New Leonardo Portrait Discovered”, a Martin Kemp press release, 27 September 2011.

“Three of the stitch marks, the ones that we can still see on the edge of the ‘Bella Principessa’, match as well as they conceivably could – Martin Kemp, Artinfo interview, “The Da Vinci Detective: Art Historian Martin Kemp on Rediscovering Leonardo’s Tragic Portrait of a Renaissance Princess”, by Andrew M. Goldstein, 11 October 2011;

“During our studies at the National Library, we inserted a facsimile of the portrait into the relevant opening of the book where the size matched very closely – Martin Kemp, item 14, “Leonardo da Vinci La Bella Principessa: Errors, Misconceptions, and Allegations of Forgery”, on the Authentication in Art website. See Fig.1, above;

“The current stitching of the volume involves five holes, whereas there are only three holes now visible along the left margin of the ‘La Bella Principessa’. However, these three holes correspond very closely to the corresponding ones in the book.” – Martin Kemp and Pascal Cotte, “La Bella Principessa and the Warsaw Sforziad”, Lumiere Technology website, September 2011.


With linear measurements a near-miss is as good as a mile. If a hole is two millimetres adrift of a stitch there is no match. Claiming “correspondences” and “close” matches between the three holes and the five stitches is problematic enough, but, as Kasia Pisarek has now re-confirmed, the three holes in the “La Bella Principessa” drawing do not correspond with three of the book’s five stitches. Moreover, Kemp’s imprecision came with a perplexing multiple caveat: “In measuring the distances between the holes and matching these distances in the book and the portrait we allowed for four potential sources of error” – Kemp/Cotte, item 13, “La Bella Principessa and the Warsaw Sforziad”, Lumiere Technology website.

If incorporating an allowance for one potential source of error would necessarily be weakening to the force of a claimed match, how might allowances have been made for four different sources of error? How were the different “potentials-for-error” calibrated and weighted one against the three others? With accounts of this attribution, too many features remain in flux. For example, explanations offered for the disparity between the drawing’s three stitch holes and the book’s five stitches have shifted. In 2011 Kemp/Cotte wrote:

“The second task was to see if the holes in the portrait and the stitching pattern in the book corresponds. There is an obvious difference. The current stitching of the volume involves five holes, whereas there are only three holes now visible along the left margin of “La Bella Principessa”. However, these three holes correspond very closely to the corresponding ones in the book…The different number of stitching holes may result from the untidy way the left margin of the portrait has been cut, or from two intermediate stitches being added when the book was later rebound in standard Zamoysky livery. The former explanation is more likely.” (Emphases added.)

The suggestion that the book might possibly have been bound originally with only three stitches seems to have been abandoned altogether. Martin Kemp now accepts in his response to Kasia Pisarek that the book always had five stitches but claims as a countering fact against this recognition that: “The irregularity and extensive damage along the left margin explains why two of the five stitch holes are no longer clearly discernible.”


The posited stitch holes cannot be said to be “no longer discernible” because there is no evidence that they were ever present. Prof. Kemp here begs a question on which this attribution turns. The roughly cut edge cannot be taken to have explained these absences. What Kemp offers, in truth, is (an implausible) hypothesis that ignores the technical exigencies of book binding and the dimensional realities of the “La Bella Principessa” sheet. When books are being made, the stitches are inserted along the line of a fold made collectively to the small number of sheets that form one of the book’s sections or “quires”. In the case of the Warsaw Sforziad, Pascal Cotte established (by taking and combining 70 precisely-focussed macro-photographs) that each quire was composed of four sheets (folios) which, when folded and stitched, comprised sixteen numbered pages. The book binder’s craft requires that the stitching occurs precisely along the crease line of the folded sheets. This careful alignment is necessary if the pages are not to cockle and for the book to open easily.

The three holes on “La Bella Principessa” have been taken to relate (- more or less, but never exactly) to the book’s central and two outer stitches. Had the “La Bella Principessa” sheet been incorporated in the Warsaw Sforziad when it was made in the late 1490s, the two inner stitch holes would be expected to be present on the sheet, even as it is today, and notwithstanding its roughly cut left-hand edge.

At Fig. 2 below, we see the white arrows and circles with which Pascal Cotte identified what are said to be “La Bella Principessa’s” three stitch holes. On the image on the left, we have drawn in red the alignment of the present three holes and have indicated with arrows where the two hypothesized additional stitching holes would be expected to be located. Both holes would fall within the present sheet despite its roughly cut left-hand edge. In the image on the right of Fig. 2 we again indicate (in black) the alignment of the present three holes, but show in red how the alignment would be disrupted had the two hypothesized additional stitch holes been situated to the left of the present sheet, as Prof. Kemp now claims in “explanation” for their absence from the sheet itself. Such a positioning would have resulted in a zigzag, not a row, of stitch holes. It is impossible to envisage how four sheets of vellum might have been folded so as to produce a neatly zigzagging crease.

Aside from the above problem, any lingering hope that this “La Bella Principessa” sheet might once have formed part of the Warsaw Sforziad will have to be abandoned in the light of Kasia Pisarek’s latest findings described below on her second examination of the Warsaw Sforziad.

Above, Figs. 2 and 3. In Fig. 2 (top) we see, in the left and right images, the white arrows and circles with which Pascal Cotte located the “three holes showing that the image was once part of a codex or manuscript”. Given that the book (the Warsaw Sforziad) from which this sheet is said to have been cut was originally bound with five stitches, had the “La Bella Principessa” sheet once formed part of that book, it would today have five stitch holes, not the present three. In Figs. 2 and 3, we indicate with arrows (in red and then in black) where the missing two stitch holes would, for the reasons given above, be expected to have been located.


When Andrew Goldstein asked Martin Kemp in an Artinfo interview in October 2011 what, on seeing the drawing, had convinced him it might be a Leonardo, he replied:

“So the initial connoisseur’s reaction merely tells you that something is worth looking at, but at any point one wrong thing can throw that all away — a later pigment, a bit of something that might come up about its history to indicate it was forged at some point, and so on. I was trained as a scientist, and if you have a scientific theory, you only need have one bit of the experiment that says, ‘this is not right,’ and the whole thing collapses. You always have to be looking for that one thing that is going to demolish the whole expectation that’s being set up.”

Kemp has given other grounds for caution when making attributions. On first encountering “La Bella Principessa”, he told Silverman, “I immediately saw it was in a different league from the others. But I was still very, very cautious. I didn’t want to jump at it because once you start believing you can summon up all the evidence you need.” (Peter Silverman, Leonardo’s Lost Princess: One Man’s Quest to Authenticate an Unknown Portrait by Leonardo da Vinci, 2012, p. 74.) Kemp had become a believer in the Leonardo attribution by 28 September 2011 when he issued a press release “The Original Source of the New Leonardo Portrait Discovered”. He added, “This (press release attached without the pics) should more or less settle the arguments – though probably not knowing the myopia of the art world.”

In 2011, after Kemp and Cotte had inserted their facsimile “La Bella Principessa” into the book, Kemp expressed himself 80 per cent confident of the drawing’s Leonardo attribution in a National Geographic film of the occasion. However, this was not the first time that a claimed “fit” between the facsimile and the book had been made and filmed by National Geographic. Peter Silverman describes in his book how, in December 2010, he had established a match in a different part of the book:

“We began by measuring the page size to see if it corresponded to “La Bella Principessa” and were gratified to see that it did, within a millimeter or two (a minute fraction of an inch)… Martin [Kemp] had surmised that the drawing would have been placed either at the very beginning or the very end of the book, but after careful examination we could find no trace of a cut page in either place. ‘May we turn each page?’ I asked. It was not a simple request. The book was nearly two hundred pages, and it would be a bit laborious for her [Anna Zawisza, the head of manuscripts] since utmost care had to be used so as not to damage the precious work in any way…It was apparent that the three pinholes where the binding had been sewn, noted earlier by Martin, which we had hoped would be a key to matching, would not be relevant, since the book had been rebound using five sutures…We slowly continued to turn each page, but there was no sign of a missing page…I had begun to abandon hope and to mentally prepare myself to return empty-handed. But then Zawisza turned page 161. There was a momentary beat of silence, and then she and I let out muffled cries. There, before our incredulous eyes, was what seemed to be the missing link, the element we longed to find: a remnant of a cut and extracted page of vellum that was the same darkish yellow as “La Bella Principessa”. We could barely contain our emotions. We measured the undulation of the remnant, and it corresponded exactly. Kathy [Silverman’s wife] and Kasia [Wozniak, art historian] came round to see for themselves, while David [Murdoch, National Geographic producer] filmed the historic moment. Even the armed guard was caught up in the excitement. Zawisza, who had carefully studied the book on past occasions, murmured how unhappy she was that she’d never noticed the missing page, now so glaringly obvious from the protruding remnant…”

That filmed historic moment was eclipsed by the moment in which Kemp and Cotte discovered a (preferred) location for the drawing at the front of the book. On her second examination of the Warsaw Sforziad, Dr Pisarek has learned that each of the five stitches in the book’s binding resulted in two holes, through which a string was passed. Thus, now that it is accepted that the book was originally bound with five stitches, each of which generated two holes, and that the “La Bella Principessa” sheet possesses only three of the necessary ten holes – and three where there should be six – there is no physical match between the drawing and the book, just as there is no documentary record of a Leonardo drawing having been bound within the book. Those who continue to see the hand of Leonardo in the drawing itself must now find an alternative history and another princess to accompany and bolster it.

Michael Daley 24 May 2016

Kasia Pisarek: A reply to Martin Kemp’s essay “Leonardo da Vinci La Bella Principessa. Errors, Misconceptions and Allegations of Forgery”

Professor Kemp has written an essay in response to my article “La Bella Principessa; Arguments against the Attribution to Leonardo”, Artibus et Historiae, No. 71, XXXVI, June 2015, pp. 61– 89.

In his essay, Prof. Kemp lists what he deems a series of errors and misconceptions in the Artibus article, but says he does not wish to address the issues of attribution I raised.

The purpose of his article is, however, an attempt to counter or undermine my findings.

I will answer his points in the order and with the numbered headings used by Kemp in his text.

[Martin Kemp] “1) Bibliographical”

[Kasia Pisarek] Martin Kemp says that most of my material is quoted from the internet and that I make only one reference to his book in my footnote 50.

This is incorrect. I make extensive reference to his book on the opening page and further references in footnotes 54, 57, 59 and 64. I have examined it as thoroughly as would be expected of any researcher. I also referred to many other books and articles which were accessed from libraries and not from the internet. These were:

M. Kemp and P. Cotte, The Story of the New Masterpiece by Leonardo da Vinci: La Bella Principessa, London, 2010; M. Kemp, Leonardo (rev. ed.), Oxford, 2004; P. Silverman, Leonardo’s Lost Princess: One Man’s Quest to Authenticate an Unknown Portrait by Leonardo Da Vinci, New Jersey, 2012; C. Geddo, ‘A “Pastel” by Leonardo da Vinci: His Newly Discovered Portrait of a Young Woman in Profile’, Artes, 2008–2009, pp. 67–87; C. C. Bambach, ‘Leonardo’s Notes on Pastel Drawing’, Mitteilungen des Kunsthistorischen Institutes in Florenz, vol. 52, 2008, no. 2/3 (Le tecniche del disegno rinascimentale: dai materiali allo stile. Atti del convegno internazionale, Firenze, 22–23 settembre 2008, ed. By M. Faietti, L. Melli, A. Nova), pp. 177–204; M. Gregori, ‘A Note on Leonardo’, Paragone, LXI, 2009, no. 723, pp. 3–4; D. Ekserdjian, ‘Leonardo da Vinci: “La Bella Principessa” – The Profile Portrait of a Milanese Woman’ (book review), Burlington Magazine, vol. 152, 2010, no. 1287, June (Attributions, copies, fakes), pp. 420–421; P. C. Marani, ‘Deux nouveaux Léonard?’, Dossier de l’art, 2012, no. 195, avril, pp. 58–63. Giannino Marchig, 1897–1983: paintings and drawings, exh. cat. London, 1988; Giannino Marchig, 1897–1983, exh. cat., Geneva, 1985; Giannino Marchig: 1897–1983: dipinti, disegni, incisioni, exh. cat. Florence, Gabinetto disegni e stampe degli Uffizi, 12 March – 5 June 1994 (Italian ed.); J. Cartwright (Mrs Henry Ady), Beatrice d’Este, Duchess of Milan, 1475–1497. A Study of the Renaissance, London, 1910; B. Horodyski, ‘Miniaturzysta Sforzów’, Biuletyn Historii Sztuki, 16, 1954, pp. 195–213; E. McGrath, ‘Ludovico il Moro and His Moors’, Journal of the Warburg and Courtauld Institutes, vol. 65, 2002, pp. 67–94; L. Syson with L. Keith, Leonardo da Vinci, Painter at the Court of Milan, exh. cat. National Gallery, London, 2011; Dizionario delle origini, invenzioni e scoperte nelle arti, nelle scienze…, Milan, 1831; B. Berenson, The Drawings of the Florentine Painters, vol. III, Chicago, 1938; Leonardo da Vinci, Master Draftsman, ed. by C. C. Bambach, exh. cat., The Metropolitan Museum of Art, New York, 2003; Z. Zygulski Jr, ‘Ze studiów nad Dama z gronostajem. Styl ubioru i wezly Leonarda’, in: Swiatla Stambulu, Warsaw, 1999 (first published in Biuletyn Historii Sztuki, vol. 31, 1969, no. 1, pp. 3–40).

The 2012 Italian version of Kemp/Cotte’s 2010 book is a translation from the English with the addition of the Sforziad hypothesis. The latter had already been published on the Lumiere Technology website and I discussed it at length.

Gnignera’s text was in Italian so I used Prof. Zygulski’s extensive knowledge of historic costume in general and coazzone in particular. The Monza catalogue (2015) was not published when I submitted my paper. The 2014 exhibition catalogue from the Galleria Nazionale in Urbino was also unavailable. To my knowledge these later publications have not yielded any conclusive evidence.

Prof. Kemp said that I have not addressed any of ‘the scientific evidence in the two books related to the lower layers of the image, the pentimenti or the condition and retouching in various media’ and goes on to say ‘contrary to Pisarek’s assertions, the interventions of restorers are documented in both books’. The latter sentence must refer to my ‘it is also strange that he did not consider that the drawing might have been retouched and repainted at a later time’ (p.79). Prof. Kemp has taken this out of context. I said that he did not mention the restorations in that particular passage of his book Leonardo (p. 210).

I not only analysed his art historical arguments in my text, but also the technical evidence presented by Cotte – which is to say:

– The trois crayons, pen and ink and bodycolour technique on vellum (unprecedented for Leonardo);

– The X-rays (inconclusive, in Cotte’s own words “did not yield significant new findings”, p. 154)

– The Carbon-14 dating of the vellum (wide-ranging 1440-1650, not constituting proof in itself as anyone could draw at any time on a blank folio removed from a manuscript)

– The quality of the vellum (rough; drawing on the hair-side; does not match the Sforziad’s smooth and well-prepared support; Birago’s illumination on the skin-side)

– The left-hand hatching (dry, timid and mechanical; on the outside of the contour of the profile, unlike in all Leonardo’s female portraits)

– The presence of three stitch holes (the Warsaw National Library Sforziad has five holes)

– The ‘knife marks’ when the folio was cut off (unnecessary, if the folio has been removed during rebinding).

– The retouchings of a later restorer (Marchig’s)

– The fingerprint evidence (no longer valid)

– The pentimenti, in the same place as in Leonardo’s Windsor portrait (a negative point, as La Bella Principessa could be based on that drawing).

At this point I would like to discuss some more scientific evidence:

On page 109:

“The support is probably the fine-grained skin of a calf”.
To the contrary, the images show an irregular grainy surface with visible follicles. Both Geddo and Turner described the support as “rough animal hide” and the surface of the vellum as being “pitted”.

“The portrait was drawn on the smooth ‘hair’ side”.
To the contrary, the hair-side has follicles so it is the rough side, not the smooth side.

Contrary to Prof. Kemp’s claim that I ignored Geddo’s contributions, I quoted her (p. 76), and she said exactly the opposite: “Besides the presence of the follicles, the rough unworked surface of the hide and its darkened, somewhat yellowish colour show that the portrait was made on the outer surface of the skin (formerly fur­covered) and not on the inner one covering the flesh, which was aesthetically the superior of the two and commonly used as a support for written documents”.

I quoted extensively from Geddo’s article “A ‘Pastel’ by Leonardo da Vinci: His Newly Discovered Portrait of a Young Woman in Profile”, in Artes, 2008–2009, pp. 67–87, on my pages 62, 76 and 88.

On page 114:

The discovered “small area of pen marks along the left edge of the support” described by Cotte as Leonardo’s “pen trials”. This would surely have no place in a drawing destined for a luxury book presented as a gift to the Sforzas.

On page 154:

A problem with the X-rays:

‘Because white chalk (calcite or calcium carbonate) does not absorb X-rays to any great extent, the luminous zones of the sitter’s face ought to have appeared grey in the X-ray. On the contrary, however, they appear very white here, indicating the presence of a significant amount of dense material in the chalks area – which seems to contradict all the physical evidence considered so far.’

Cotte attributes this anomaly to the technician supposedly over-exposing the plate. This confirms his own observation that X-rays “are vulnerable to diverse interpretation”.


[KP] I did not say that the Marchigs were involved in forgery of any description. What I did say was that Giannino was familiar with Leonardo’s technique as a restorer and a “Leonardesque painter”. He was able to make such a drawing if he had wanted to, but clearly he had not tried to sell La Bella Principessa as a work by Leonardo.

Prof. Kemp does not explain, however, why the drawing had no provenance prior to Marchig’s ownership, and, as Michael Daley has recently pointed out, Professor Kemp and the drawing’s owner, Peter Silverman jointly trawled Berenson’s archive in hope of finding some pre-Marchig record but found none.

[MK] “3) The assertion that there is an ‘almost total absence of close comparisons with unimpeachable works by Leonardo.’”

[KP] The offending phrase above was written not by me but by David Ekserdjian in “Leonardo da Vinci: ‘La Bella Principessa’ – The Profile Portrait of a Milanese Woman”, The Burlington Magazine, vol. 152, 2010, no. 1287, June (Attributions, copies, fakes), pp. 420–421.

I used the same Leonardo comparisons as Prof. Kemp, but where he saw striking similarities, I saw possible imitation.

As to Cecilia Gallerani, although the structure of the eye looks comparable, it is round, soft and alive in Leonardo’s portrait, and dry, linear and lifeless in La Bella Principessa. The iris is drawn as a flat disc and the eyelid is marked with clear cut lines, unlike in nature.

Cotte states on p. 177: “Leonardo, for example, consistently made the bottom of the eye’s iris coincide exactly with the edge of the lower eyelid”.
This is not always the case. In Portrait of a Woman in Windsor or in La Belle Ferroniere the iris does not touch the lower eyelid, while in some works by Leonardo’s followers it does.

[MK] “4) The lack of records of Leonardo making the drawing”

[KP] Prof. Kemp wrote that all Leonardo’s known works are “unrecorded in his writings”. Leonardo does in fact mention two Madonnas in a sheet of sketches (Gabinetto Disegni e Stampe degi Uffizi, Florence, 446 E) inscribed: “…1478, I began the two Virgin Marys”, possibly referring to the Benois Madonna in St. Petersburg. He also mentions in a note his sculpture of the Sforza Horse (Ms. C, fol. 15v; R 720; B 44; V 53). And in an undated letter (about 1491-95), he writes together with De Predis about being underpaid for the Virgin of the Rocks.

The 16th century record in the Zamoyski collections Kemp refers to applies only to the Sforziad, not to the drawing, which was unrecorded there.

[MK] “5) ‘The entirely unusual for Leonardo medium of vellum commonly found in manuscripts led Prof. Kemp and his colleagues, including David Wright, Emeritus Professor of Art History at the University of South Florida, to search fifteenth-century codices for an excised illumination.’”

[KP] Prof. Kemp writes that “the author’s narrative of an extensive search is imaginary.”

But Dalya Alberge wrote in her article, “Is this portrait a lost Leonardo?” in The Guardian, on 27 September 2011:

“Earlier this year, he [Prof. Kemp] embarked on what he describes as a ‘needle-in-a-haystack’ search for a 15th-century volume with a missing sheet. […] Against the odds, Kemp tracked the volume down, to Poland’s National Library in Warsaw.”

[MK] “6) Forging a Leonardo?”

[KP] To be clear, I only said that the drawing could be a compilation of Leonardo’s works and other sources such as a bust by Cristoforo Romano.

[MK] “7) Pisarek’s reliance on Julia Cartwright”

Cartwright’s work is the only one specifically on Bianca Giovanna Sforza I could find in the English language. As there are no other secure portraits of Bianca Giovanna, the Principessa hypothesis is not supported by any evidence either.

Cartwright’s identification of The Musician as the portrait of Galeazzo Sanseverino, Bianca’s husband, looks convincing. The companion portrait in the Ambrosiana could then be showing his wife Bianca, but she looks very different to La Bella Principessa. Also Professor Carlo Pedretti in his Leonardo: The Portrait, 1999, called the Ambrosiana portrait (p. 23) “a probable portrait of Bianca Giovanna, the illegitimate daughter of Francesco Sforza.”

[MK] “8) Bianca Maria Sforza and earlier scholarship”

Even if the identification and dating of the portrait ‘pre-dates the research into the Sforziad’, as stated by Prof. Kemp, there is still the problem of the dating and the too ‘archaic’ style of the drawing.

Kemp does not explain why Vezzosi and Turner identified La Bella Principessa as Bianca Maria Sforza, even if she looks so different to her other known likenesses.

[MK] “9) Cutting out the portrait from the Sforziad in Warsaw”

[KP] There is no evidence that the folio was in the Sforziad and was removed during rebinding. But if it were so, there would be no need to cut out the folio, only to remove it as a complete sheet. A complete sheet (two folios) is indeed missing in the book, which would eliminate the need for excision.

If the folio had been removed during rebinding for its beauty or high value by the Zamoyskis, it would have been recorded in their collections, but there are no such records. Kasia Wozniak’s research has not found any evidence to this effect.

Her hypothesis that the drawing went to the Czartoryski collections in Pulawy where it was identified as by Leonardo is also so far unsubstantiated. There were no such records in the Czartoryski collections. The late Director of the Czartoryski Museum, Prof. Zygulski Jn. never mentioned the existence of a Leonardo drawing in their collections.

The Bona Sforza drawing listed in the 1815 inventory of the Temple of Sybil in Pulawy and mentioned by Wozniak as the possibly misidentified Bianca Giovanna Sforza, refers to a miniature watercolour on vellum illustrated in D. Dec and J. Walek’s, Czasy! Ludzie! Ich dziela. Teatr obrazów ksieznej Izabeli, listed as no. 99. ‘Polish, 16th century’.

There is no connection between my article in Artibus and the interests of the National Library in Warsaw.

[MK] “10) The foliation and inserted paper pages”

[KP] Prof. Kemp himself used the word ‘codex’ to describe the Sforziad in his book Leonardo: Revised edition, 2011, p. 256:
“this tender and refined formal portrait in ink and coloured chalks on vellum has been cut from a codex (a book), namely the copy of the Sforziada in Warsaw produced for Galeazzo Sanseverino.”

The three folios missing in the Warsaw Sforziad were originally left blank as in the other copies of the book in Paris or London.

Geddo described the drawing’s “apparent crudeness in the preparation of the parchment” and “the rough unworked surface of the hide” (A Pastel by Leonardo da Vinci… reprinted in P. Silverman, Leonardo’s Lost Princess, pp. 219-220). This is not true of the Sforziad’s parchment.

I have seen the Sforziad on two occasions. Once in the summer 2012 and more recently in March 2016; the parchment is finely grained and of high quality, as expected in a luxury book for the Sforzas.

My illustrations Fig. 7, 10 and 11 indeed show paper pages. Because they look so similar it is easy to mistake the paper pages for the vellum ones. My Fig. 6 and Fig. 8 are vellum and they look very similar in colour and texture.

Kemp states that I inaccurately said that his “reconstruction of the insertion of the drawing in the Warsaw Sforziad looks unrealistic, as it is facing a printed page”. He wrote that “The reconstruction shows that the portrait would have faced a blank page”. This is incorrect. His “Fig. 12 – ‘Hypothetical Reconstruction of La Bella Principessa as folio 6r” in the online article “La Bella Principessa and the Warsaw Sforziad” does face a printed page.

[MK] “11) Iconography”

[KP] According to Kemp, Horodyski’s pioneering research on the Sforziad I support ‘has been superseded in the light of more detailed knowledge of Sforza court iconography and analyses by later scholars, including Wright’.

Horodyski suggested Gian Galeazzo Sforza and his offspring, such as Bona Sforza, the later owner of the book in Warsaw as the recipient of the Sforziad.

But according to D.R.E. Wright in his article on the Lumiere Technology website (which has now been removed), M. L. Evans and E. McGrath, the Warsaw Sforziad was destined for Galeazzo Sanseverino, Bianca’s husband.

So why is Galeazzo Severino’s profile absent on the Warsaw frontispiece by Birago, where Ludovico’s is in London and Gian Galeazzo’s in Paris, the recipients of the other Sforziads?

Moreover, Galeazzo Sanseverino was not part of the Sforza dynasty and Bianca was illegitimate. All the copies of the Sforziad on vellum were dedicated to members of the Sforza family.

Horodyski’s reading of the symbolic content of the Birago frontispiece more logically pointed to the death of Gian Galeazzo: the lack of the recipient’s profile as emblem, the missing figure of Gian Galeazzo in the boat with Ludovico il Moro, the tears in the handkerchief, the sarcophagus, the broken shield with the initials GZ, the crest with one half with arms of Milan and the other of Aragon for Gian Galeazzo.

After my defence of Horodyski’s interpretation, an Italian scholar Carla Glori published online her very detailed new iconographic study of the illumination: ‘The Illumination by Birago in the Sforziad incunabulum in Warsaw: in defence of Horodyski’s thesis and a new hypothesis’. I am quoting her extensively below.

She said: “The incunabula with the illuminations now in London and Florence were the property of Ludovico il Moro, given the recurrence of the central upper figure of a moor, and the presence of the ensign of the Duke of Bari with his devices called “la scopetta” (the little broom) and “I due fanali” (the two beacons). The Paris and Warsaw incunabula were the property of Duke Gian Galeazzo Sforza and his family, because they are reproducing the devices of Gian Galeazzo himself and of his father, Galeazzo Maria.”

The sieve, which was said to be the emblem of Galeazzo Sanseverino, was Gian Galeazzo’s personal device (created by his father) called “il buratto” (a sieve held by two hands) with the motto TAL A TI QUAL A MI, as illustrated by the “Cassone dei tre Duchi”, Sforza Castle, Milan, 1479-1494.
Glori also added that the sieve device (“il buratto”) is duplicated in symmetrical position in the central area of the Warsaw illumination.
She concluded that “the Warsaw illumination was dedicated to the memory of the deceased Gian Galeazzo Sforza and his family, and that it was certainly dated after his death (1494)”.

Her argument against the Sanseverino coat of arms or imprese supposedly identified in the Birago’s illumination includes:

– The missing Aragonese “A” in the device of the “three intertwined rings with diamonds” appearing on the Warsaw illumination
– The missing Sanseverino NOSTRO È IL MESTIERE motto
– The fact that the hybrid coat of arms of the Warsaw illumination does not correspond to the coat of arms of the Sanseverino dynasty; it should be silver/white not gold/yellow. “Every armorial certifies that the field (“campo”) of the Sanseverino coat of arms was “SILVER” (white), while the field (“campo”) of the coat of arms in the illumination of Warsaw is “GOLD” (yellow)”. “According to Horodyski’s logical and symbolical interpretation, the emblem is an artistic fusion of the traditional emblem of the city of Milan with the yellow and red lines of the Aragona coat of arms.”
– The reference of the initials “GZ” (appearing in the ducal documents and iconography, also on the ‘Cassone dei Tre Duchi’) to the memory of the deceased dukes Galeazzo Maria and Gian Galeazzo, not Galeazzo Sanseverino.
– The absence of any reference to Galeazzo Sanseverino and his biography such as the tournament lance of the famous jouster, while the ducal arms on the “Cassone dei Tre Duchi” are present: the round shield, the quiver with arrows and the sword. The depicted starry armour on the left is not the typical armour of a jouster. It is empty as Gian Galeazzo is dead, and it is almost identical to the one worn by him in the Paris illumination.
– The presence of a body of heraldic devices celebrating the Visconti-Sforza dynasty and referable in particular to Galeazzo Maria Sforza and his son Gian Galeazzo such as the greyhound/the tree/the divine hand (Francesco Sforza); the “capitergium” device (a bandage with a knot) dedicated to Gian Galeazzo Visconti, the first Duke of Milan, celebrating the Visconti dynasty; the rising waves (“onde montanti”); the three intertwined rings with diamonds (“i tre anelli intrecciati con diamante), by Muzio Attendolo, the founder of the Sforza dynasty: it was probably given to him in 1409 by the Marquis of Ferrara Niccolò II d’Este after the conquest of Reggio Emilia. They are not emblems of Bianca Giovanna as was advanced, but of members of the Sforza family in general.

According to Glori, “we have no evidence that the incunabulum now in Warsaw was confiscated in Milan during the French invasion. In 1517 Antonio de Beatis saw some precious incunabula in the Royal Library of Blois, but he did not cite the Sforziad as being amongst them in his autograph manuscript XF28 of the National Library in Naples. It is plausible that the incunabulum now in Warsaw was a wedding gift to Bona Sforza from her mother Isabella; I propose also the hypothesis that she received the gift from her aunt Caterina Sforza, probably when she left Milan with Isabella after the downfall of Ludovico il Moro.”

[MK] “12) Betrothal and Marriage”

[KP] Although the word ‘betrothal’ might have been more appropriate than ‘marriage’ in the case of Bianca Sforza, others also wrote that her ‘wedding’ (or nuptials) took place in 1490 or late 1489.

Julia Cartwright in her Italian Gardens of the Renaissance and Other Studies, 1914, (reprint 2013) wrote p. 174: “On the 10th of January, 1490, the wedding [of Bianca and Galeazzo] was solemnised in due splendour in the Castello of Milan (…)”.

Edward McCurdy in The Mind of Leonardo da Vinci, New York, 2013, wrote “Bianca Sforza, a natural daughter of Ludovic who in 1489, while still a child, was married to the famous captain Galeazzo di Sanseverino.” p. 301.

Wikipedia entry for Galeazzo Sanseverino also says “He was married to Bianca, illegitimate daughter of Ludovico Sforza, in 1489.”

[MK] “13) The Technique”

[KP] It was Prof. Kemp himself who said that Leonardo never worked on vellum in his book The story of the new masterpiece… p. 35: “There are no other known works by Leonardo on vellum, but there is previously neglected evidence of his interest in making coloured images on prepared animal skin.”

Turner also wrote in his online Statement concerning the portrait on vellum by Leonardo, p.3: ‘Also apparently unprecedented [for Leonardo] is the use of vellum or parchment as a support for the new portrait.’

Geddo also wrote: “The use of parchment was until now unknown in the work of Leonardo (…)”, in P. Silverman, Leonardo’s… p. 226.

In reference to Jean Perréal and dry colouring, the quotation in full is as follows: “Piglia da Gian de Paris il modo di colorire a secco e’l modo del sale bianco e del fare le carte impastate, sole e in molti doppi, e la sua casetta de’colori”.

The translation for carte impastate as ‘paste-board’ is not anachronistic, as it was used as early as 1760 in Joseph Baretti’s, A dictionary of the English and Italian languages, Vol. 1: “Cartone [composto di piu carte impastate insieme] paste-board.”

I have only inspected the vellum of the Sforziad in Warsaw, not the vellum of the drawing, but both Geddo and Turner have described it as ‘rough animal hide’. This is most certainly not what you will find in the Sforziad in Warsaw.

[MK] “14) Dimensions”

[KP] According to Kemp and Cotte, the dimensions of the vellum pages of the Sforziad vary from 33.0 to 33.4 cm in height, while the drawing is 33 cm high.

I have carefully checked the dimensions with the Librarian in March 2016. All the pages are at least 33.4 cm high and more, up to 33.7 cm. The size of 33 cm would be far too small for the book.

The 5 holes in the book are in fact all double holes. Each of the 5 holes is two small holes, between which a string passes. The distance between the two small holes is about 3 mm. The double holes were never mentioned by Kemp or Cotte.

According to the conservator who was present at the time of my last visit, this is the binding that follows the original binding as there is no damage of any kind. So in total there were as many as 10 small holes, not 3 single ones as in the drawing.

I measured the distances between the 3 holes that Kemp and Cotte measured in La Bella Principessa. The measurements were taken from the middle of the double holes.

The distance between the bottom hole and the middle hole is 11.35 cm in the Sforziad, while in the drawing it is 11.06 cm.

The distance between the middle hole and the top hole is 11.7 cm in the Sforziad, while in the drawing it is 11.44 cm.

[MK] “15) The profile and the cartoon portrait of Isabella d’Este”

Above, Fig. 4: A comparison of La Bella Principessa with Leonardo da Vinci’s, Portrait of Isabella d’Este, c. 1499–1500, Paris, Musée du Louvre.

[KP] The portrait of Isabella d’Este shows the face in profile but the body in three-quarters, unlike La Bella Principessa. The former is also unfinished, rendered softly with the sfumato effect, fluid in execution, while the shading is on the inside of the profile.

La Bella Principessa is shown in full profile, highly finished, rigid and linear, while the shading is on the outside of the profile.

The similarities between the two profiles are only superficial.

The use of the word ‘repaint’ was an incorrect translation of the French word ‘repentir’, which means pentimento. I would like to mention that I have a good enough grasp of drawing techniques as I also trained as a copyist of Old Masters and an art restorer.

[MK] “16) Left-handedness”

[KP] I disagree that “the left-handed execution cannot undermine the attribution”, as it indicates the intention to imitate Leonardo. None of his collaborators or followers were left-handed, so the drawing is either by him or by an imitator/forger.

[MK] “17) The costume”

Above, Fig. 5: Comparisons of La Bella Principessa with Leonardo da Vinci’s, Head of a Woman, c. 1488–1490, National Gallery, Parma (left); Leonardo’s Portrait of a Woman in Profile, c. 1489–1490, Windsor Royal Collection (right); and, Gian Cristoforo Romano’s sculpture, Bust of Beatrice d’Este, c. 1491, Paris, Musée du Louvre (top).

[KP] The simplified and flatly rendered dress as well as the coazzone hairstyle do indeed show similarities with the sculpted busts by Gian Cristoforo Romano (Bust of Beatrice d’Este) and Francesco Laurana. But this could be a negative point, as the drawing could be based on one of these busts.

Incidentally, the opening in Laurana’s sculptures differs from that in La Bella Principessa. In the former it is a wide horizontal cut facilitating the movement of the arm, while in the drawing it is a triangular hole which doesn’t seem to play such a role.

Above, Fig. 6: Gian Cristoforo Romano, Bust of Beatrice d’Este, c. 1491, Paris, Musée du Louvre.

Above, Fig. 7: A comparison of the knots on La Bella Principessa’s dress and Gian Cristoforo Romano’s Bust of Beatrice d’Este, c. 1491.

[MK] “18) The fingerprint”

[KP] The fingerprint evidence which was originally published in the book as “strongly supportive of Leonardo’s authorship” is now considered invalid.

It was not possible to compare the palm imprint to Leonardo’s other examples, and it was described as perhaps unintentional as it is single and isolated, unlike in the execution of Cecilia Gallerani, where many imprints were found where the blending of hues had taken place.

[MK] “19) On Method”

[KP] The “accumulative build-up of different types of evidence” against the attribution to Leonardo is strong, but, as mentioned above, my main arguments were not addressed by Prof. Kemp.

Why is the shading on the outside of the profile? What is the significance of the hand writing on the reverse of the drawing and why was it not investigated? Why is the profile of La Bella Principessa so similar to that in the sculptural Bust of Beatrice d’Este by Cristoforo Romano? Why is the knot on the dress similar to the one on the bust and unlike other Leonardo’s knots? Why are the proportions of the face flawed? Why the vellum of the drawing was described as ‘rough animal hide’ and could it be part of the luxury book of the Sforziad? Why Marchig’s friend the famous Renaissance expert Bernard Berenson had not attributed the drawing to Leonardo? Why there is no known provenance prior to Marchig’s in the 1950s?

In the field of attributions the level of inconsistencies and contradictions always undermine any evidence in favour of a proposed attribution.

[MK] “20) The damaging allegation in the opening to Pisarek’s article that the owner was to set up ‘non-profit-making foundation for multi-disciplinary Classical and Renaissance studies near Florence, to be headed by Professor Martin Kemp’.”

[KP] I made no such allegation. This was a quotation from an article by Simon Hewitt who supports the attribution to Leonardo. This information was published in an Antiques Trade Gazette article by Simon Hewitt in 2009. The article can be found here.

In my article I also wrote: “Prof. Kemp and his colleagues are no doubt genuinely convinced of the authenticity of the drawing, as well as highly enthusiastic about the rediscovery.” This shows that I in no way question Prof. Kemp’s integrity on this matter, only the methodology and the results of the proposed attribution.

Kasia Pisarek, 24 May 2016

Connoisseurship in Action and in Peril

“WHEN THE FIRST catalogue of the Ashmolean Museum’s Sculpture collection (in three volumes) was undertaken more than twenty-five years ago, the present author decided to exclude works made before 1540…”

“…The decision was partly determined by the quantity of the material but also in recognition of the special attention that the finest of the small bronzes given and bequeathed by Charles Drury Fortnum deserved. It is indeed because of these that the Ashmolean’s sculpture collection is the most important in the United Kingdom after that of the Victoria and Albert Museum, London…”

Above and below: The ‘Fortnum Venus’, attributed to Francesco Francia or his circle. c. 1500-05. Bronze, 26.1 cm. high. (Ashmolean Museum, Oxford).

Above, detail of St Jerome, by Cosimo Tura. c.1470. Panel, 101 by 57.2 cm. (National Gallery, London).

“The first of Jeremy Warren’s three volumes under review here includes these very figures. One of them is the
exquisite Venus associated with the Bolognese painter and goldsmith Francesco Francia (cat. No.20). It is one of the earliest responses in the Renaissance to the antique female nude, and perhaps belonged to a narrative group for, as her drapery falls, her hand is cupped as if to receive the apple from Paris…”

“…In conclusion it should be noted that the cataloguing of the permanent collection is not now often considered an essential part of the curator’s job. Any museum director who is chiefly an impresario is inevitably attentive to fashionable ideas and taste, whereas the cataloguer of a permanent collection is bound to give his or her attention to neglected artists, and to works by artists in which they are not initially interested, but which beg the question as to why the appealed to their predecessors – collectors, scholars and curators long deceased. This provides an antidote not only to the narrow outlook of the popular exhibition machine, but to that of academic institutions where scholars, having achieved promotion by imitating their seniors, are obliged to compete in appealing to the young consumers upon
whose favour the prosperity of the institution depends.”

~ NICHOLAS PENNY: “Sculpture in the Ashmolean”, Burlington Magazine, January 2016. A review of Catalogue: Medieval and Renaissance Sculpture. A Catalogue of the Collection in the Ashmolean Museum, by JEREMY WARREN, 3 vols., continuously paginated, 1188 pp. incl. numerous col. + b. & w. ills. (Ashmolean Publications, Oxford, 2014, £395.)


“It’s a new dawn, it’s a new day… what new art exhibitions will make you feel good? Read our list of the top five shows opening this season to get inspired.

“Discover the enduring impact of Botticelli and Delacroix on the art world and prepare to be star struck at exhibitions celebrating 100 years of British Vogue and the incredible beauty of our Solar System.

“Need a second opinion? Watch art historian Jacky Klein’s guide to the season’s unmissable art exhibitions.”

~ THE ART FUND, a mailing, 8 December 2016.


“Museums have become places where we take part in social as well as learning activity. It is easy to be cynical about
the impact of the café, restaurant or shop spaces on the culture and character of museums, but such facilities have made museums less daunting, more welcoming and more open to general visitors. However, such [democratization] needs to go deeper than the provision of opportunities to purchase or to consume.”

NICHOLAS SEROTA, the director (- since 1988) of the Tate, at the Leeum Samsung Museum of Art in Seoul.


“One of the most influential books in the twentieth-century art history was Erwin Panofsky’s Studies in Iconology. In this seminal work, the very latest German approaches and method were presented to an Anglo-Saxon audience for the first time. One of the dazzlingly learned chapters was devoted to the figure of Cupid wearing a blindfold, and Panofsky showed how this theme could be traced back to the writings of the medieval moralists. For thinkers of this Christian stamp, lovers were metaphorically blind, since they were ‘without judgement or discrimination and guided by mere passion’. But later, in the Renaissance, ‘moralists and and humanists with Platonizing leanings’ contrasted the figure of the blindfold Cupid with another kind of…”

~ PAUL TAYLOR. Introduction, CONDITION The Ageing of Art, Paul Holberton Publishing, 2016. ISBN 978 1 907372 79 7

Illustration: Jean-Auguste-Dominique Ingres, The Composer Luigi Cherubini with the Muse of Lyric Poetry, 1842, detail.

“…Although Bode was extremely learned in many branches of art history and had his information always ready – he had little need of written notes and photographs – he was no scholar in the true sense. He was more fond of action than reflection, and was far from having a just and objective mind. Moreover, the distracting practice of uninterrupted daily work hindered a comprehensive synthesis of realization. The deepest motive of the scholar, namely, disinterested love of truth, could not function productively in a nature always resolutely aiming for effectiveness and visible results. Bode was utterly unphilosophical, he considered no affair from more than one side. He saw black or white, good or bad, advantageous or inimical; he knew no intermediary steps. To forsee obstacles, to fear, to guard against them was not his way. Consequently he became angry as soon as he encountered opposition. He made more mistakes by acting than by failing to act. He was a hunter, not an angler.”

~ Max J. FriedländerReminiscences and Reflections, Edited from the literary remains and with a foreword by
Rudolf M. Heilbrunn, 1969, New York.

“…Anyway, getting back to the book that was sent to me, its title was rather grand and pompous: La bella Principessa – the beautiful princess. Or, as I knew her, ‘Bossy Sally from the Co-op’. I’m a bit unsure of how to talk of this because the book was written by an eminent Oxford professor and must have been quite an effort. I don’t want to ruffle any feathers or cause problems but I nearly swallowed my tongue on reading its supposed value – £150 million! It would be crazy for a public body to pay such a sum. So I feel the need to say something about it.

“The drawing is thought by some to be a work by Leonardo da Vinci, but it does divide opinion and it wasn’t included in the National Gallery’s Leonardo show of 2011, a show which I thought was really well done except for it being staged ‘underground’ in the Sainsbury Wing basement…

“I drew this picture in 1978…It was done on vellum, quite a large piece to find unfolded and without crease lines. I did it on vellum because at that time I couldn’t make old paper yet…The first thing I had to do was sand off the writing with 600-grit wet and dry paper. That done, it looked too new for anything old to look right on it, so I turned it over and did the drawing on the other side. That is why the drawing is done on the hair side of the vellum instead of the much-preferred ‘flesh side’. The texture of the sanding should still seen on its reverse.

“As I said, the face is of 1970s vintage, and I think that shows in the drawing…The drawings of Leonardo and Holbein especially have always impressed me with their fineness of line and detail, and in my view they must have been done under some magnification…The vellum is mounted on an oak board…before drawing on it, the vellum was stuck to the backboard with cabinet maker’s pearl glue, so it needed to be under a weighted press for a while to allow the glue to go off without ‘cockling’ the vellum. ‘Cockling’ is the effect you see on paper when you try to paint a watercolour without soaking and stretching the paper first. On vellum the dampness looks like blisters or a cockle bed on the shore. It’s caused by the water content of the glue, so the thing needs to be under a heavy press to dry it flat.

“After a bit of experimentation, and just to prove a point to myself, I lightly traced the drawing I’d invented onto the vellum (I’m sure the graphite can still be detected) and started to draw the image in hard black chalk – carbon black in gum arabic – using a pair of jeweller’s magnifying glasses. It took some time to get used to working like that, and I had to go to back to practising on papper for a while so as not to bugger things up…

“It was done in just three colours – black white and red – all earth pigments based in gum arabic, with the carbon black mostly gone over with oak gall ink. To be a bit Leonardo-like or even Holbein-like – they were both left-handers – I put in a left-hander’s slant to it…The Leonardo book [“La Bella Principessa ~ The Story of the New Masterpiece by Leonardo da Vinci”, By Martin Kemp and Pascal Cotte, with contributions by Claudio Strinati, Nicholas Turner and Peter Paul Biro] seems to put great store by the apparent leftyness of the drawing, but it can be shown up very easily. With the face on the vellum facing left, just turn the drawing clockwise to face her skyward, and hatch strokes from profile outwards in the normal manner…Incidentally, the book points out a palm print on the neck area, just the spot a right-hander doing an impersonation of a left-hander might rest their hand whilst doing the background hatching.

“Although I am no Oxford professor, I could list umpteen reasons for not thinking this drawing to be by leonardo…
The book mentions several holes on one margin as evidence that it has been bound into a volume, and also mentions some later ‘restoration’. I did not do these things, and don’t know who did, or where it went on its later travels. Looking closely at the picture in the book, it looks to have had the left margin peeled back an inch or so and has been restuck, not very well, especially at the bottom left. Could this be from when the left margin was pierced and roughly re-cut by someone else?

“I sold it for less than the effort that went into it to a dealer in Harrogate in late 1978 – not as a fake, or by ever claiming it was something it wasn’t. I can’t really say any more on it. At least it may now be known for what it is.”

~ SHAUN GREENHALGH – A Forger’s Tale, Published by ZCZ Editions in 2015

In the past it was customary for scholars to advance claims of attribution for particular works of art in scholarly journals and then wait to see how peers and colleagues responded to their evidence and arguments. Increasingly, we are seeing co-ordinated promotional campaigns of advocacy by players and owners who eschew venues of debate and sometimes denigrate sceptics and opponents. We invited two of the leading advocates of “La Bella Principessa’s” Leonardo’s authorship – Martin Kemp and Nicholas Turner – to speak at our recent conference on connoisseurship. Both declined. Neither our post of January 2014 (“Art’s Toxic Assets and a Crisis of Connoisseurship ~ Part II: Paper – sometimes photographic – Fakes and the Demise of the Educated Eye”) nor our colleague Kasia Pisarek’s article “La Bella Principessa: Arguments against the attribution to Leonardo” that was published in the June 2015 issue of the scholarly journal artibus et historiae have been challenged in print.
(For a pdf copy of Kasia Pisarek’s article, please write to: For our conference, see Art, Law and Crises of Connoisseurship and Recap of Art, Law and Crisis of Connoisseurship Conference.)

The da Vinci Detective: Art Historian Martin Kemp on Rediscovering Leonardo’s Tragic Portrait of a Renaissance Princess – an ARTINFO Interview by Andrew M. Goldstein, 17 October 2011. Extracts:

…[Andrew Goldstein] “They weren’t the only ones to differ on the attribution of the painting, and when you first announced that you believed it to be a Leonardo, a lot of people disagreed. One museum director even told the Telegraph’s Richard Dorment, anonymously, that it was a “screaming 20th century fake, and not even close to Leonardo himself.” Has there been any reversal since then?

[Martin Kemp] “I don’t know who this anonymous person was, but we carbon-dated the parchment and that eliminates it from being a screaming modern forgery. If it were a forgery, it used things that we’ve only recently discovered about Leonardo’s technique in the last 20, 30 years. The fact that it was owned by Giannino Marchig takes it outside the period when it could be a forgery, knowing what we know now, so that’s not an option. The ultra-violet turns up retouching, and it’s very clear it has been heavily restored, but most objects 500 years old, including the “Salvator Mundi”, which is the new picture being shown in the National Gallery.

[…] “One thing that critics of your ‘Principessa’ attribution tend to bring up is the involvement in your research of Peter Paul Biro, a fingerprint expert whose credibility was questioned. What is your opinion of him?

“Well, Biro I knew of as someone who’d specialized in fingerprints and paintings, so we asked him to look at the fingerprint that is in the upper left side of the ‘Bella Principessa.’ I had data on finger prints and finger marks in other Leonardo paintings, and he said one of these matched – not astoundingly, because it’s just the tip of a finger, and one doesn’t rely on fingerprints on vellum. It wouldn’t convict anybody in a court of law. You need more than that. So he did a limited job here, and we didn’t depend too much on that evidence. The press liked it, of course because it was cops and robbers stuff.

“I would not now probably say much about it at all, because on reflection I don’t think we have an adequate reference bank of Leonardo fingerprints. I’ve talked to fingerprint specialists, and they typically require a full set of reference prints. We don’t have that for Leonardo. My sense is – and this is Pascal’s sense, too – that it’s probably premature, given what we know about Leonardo’s fingerprints, to come up with matches at all. But the job Biro did was perfectly straightforward. There were no grounds for dishonesty. Peter Paul Biro is suing the New Yorker*, but I can’t comment at all upon the court case because that’s about things that I know nothing about, so it’d be totally improper. But he did work for us, which I now, let’s say, place less reliance on, simply because, on reflection, I think the fingerprint evidence is rather slippery.

“Because of the work you have done to bring the ‘Principessa’ into the fold of acknowledged Leonardos, some say you have crossed from the realm of scholarship to something more like advocacy. How do you explain your passion for the portrait?

“I would say that one of the differences between being a historian of art and being a scientist, as I was trained, is that you’re dealing with objects that are deliberately communicating with something other than just our intellect. So, for me, it’s not a dry process. You begin with the feeling it’s special, and if it stands up to the research, you end with the feeling that it’s special, and I make no apology for that. I’ve been critized as acting as an advocate for it, but if I’m writing, as I am in ‘Christ to Coke’, about the ‘Mona Lisa’, I’m an advocate for that too, because it’s a miraculous picture. Also, when I’m writing about the Coke bottle, I’m not an advocate for Coke as a drink. I hate it. But it’s one of the all-time great
bits of product design and I’m happy to say that.”

“*THE MARK OF A MASTERPIECE” by David Grann, The New Yorker, 12 July 2010 ~ Extracts:

[1] “But he [Martin Kemp] also relies on a more primal force. ‘The initial thing is just that immediate reaction, as when we’re recognizing the face of a friend in a crowd,’ he explains. ‘You can go on later and say “I recognize her face because the eyebrows are like this, and that is the right colour of her hair,” but, in effect, we don’t do it like that. It’s
the totality of the thing. It feels instantaneous.'”

[2] “Moreover, according to [the curator of Drawings at the Metropolitan Museum of Art, Carmen] Bambach, there was a more profound problem: after studying an image of the drawing [La Bella Principessa] – the same costume, the same features, the same strokes that Kemp examined – she had her own strong intuition. ‘It does not look like a Leonardo’, she said.”

[3] “When such a schism emerges among the most respected connoisseurs, a painting is often cast into purgatory. But in January, 2009, Kemp turned to a Canadian forensic art expert named Peter Paul Biro who, during the past several years has pioneered a radical new approach to authenticating pictures. He does not merely try to detect the artist’s invisible hand; he scours a painting for the artist’s fingerprints, impressed in the paint on the canvas. Treating each painting as a crime scene, in which an artist has left behind traces of evidence, Biro has tried to render objective what has historically been subjective. In the process he has shaken the priesthood of connoisseurship, raising questions about the nature of art, about the commodification of aesthetic beauty, about the very legitimacy of the art world.Biro’s research seems to confirm what many people have long suspected: that the system of authenticating art works can be arbitrary and, at times, a fraud.”

[4] “Biro asserted that he had uncovered the painting’s ‘forensic provenance,’ telling a reporter, ‘The science of fingerprint identification is a true science. There are no gray areas.’ Having developed what he advertised as a ‘rigorous methodology’ that followed ‘accepted police standards,’ he began to devote part of the family business to authenticating works of art with fingerprints—or, as he liked to say, to ‘placing an artist at the scene of the creation of a work.'”

[5] “But the International Foundation for Art Research, a nonprofit organization that is the primary authenticator of Pollock’s works, balked, saying that Biro’s method was not yet ‘universally’ accepted.

[6] “In 2009, Biro and Nicholas Eastaugh, a scientist known for his expertise on pigments, formed a company, Art Access and Research, which analyzes and authenticates paintings. Biro is its director of forensic studies. Clients include museums, private galleries, corporations, dealers, and major auction houses such as Sotheby’s. Biro was also enlisted by the Pigmentum Project, which is affiliated with Oxford University.”

[7] “Biro told me that the divide between connoisseurs and scientists was finally eroding. The best demonstration of this change, he added, was the fact that he had been commissioned to examine ‘La Bella Principessa’ and, possibly, help make one of the greatest discoveries in the history of art.”

[8] “After he first revealed his findings, last October, a prominent dealer estimated that the drawing [‘La Bella Principessa’] could be worth a hundred and fifty million dollars. (The unnamed ‘lady’ who had sold it at Christie’s for less than twenty-two thousand dollars came forward and identified herself as Jeanne Marchig, a Swedish animal-rights activist. Citing, among other things, the fingerprint evidence, she sued the auction house for ‘negligence’ and ‘breach of warranty’ for failing to attribute the drawing correctly.)”

[9] “Ellen Landau, the art historian, said that she was ‘absolutely convinced’ that the paintings were by Pollock. Biro was sent a photograph of a fingerprint impressed on the front of one picture. He identified six characteristics that corresponded with the fingerprint on the paint can in Pollock’s studio—strong evidence that the work was by Pollock. But, as more and more connoisseurs weighed in, they noticed patterns that seemed at odds with Pollock’s style. Meanwhile, in sixteen of twenty art works submitted for analysis, forensic scientists discovered pigments that were not patented until after Pollock’s death, in 1956. At a symposium three years ago, Pollock experts all but ruled out the pictures. Ronald D. Spencer, a lawyer who represents the Pollock-Krasner Foundation, told me, ‘Biro can find all the fingerprints he wants. But, in terms of the marketplace, the Matter paintings are done. They are finished.'”

[10] “Reporters work, in many ways, like authenticators. We encounter people, form intuitions about them, and then attempt to verify these impressions. I began to review Biro’s story; I spoke again with people I had already interviewed, and tracked down other associates. A woman who had once known him well told me, ‘Look deeper into his past. Look at his family business.’ As I probed further, I discovered an underpainting that I had never imagined.”

[11] “During the eighties and early nineties, more than a dozen civil lawsuits had been filed against Peter Paul Biro, his brother, his father, or their art businesses. Many of them stemmed from unpaid creditors. An owner of a picture-frame company alleged that the Biros had issued checks that bounced and had operated ‘under the cover’ of defunct companies ‘with the clear aim of confusing their creditors.’ (The matter was settled out of court.) As I sifted through the files, I found other cases that raised fundamental questions about Peter Paul Biro’s work as a restorer and an art dealer.

[12] “Biro refused, multiple times, to divulge where he had obtained either of the paintings. According to the Wises, Biro insisted that the person who sold him the paintings was in Europe, and that it was impossible to contact him.

[13] “Sand sought proof of a financial transaction—a check or a credit-card payment—between Biro and Pap. Biro, however, said that he had obtained them in exchange for two musical instruments: a Steinway piano and a cello.

[14] “Sand was incredulous: ‘Is Mr. Pap a music dealer or is he an art dealer?’ After Biro could not recall where he had originally purchased the cello, Sand suddenly asked him, ‘You ever been convicted of a criminal offense, sir?’


‘You are certain of that?’

‘Yes,’ Biro said.”

[15] “Throughout the trial, the Biros and their attorneys maintained that the two paintings sold to the Wises were authentic, but to make their case they presented an art expert who was not a specialist on Roberts, or even on Canadian art. On September 3, 1986, the court found in favor of the Wises, and ordered Peter Paul and Geza Biro to pay them the seventeen thousand dollars they had spent on the pictures, as well as interest.”

[16] “Lawsuits had piled up against Peter Paul Biro and his family business. In two instances, there were allegations that art works had vanished under mysterious circumstances while in the care of Peter Paul. In one of the cases, Serge Joyal, who is now a senator in Canada, told me that he left a nineteenth-century drawing with the Biros to be restored. Before he could pick it up, Peter Paul notified him that it had been stolen from his car and that there was no insurance. Biro, however, never filed a police report, and Joyal says that Biro pleaded with him to wait before going to the authorities. During their conversations, Joyal says, Peter Paul acted evasive and suspicious, and Joyal became convinced that Biro was lying about the theft. As Joyal put it, ‘There was something fishy.’ Though Peter Paul said that there was nothing ‘suspect’ about his behavior, and that he should not be held liable, the court awarded Joyal seven thousand dollars, plus interest.”

[17] “Within Montreal’s small art world, there were whispers about Peter Paul Biro and his father. But the lawsuits appear to have attracted virtually no public attention. In 1993, Peter Paul Biro filed for bankruptcy, and he never paid many of the judgments against him, including what he owed the Wises and Joyal. Lipsz’s lawyer said of Biro, ‘He oiled his way out of that whole thing. . . . He got away scot-free.'”

[18] “Biro was part of an effort to launch a venture named Provenance, which would provide, as he put it, the ‘clever strategy’ necessary to sell ‘orphaned’ paintings for tens of millions of dollars. According to a business prospectus, marked confidential, Provenance would acquire art works that had been forensically validated by Biro and several colleagues, and sell them in a gallery in New York City. The company chose a thumbprint for a logo.”

[19] “Provenance was cleverly tapping into the public’s desire to crack open the art world, offering the tantalizing dream that anyone could find a Pollock or a Leonardo or a Turner languishing in a basement or a thrift shop. The company combined the forensic triumphalism of ‘C.S.I.’ with the lottery ethos of ‘Antiques Roadshow.’ (An associate producer at ‘Roadshow’ had already sent Biro an e-mail about possibly doing a segment on the Parkers’ ‘unbelievable discovery.’)

[20] “Biro previously had been suspected of creating an investment scheme around a seemingly precious object, with the promise that it would eventually reap huge profits. In the late nineteen-nineties, he persuaded a Canadian financial adviser, Richard Lafferty, who is now dead, to invest in a venture to authenticate and sell a work purportedly by Raphael’s disciple Perino del Vaga. Three of Lafferty’s colleagues confirm the story, as do letters, memorandums, and other documents.”

[21] “By the fall of 2005, Ken Parker had begun to look into the people behind Provenance. It turned out that Tod Volpe, in the nineties, had defrauded his art clients, including Jack Nicholson, of nearly two million dollars, and had served two years in prison. Parker discovered that one of Volpe’s principal partners in Provenance was also an ex-con, who had done time for tax evasion and for running a drug-smuggling operation in the United States. (Volpe told me, ‘We all have skeletons in our past.’) Parker confronted Biro, who, in a subsequent e-mail, told Parker that he had ‘severed all communication with Volpe.’”

[22] “And only months after rescinding his request for money he asked the Parkers to fund another new project: a privately endowed department for him and a colleague at Oxford University. ‘Naturally it is 100% tax deductible,’ Biro wrote, in an e-mail. ‘Those who support the foundation of a bold and new department for us at Oxford will have their name on a plaque or have the department named after them such as “The Ken Parker Department for Forensic Art History.” Sounds cool?'”

[23] “When a forgery is exposed, people in the art world generally have the same reaction: how could anyone have ever been fooled by something so obviously phony, so artless? Few connoisseurs still think that Han van Meegeren’s paintings look at all like Vermeers, or even have any artistic value. Forgers usually succeed not because they are so talented but, rather, because they provide, at a moment in time, exactly what others desperately want to see. Conjurers as much as copyists, they fulfill a wish or a fantasy. And so the inconsistencies—crooked signatures, uncharacteristic brushstrokes—are ignored or explained away. “

[24] “In the case of ‘La Bella Principessa,’ Biro did not handle the drawing, and was sent multispectral images from another laboratory, which he then developed and enhanced. Martin Kemp, the Leonardo scholar, told me, ‘In terms of what Biro did for us, I have absolutely no problems with any potential ethical issues.’ He emphasized that his opinion of the drawing did not depend on the fingerprint evidence: ‘I’m entirely confident that it is by Leonardo.'”

[25] “A final verdict on whether ‘La Bella Principessa’ is genuine may not be reached for years, but more and more connoisseurs have voiced doubts. Skeptics express surprise that there is no apparent historical record for the drawing, given that Leonardo was one of Italy’s most famous painters during the Renaissance. They note that vellum lasts for centuries, and that it would be easy for a forger to obtain old sheets. Many of the critics share the view of the Met’s Carmen Bambach: it just doesn’t look like a Leonardo. ARTnews, which has reported on Wertheim’s findings, recently interviewed Klaus Albrecht Schröder, the director of the Albertina Museum, in Vienna. ‘No one is convinced it is a Leonardo,’ he said. David Ekserdjian, an expert on sixteenth-century Italian drawings, wrote in The Burlington Magazine that he ‘strongly suspects’ it is a ‘counterfeit.’ Other art critics have suggested that Kemp has succumbed to a fantasy.

[26] “Whereas Biro had once spoken of the absolute objectivity and infallibility of fingerprint analysis, he now sounded more like a connoisseur than like a scientist. ‘I’m trying to define, for example, what is the point that something becomes a matter of interpretation,’ he said. ‘In other words, where is that line? O.K., on the one hand, fingerprint practitioners state that fingerprint identification is a science. I’m more toward the other side, where I’m convinced by my own personal experience that it is very much like connoisseurship, because of . . . things I see they don’t.’”

[27] “I asked him whether he might have been wrong in suggesting that Leonardo had ever touched ‘La Bella Principessa.’ He looked up at the sky and said, ‘It’s possible. Yes.’”

In July 2011 ADWEEK reported that Biro had sued The New Yorker and David Grann for $2 million on twenty-four charges of false and malicious defamation – Forensic Art Expert Peter Paul Biro Sues New Yorker for Deformation. The action failed on every count (see “Art Authenticator Loses Defamation Suit Against The New Yorker).

CODA: Faking A Picasso and a Provenance

In June last year, we and Martin Kemp were asked (separately) by the Daily Mail to comment on a claimed Picasso painting that supposedly had been found by an artist among his late mother’s belongings in an old suitcase in the loft – “Is this a long-lost Picasso? Painting that bears a strong resemblance to Cubist’s work is discovered rolled up in a battered suitcase that hasn’t been touched in 50 years” – the Mail reported:

“…Last month, a 1955 Picasso painting – Les Femmes d’Alger – broke auction records when it sold at Christie’s for a staggering £115 million.

“Art expert Michael Daley told MailOnline his first instinct was that the painting was not a Picasso. ‘The colours and all of that are right but I think there is too much incidental detail and not enough decisive interest or exploration. This is
all sharp edges and brown tones and what-have-you, but it doesn’t have the driving sense of Picasso getting at something, getting at a figure trying to see something about it. This is more decorative. It can be no more than an instinctive first impression, but to me it doesn’t look like a Picasso Cubist painting from that period.’

“Martin Kemp, emeritus professor of art history at Oxford University, was also skeptical. Prof Kemp said: ‘It’s pretty close to Picasso’s portrait of Daniel-Henry Kahnweiler, the dealer whom he painted in 1910 and is in the Art Institute of Chicago. This looks to be a relatively competent but not excellent near copy of it – the signature is very dodgy. Without doing a full art historical analysis – and I’m not a Picasso expert – I would say someone has made a variant of the Picasso portrait. Also the signature with the rounded S’s doesn’t seem to correspond to Picasso’s way of signing and thirdly the canvas doesn’t look quite right for Picasso of that period.
But I hope for the sake of the owner that I’m wrong.'”

On 6 July 2015 The Scotsman disclosed that the discovered ‘Picasso’ was a fake (“Picasso painting in Fife attic ‘a hoax’”). The faker, Dominic Currie, reportedly described his purported-Picasso (which had been about to be inspected by Christie’s) as “a piece of performance art [made] in order to raise awareness of the struggling artists in Scotland.” The fake is now being offered for sale on the website of a picture restoration business.

Michael Daley. 10 December 2016.

12 December 2016. Postscript: “Bode gave an opinion on art works from many fields, on Netherlandish painting of the seventeenth century, on Italian sculpture, Persian rugs, majolica, German wood sculpture, and many other things. His writings were for him always a means to an end, and were frequently controversial and advocatory. Almost always he judged correctly, and at the time he expressed them, his communications were of significance and infinitely enriched knowledge in many fields. But the tragic fate of the ‘connoisseur’ lies in the fact that yesterday’s new, strikingly accurate definitions are today’s common properties and banalities, that only the mistakes linger in the memory under the name of the originator.”

~ Max J. Friedländer on Wilhelm Bode – Reminiscences and Reflections.

The Perpetual Restoration of Leonardo’s Last Supper, Part 2: A traumatic production of “a different Leonardo”

14 March 2012

The unhappy $8m Olivetti-sponsored restoration of Leonardo’s Last Supper began in 1977 with a repair of small flakes of detaching paint. It morphed, on very grand institutional technical advice, into a promised liberation of all of Leonardo’s surviving original paintwork. It ended after twenty-two years amidst widespread recriminations and as a distinctly mongrel work showing alarmingly little original paint and very much alien “compensatory” and “reintegrating” new paint.

Prospective major restorations are often presented as elegant technical answers to some urgent conservation necessity the resolution of which promises magnificent artistic gains. In reality, the interface between technical intervention and artistic outcomes constitutes art restoration’s fault-line and does so in a field that is notoriously subject to the law of unanticipated consequences. One of the commonest surprises is how greatly the coherence of a work had depended on earlier restoration repairs that were removed on the grounds of being alien impurities. Like Humpty Dumpty, radically stripped works often prove to be wrecks that have to put back together again and many a restorer discovers – too late – that it is easier to take to pieces than to reassemble [see Endnote 1]. The resulting changes made during restorations are often presented as “discoveries”, “recoveries” or “recuperations” when on close examination they prove to have been plain errors. One such unwarranted, unsupported, insupportable case is shown here.

Towards the end, the restorer, Pinin Brambilla Barcilon, expressed the hope (Art News March 1995) that her restoration might be the last because: “The less you restore a work of art, the better its chances of survival. Each time you touch a work, it suffers a trauma, no matter how carefully you operate.” In November 1998 the Art Newspaper reported that when Mrs Brambilla was restoring a crucifixion on the wall opposite the Last Supper in 1978, she noticed that “fragments of painting were peeling off Leonardo’s work before her eyes.” The Guardian of July 21 1997 reported she had noticed “bits of painting falling from the Last Supper” and that after experts from Italy’s central institute for restoration in Rome (Istituto Centrale di Restauro) had been called in “the decision to restore the painting was swiftly taken”. Too swiftly, many Italian experts believed.

Although conservation necessities are sometimes exaggerated [see Endnote 2], with paint losses, determining the extent, cause, and remedy must always be the top priority – the equivalent of fixing a building’s leaking roof. As seen in Part 1, when the Last Supper was disintegrating to the touch after the Second World War, the then restorer, Mauro Pelliccioli, fixed the problem by embedding the paint in litres of shellac. He also won some critical praise for uncovering most of Leonardo’s own surviving paint from restorers’ over-paint. Crucially, however, he tackled the disintegrating paint first (during 1947-49). Only when the shellac was settled and the paint completely secure did he begin scraping off restorers’ repainting (during 1952-54). In this most recent restoration, despite the problem of paint detachment, work began with an intended systematic removal of the remaining repaints.

Effectively, in the last restoration the authorities undertook an all-or-nothing gamble with a masterpiece. Against the certainty that shedding the old would be disruptive of the familiar and the still-surviving, they bet that the recovery of some more fragmentary, talismanic relics of Leonardo’s paintwork would outweigh the scale of accompanying losses and newly exposed bare wall. This presumptuous naivety was to prove disastrously wrong-headed. First, as scientific tests of paint fragments (published in Studies in Conservation, August 1979) were to warn, the distinction between original paint and later restorers’ overpaint was not at all easy to establish: “the dividing line is much less clear cut”. (This was hardly surprising given the work’s earlier exposure to corrosive cleaning agents and heated metal rollers.) Second, Pelliccioli had already uncovered most (two thirds, he believed) of what was taken to be Leonardo’s surviving paint. While there was not all that much more to recover, there was, artistically, still very much to lose. Pelliccioli had left repaints in place precisely where they covered only bare wall – which is to say, where they held the image together.

The paradoxical consequence of this pursuit of original material was that the old ill-preserved yet somehow-maintained “theatrical” illusion that Leonardo had originally created was greatly undermined. One narrow specialised purist concern for what was “original” and “authentic” material was set against another larger more elusive aesthetic/artistic concern for what had been intended; for what was yet struggling to survive. Achieving the liberation of fragmentary and injured archaeological material imperfectly adhering to a damaged moisture-prone vertical surface came at the cost of eliminating all that had maintained and prolonged Leonardo’s decaying but originally mesmerising artistic illusion incorporated into the space and fabric of a large room. (Kenneth Clark had spoken fondly of “these ghostly stains upon the wall”.) The work was remorselessly stripped down to the sum total of all previously accumulated injuries in order that the resulting wreck might then be put back into some presentable aesthetic form more suited to today’s tastes. The ideological/art historical rationale offered for this purgative exercise was that every age has the right to make its own Leonardo…that the Leonardos that had come down to us from the past were somehow deficient, obsolete, culturally-contaminated; that we now simply know better. The preposterous nature and Futurist flavour of this relativist conceit (“Every previous generation has erred, we, standing outside of history – or at its end – will now get things right”) might have been held self-evident. Leo Steinberg, evidently unsettled by this recent spasm of historical/aesthetic cleansing, quoted Jonathan Swift: “Last week I saw a woman flayed, and you will hardly believe how much it altered her person for the worse.” (“Leonardo’s Incessant Last Supper”, 2001.)

The purist shedding of earlier repaints regardless of their antiquity and artistic functions, necessarily guaranteed that Leonardo’s work would be “altered considerably”, as Carlo Bertelli, the then director of the restoration, acknowledged in the catalogue to the 1983 Washington National Gallery of Art exhibition of Leonardo’s studies for the Last Supper. A later defender of the restoration, Giovanni Romano, not only applauded the creation of “a different Leonardo” during “a great restoration” (Il Giornale dell’arte, April 1999) but fawningly added that he would be “satisfied with a restoration of this sort every year.” Throughout this era of vaultingly high ambition, the restoration community needed the biggest possible “sell”: nothing less than a “New Michelangelo” was said to have emerged during the Sistine Chapel ceiling restoration. It became a commonplace that revolutionary restoration “discoveries” required the very “rewriting of art history”.

There is no mystery about how the latest calamity came about. In the 1983 catalogue, the Washington National Gallery curator, David Allan Brown, duly relayed the twin official reasons for the restoration: that the paint (which, under a microscope, resembled “the scaly skin of a reptile” ) had not remained secure; and, that Pelliccioli had not removed all earlier repaints. The repaints had to come off because they were now “threatening the stability of the original colour.” Not explained, was how this was so, or how much original paint had been left by Pelliccioli. In the July 2 1995 New York Times, Bertelli recalled having been “certain that there was enough beneath the additions to warrant this restoration”. He added that “Mrs Brambilla and I had examined the surface with a microscope, and we were surprised to see how much of Leonardo’s original work remained”.

In 1983 Allan Brown noted that “The expectation that a considerable portion of the original might survive [had given] a strong impetus to the decision in the late 1970s by the Superintendency of Fine Arts for Lombardy (at that time directed by Franco Russoli) in consultation with the Istituto Centrale di Restauro, to take up again the unfinished work of cleaning the picture.” This would suggest that so strong was the desire to revive and complete Pelliccioli’s unfinished aesthetic programme that the operation was begun even as the technical solution which had originally made that aesthetic objective possible was said to be failing. Whatever gains might have been hoped for or anticipated, by May 1998, Bertelli admitted (in Art News) that “Now we can see only a few square feet [of original Leonardo paint] but they are by the master” and, on a Channel 4 documentary that year (“The Lost Supper”), he characterised Leonardo’s mural as a “ruin” (“una rovina” ). In the November 1998 Art Newspaper Brambilla said, self-contradictorily, that the repaint had had to be removed out of fear that condensation might become trapped between “the artist’s original paint and the successive layers of paint”, and, that constant environmental conservation measures would henceforth be necessary because “the layers of repainting are no longer protecting the original paint.”

Which was the case? If the repaint was protecting the Leonardo paint, what had been causing the original detachments? Given that the 1979 tests mentioned above had established that it was not always possible – even under ideal laboratory conditions – to “decide exactly on the dividing line – both for areas and for layers – between what remains of the original and materials pertaining to later interventions”, how great could the risk have been of moisture insinuating itself between the original and the subsequent paint layers? On whatever technical premises it rested, when the restoration proper began on the better preserved right-hand side of the mural, the attempted removal of all previous restorers’ repaints and consolidations of paint, inevitably constituted a prolonged and sustained assault on the mural’s fabric – as Brambilla herself candidly described in the March 1995 Art News:

Here we have a surface that is completely ruined, disintegrated into tiny scales of colour that are falling off the wall. We have to clean each one of these scales six or seven times with a scalpel, working under a microscope…Here I can clean an area one day and still not be finished, because when the solvent dries it brings out more grime from beneath the surface. I often have to clean the same place a second time, or even a third or a fourth. The top section of the painting is impregnated with glue. The middle is filled with wax. There are six different kinds of plaster and several varnishes lacquers and gums. What worked on the top section doesn’t work in the middle. And what worked in the middle won’t work on the bottom. It’s enough to make a person want to shoot herself.”

Could Pelliccioli’s already failing shellac have survived these repeated traumatic assaults with solvents and scalpels on all the glues, waxes, lacquers and gums within the paint-film? Had some new superior quick-acting consolidant been identified or manufactured? What were the structural consequences of this apparent removal of every atom of previous consolidations of the paint? Brambilla has said of the detaching paint “To re-adhere the fragments we used wax-free shellac in alcohol, the same adhesive as Pelliccioli applied during his intervention of 1947″. So, in other words, just some more of the same. If Brambilla’s best English wax-free shellac lasts no longer than Pelliccioli’s, we might expect another restoration within twenty years or so.

One thing is clear: the technical underpinning of the restoration, and the swiftness with which its unquestionably radically transforming methodology was applied, were both challenged by Italian experts. On July 2 1995 the New York Times reported that Mirella Simonetti, a Bologna-based restorer, protested: “There was never any doubt in their minds. They decided how to proceed without even conducting the proper analyses to determine how much of the original painting remained. They didn’t even submit their findings to an international committee of experts.” The Florence-based diagnostician, Maurizio Seracini, who had been called to examine the Last Supper after the restoration began, complained: “I think that Mrs Brambilla has worked in good faith. But you don’t decide to restore a masterpiece like the ‘Last Supper’ on the basis of what you see under a microscope. It’s simply irresponsible.” Seracini added “I myself have not seen any definitive scientific proof that restoration was really needed.”

That the authorities had not known how much original Leonardo paint might survive had been tacitly acknowledged as early as 1983, when, with the restoration one third completed, David Allan Brown could speak only in relative terms: “By comparison with other, well-preserved murals of the time, Leonardo’s detailed execution is almost entirely lost.” Even when the restoration was eventually finished (or halted) there was no agreement among the protagonists themselves on how much had survived. Carlo Bertelli, the director of the Brera Art Gallery in Milan, who effectively initiated the restoration, put the figure at 20%. Pietro Marani, the prolific Leonardo scholar who advised Brambilla from 1985 and became co-director of the restoration in 1993, once said that “no more than 50%” survived and later more ambiguously claimed that 90% had survived “in parts”. Giuseppe Basile (later the director of the restoration of Giotto’s Arena Chapel frescoes) put it at “about half”. Giorgio Bonsanti, the director of the Florence-based laboratory Opificio delle Pietre Dure, put it at “possibly 20%”. Giovanni Urbani, the director of the Istituto Centrale di Restauro between 1973-83, and the director of the Brancacci Chapel restoration, thought 25% had survived.

In 1989 a Milan town councillor, Maria Bonatti, brought an (unsuccessful) action against the restorer for accelerating the mural’s decay – a charge also made by the painter, essayist and paint materials expert Mario Donizetti, who held that it would “disintegrate more rapidly than before.” Ten years before the restoration ended, in November 1989 the Art Newspaper reported “Over the years work has been stopped repeatedly, sometimes following changes at the helm of the Milanese Soprintendenza and the Instituto in Rome, other times simply to allow the whole project to be reconsidered.” Eventually the proceedings quickened dramatically. In his 2001 book, Leo Steinberg recalled encountering the restorer and three young assistants in 1998 “all huddling at lower left scraping away”. On the cleared wall, “more filling was needed – and it had to be done fast (a deadline had been imposed from on high), so that this must-see tourist attraction would show decent finish to the daily sightseers.”

In 1983 Bertelli had said in National Geographic that Brambilla was taking a week to clean an area the size of a postage stamp. He quoted her professional plaint: “It’s difficult. The work is hard and tiring. It creates much physical tension bending over the microscope. After a few hours my eyes grow blurry. I may come every day for months. Then I must take an extended break. There is also the psychological tension. All the eyes of the world that know Leonardo are watching what I do. Some nights I do not sleep.” The pressure intensified as the restorer inched her way towards the central figure of Christ. In April 1998 the Art Newspaper reported that “Hundreds of tourists (mostly Japanese) last month lined up…to visit Leonardo’s Last Supper. The painting was back on view after having been closed for two months to allow restorers to work on the faces of Christ and the apostles.” That report evidently escaped the attention of the National Gallery’s then director, Neil MacGregor, who wrote in his 2000 BBC book “Seeing Salvation”:

When the latest restoration was unveiled in 1999, all hell broke loose, and the admirably scrupulous restorer in charge was vilified in much of the world’s press…Among the wilder accusations, fears were expressed that the face of Christ had been altered. Happily these proved to be groundless.”

That the unveiling was badly received is beyond dispute, but if vilification was in evidence it was aimed by defenders of the restoration at their critics (see right). The extent to which the face of Christ was altered, and the evolving means by which the restorer came to impose her own distinctive, ahistorical, arguably arbitrary aesthetic reading on the unprecedentedly vast, fully-exposed areas of paint loss in a quest to “bring it back to its original colors and composition”, will be examined in Part 3. Here (right) we examine a single unwarranted change that was made to the design of drapery on Christ’s right arm and then presented as a restoration recovery.

Michael Daley

ENDNOTES [1] In the 1962-64 National Gallery Annual Report, the then director, Sir Philip Hendy, described how with the great Uccello panel from the Battle of San Romano series, the stripping down (which had begun and 1959 and was still ongoing) had exposed a greatly damaged surface. As a result, after its characteristic Gallery “complete cleaning“, it was realised that “To restore scrupulously takes very much longer than to create freely, and the task of pulling the picture together again could have been further prolonged.” [2] With the restoration of Michelangelo’s Sistine Chapel ceiling it was said in 1986 (six years into the restoration) that “various checks [had] ascertained that in several places minute flecks of colour were lifting” and that this had “necessitated an immediate restoration.” In 1987 it was said that extensive areas of flaking were progressively worsening and threatening an imminently “uncontrollable situation”. By 1988 Vatican spokesmen were claiming that the weight of encrustations upon the paint surface was causing it to break away from its ground. By 1989 it was said that the glues had “shrunk and puckered” causing “scabs” to fall away “pulling pigment with them”. It was said that this “slow destruction by glue-pox” was “the Vatican’s principle motivation for cleaning the ceiling“. When I asked in 1990 how big the puckerings were, a Vatican spokeswoman said “Oh! Some are as big as your hand.” Soon after, in 1991, the problem de-escalated: initial investigations were acknowledged, once more, to have encountered “minute desquamations and loss of pigment.”

Comments may be left at:

Printer-friendly PDF version of this article

Above, Fig. 1: The opening section of TIME magazine’s June 7th 1999 coverage of the unveiled Last Supper. Such reports generated public denunciations of the critics. One, by Georgio Bonsanti in the May 1999 Art Newspaper, targeted this organisation:
At the root of the controversy are the declamations of the international brigade, Art Watch International, now nearly ten years old, led by the American art historian James Beck and the English journalist [sic] Michael Daley. These crusaders are convinced that masterpieces are being desecrated by poor restoration all over the world but especially in Italy…”
Bonsanti further claimed that Italy’s restorations were being attacked by outsiders because that country “tolerates dissent”, evidently forgetting that James Beck (a professor of art history at Columbia University, who lived in Italy for part of each year, whose wife was Italian and whose children are half Italian) had been hounded simultaneously in several Italian cities on charges of criminal slander (which carried, in addition to horrendous potential financial penalties, the risk of a three-years jail sentence) for criticising the restoration of a masterly piece of sculpture by the subject of his life-long studies, Jacopo della Quercia. (For Michael Daley’s reply, see “Was it necessary?”, the Art Newspaper, July/August 1999.)
In a more specifically abusive vein Bonsanti alleged that:
All standards of accuracy and correctness were laid aside in the scramble to report the supposedly disastrous restoration of the Last Supper…I am not suggesting that critics should be silenced or that a moratorium is desirable, but criticism should be based on observable fact, it should be a technically competent, well-researched and accurate reaction.”
As it happens, all of our criticisms are based on observable, demonstrable (and photographically reproducible) facts. How else, in visual art, might criticisms proceed? As a case in point we discuss and illustrate below a change that was made to the Last Supper and then presented as one of a number of “noteworthy recoveries” when, manifestly, it was not a recovery but an adulteration. To judge restorations it is necessary to know what was done and what was said to have been done by those in authority. For many of those of us who had the privilege of seeing from the scaffold what was being done to Leonardo’s work, the experience was alarming. The removal of everything except that which was deemed original Leonardo paint meant that greatly more “not-Leonardo” was emerging than Leonardo. The second consequence of the stripping was that what little original paint survived, was lost in the visual clamour of the surrounding cracked and discoloured wall preparation layers. It would have been inconceivable to leave the tiny emerging islands and archipelagos of original paint adrift in the vast sea of ruined wall. At close quarters it was evident that this was not only an extreme restoration, it was also methodologically self-defeating in its dogmatic pursuit of “pure” material. As fast as earlier restorers’ paint was eliminated, fresh paint was needed to ameliorate the losses and impart a pale impression of continuity, coherence and legibility (“The goal of pictorial integration was to achieve a sufficient legibility of gesture, pose and modeling” – Brambilla). The purging of paint and the attendant debilitation of imagery gave rise – as is soften the case with radically deconstructing cleanings – to a misconceived, ahistorical reconstruction that altered the design of the picture – in this case, even, that of Christ himself. While doing so in the name of historical authenticty, historical testimony was defied to a degree that beggars belief.
The change in question was made to the sleeve of the tunic on Christ’s right arm. It is shown here at Figs. 5-9, as it appears in the “official” published accounts. As seen in Figs. 6 and 8, the alteration was made with fresh paint to stripped areas. On the “evidence” of some fragments of red paint located in a clearly distressed section of the mural, the sleeve drapery was extended by fresh repainting so as to come to rest on the top of the table which it had originally tucked behind – as countless drawn, engraved and painted copies of the Last Supper testify. By a fluke of publishing, we can show directly comparative photographs that capture the genesis of this alteration. They are found in two books. The first is Pietro Marani’s sumptuous, large format 1999 “Leonardo da Vinci ~ The Complete Paintings” (hereafter: Marani 1999). In it there is a photograph of Christ (shown here in Figs. 6 & 8). It was taken when the figure had been stripped down and largely but not entirely retouched. The second source is the beautifully photographed and produced 1999 book-of-the-restoration “Leonardo ~ The Last Supper”, by Pinin Brambilla Barcilon and Pietro C. Marani with Antonio Quattrone’s photographs (hereafter: Brambilla/Marani 1999). In this book there is another photograph of the Christ but this time it is after all the retouching had been completed. This plate is seen at Figs. 7 & 9. The differences between the two states are worth a thousand words. But before discussing their testimony in detail, the status of the many copies of the Last Supper should be considered. We take as fair testimony of Leonardo’s original treatment of the drapery, copies of the Last Supper made in the first eleven decades of its life. Fig. 2 is a detail of an engraving thought to have been made within a couple of years of the Last Supper’s completion. Although primitive in style, it clearly shows that the sleeve drapery is cut off by the table top and does not not rest upon it. Over a century later, Rubens (or an associate) made a copy in ink and wash. Here too the sleeve drapery is cut off by the table.
Above, top left, Fig. 2: Detail of the engraving given to Giavanni Pietro da Birago and thought to be the earliest copy (c. 1500) of the Last Supper which was completed in 1498. This work comprises a visual record of the Last Supper’s appearance before the mural’s notoriously rapid physical decomposition was in train. Enlarged, the clarity of delineation attests to a feature that may not be so evident elsewhere: the (blue) mantle does not yet simply disappear in the zone of shadow behind the table and between the arm and torso (as is already recorded at Fig. 4), but is seen to turn briskly around the waist of Christ. Above, top right, Fig. 3: Detail of an ink and wash copy given to Rubens or an associate, and of c. 1600-08. It is possible that the Rubens copy may have been made not from Leonardo’s Last Supper in Milan but from the large full-size copy of it seen at Fig. 10. But that copy, too, attests to the drapery being cut off by the table. This table/drapery relationship had thus remained unchanged for more than a century. Above, Fig. 4: Detail of the 1616 oil on canvas copy by Il Vespino (Andrea Bianchi). This copy had been specifically commissioned in 1612 to record the then condition of the already alarmingly decaying mural.
Above, Fig. 5: Leonardo’s Christ, as seen before the 1977-1999 restoration. It is clear when comparing this state with the copies of the painting shown at Figs. 2, 3, 4, 10 & 11, that in the course of earlier restorations, the drapery of Christ’s right arm had been slimmed down on its inside edge by encroachments of the shaded zone between the table top and the figure.
The key role served by certain copies in the last restoration might be mentioned. Pietro Marani (in Brambilla/Marani 1999) discusses the assorted values of the many and various copies as testimony. He gives a list of fifty principal and “more faithful copies”. Pre-eminent among these are the two full-size oil on canvas copies of the Last Supper. One is by Giampietrino, a student of Leonardo’s (see Fig. 11). It is given to c. 1520. The second is given by Marani (on not very clear grounds) to an anonymous Flemish artist, but it was formerly attributed to another Leonardo follower, Andrea Solario. For convenience we refer to it as the Tongerlo copy, after the Abbey in Belgium which has owned it since 1545. A third large copy is the oil on canvas by Il Vespino, shown above at Fig. 4. At the time of last restoration of the Last Supper the Il Vespino had recently been restored and put on display in Milan at the Ambrosiana. The two full-size canvases played critical roles in the last restoration. The Tongerlo copy had been damaged in 1929 and was restored in 1932, 1952 and again in the 1990s prior to being housed in what is known as the Da Vinci Museum. The Giampietrino (owned by the Royal Academy) was borrowed as an aid to the last restoration and cleaned to that end by Pinin Brambilla Barcilon. Marani says of the copies generally that their value “cannot be overestimated”. From the full-size Giampietrino and Tongerlo copies, Brambilla had taken tracings – and these, Marani reports (Brambilla/Marani 1999), demonstrated:
…how the painters faithfully traced the most important elements of Leonardo’s work – the heads, hands and other principle outlines of the various figures – but then assembled the pieces into into complete images that did not always take into account the original intervals between the figures and the painted spaces in the mural.”
That the two full-size copies shared essential design properties both with one another as well as with Leonardo’s mural is of the utmost importance. Such a triangulated matching of imagery shows that the Tongerlo and Giampietrino copies were taken directly from the mural itself after completion, and were not taken from cartoons that might have been made in preparation for the painting during the course of which revisions were made. This in turn means that the restorers between 1977 and 1999 had in their possession unquestionably reliable guides to the designs and the component parts of Leonardo’s figures as they had been finished by him on the mural itself – if all three matched, none could be inaccurate. There could be no grounds for departing from the commonly held “principle outlines of the various figures”.
Above, Fig. 6: Christ, as seen (in Marani 1999) before the completion of the 1977-1999 restoration. This illustration captured a curious stage of restoration. We see that after the stripping of earlier restoration repaints, the extensive losses to the blue drapery of the mantle and the red drapery of the tunic had been ameliorated by overall applications, respectively, of lighter blue/grey and lighter red paints. Of critical importance is the fact that at this stage a crescent shaped, as-yet untinted zone, sits unresolved between the red and the blue passages, as if there is confusion about where to place the boundary between the coloured zones of the two draperies. A reason for hesitation is not hard to divine: this untreated area not only sits between the red and the blue draperies, it also also runs down across the table cloth.
Above, Fig. 7: Christ, after the completion of the 1977-1999 restoration (as in Brambilla/Marani 1999). It can now be seen that the junction of draperies at the inside of the red sleeve and the blue mantle in Fig. 3 was moved rightwards by an extension of the pale red paint. But at the same time, this light red retouching has also been carried downwards, overlapping the table cloth before turning upwards so as to terminate behind the wrist, and thereby impart to the sleeve drapery a muff-like (or puffball-like) configuration from the centre of which the forearm now emerges. As well as being an unwarranted falsification of Leonardo’s design, this change insinuates a solecism: in the laws of artistic drapery, material hangs from and partially expresses underlying human forms, it does not provide autonomous enclosing receptacles for them (like pots for a lobster, as it were). As shown below, no authority exists in the painting’s many copies for this change of design.
Above, Fig. 8: An enlarged detail, again showing the sleeve/mantle relationship after some tinted infilling had taken place but before the completion of the restoration. The dark shape in the bottom left corner is the mantle of the apostle John. It would seem that this photograph was taken at the point when the restoration was just about to pass to the left of Christ. It would seem also to confirm that the tinted infills were being made pretty much as the stripping down was taking place. Here we can see that the razor-sharp delineation of the architectural forms is not a happy by-product of the stripping down but is almost entirely a subsequent reconstruction effected with superimposed overall painting on a zone of almost total losses of original paint. It looks as if the tinted brown repaint on the background wall was stopped short of the not-yet stripped drapery of John, thus producing the temporary effect of a coarse outline or halo.
Below, Fig. 9: The section at Fig. 8 after completion of the restoration. Note how the repainted tones on the architecture have been brought to a sharp and precise relationship with the contour of the arm drapery. Is that seeming precision of draughtsmanship authentic or spurious? How well does it compare with the sleeve/wall boundary on the two full-size copies shown below? Of this back wall, Pinin Brambilla Barcilon writes:
The entire wall, however, was characterised by the widespread loss of colour…Once the superficial repaint had been removed, it was clear that we had to press on with cleaning in order to achieve some sort of visual coherence between the back wall and the side walls…Pictorial integration meant that the wall had to be restored using coats of dark toned water colour to define the shadows. Obviously the flakes of original colour remain completely visible…”
Of the sleeve drapery the restorer writes:
Like the mantle, Christ’s robe also required the removal of extensive repaint, as the heavy red tone detracted noticeably from the lovely original vermilion passages. The thick adhesives made the removal particularly difficult, so it progressed with the repeated application of compresses which managed to dissolve the film of glue completely. The cleaning redefined the original articulation of the folds, both at the neckline and at the sleeve cuff. The cuff which had been covered by the mantle’s repaint now revealed beautiful violet flakes, composed of a blue base glazed with a crimson lake to define the shadow area…”
The very concept of “redefining” the original is unsettling if not oxymoronic. Note that while there is excitement at the discovery of “beautiful violet flakes” (as seen through a microscope?) in the stripped down wreckage, there is no mention of the table cloth or discussion of the changed drapery/table relationship. The restorer continues:
Other noteworthy recoveries included…”
Above, Fig. 10: Detail of the Tongerlo Abbey copy of which Steinberg writes:
“…given its size, its high quality, and general accuracy, the Tongerlo copy ranks with the finest surviving testimonies to the near-lost Leonardo.”
Above, Fig. 11: Detail of the large, c. 1520 copy of Leonardo’s Last Supper that is owned by the Royal Academy and attributed to Giampietrino. In the catalogue to the National Gallery’s recent “Leonardo da Vinci – Painter to the Court of Milan” exhibition, Minna Moore Ede wrote:
Given the deteriorated state of Leonardo’s Last Supper mural today, the question of which of the early painted copies can be said to be most faithful to the original is of particular and tantalising importance. Always viewed as among the most accurate is this scale copy by Giampietrino…Believed to have been a live-in apprentice of Leonardo’s during his first Milanese period (probably joining the workshop in the mid 1490s), Giampietrino would have been present during the period when Leonardo was preparing and painting the Last Supper, perhaps even assisting his master.”
When told of our objection to the redrawn sleeve of Christ, Pietro Marani reportedly responded: “A small piece of drapery. Oh, my God.” and contended that Giampietrino might have misunderstood the position of Leonardo’s drapery.
Click on the images above for larger versions. NOTE: zooming requires the Adobe Flash Plug-in.