285 The Cotswolds Rembrandt

A country art auction in England made the front pages all over the world when 2.2 million pounds was paid for a painting that looks a lot like a Rembrandt self-portrait. Is it? Schwartz thinks it is, and supplies an analysis to explain why. At the same time, he shows how the published opinions of the Rembrandt Research Project could have led to the rejection of the painting by the experts consulted by the owner and the auction house. More like an article than a column.

Laughing man (Rembrandt?) in a gorget. Whereabouts unknown


On October 26th an unusual painting was sold at a country auction house in the Cotswolds, Moore Allen & Innocent. The subject looks for all the world like a fanciful self-portrait of the young Rembrandt, a laugh on his face and a gorget of the kind he never had to wear in real life around his neck. The painting, in oil on copper, is totally unknown in the literature on the master.

Lot 377, “The young Rembrandt as Democrates [should have been Democritus] the laughing philosopher” was attributed in the auction catalogue to “A follower of Rembrandt,” with an estimate of £1,000-1,500 pounds. Because higher bids were submitted before the sale, bidding that day opened at £5,800.

The auctioneer’s website tells the rest: “At times, the bids were rising in increments of £100,000. By the time bidding reached £330,000 only three dealers in the room – all on mobile telephones to anonymous clients – were in the running. When the bidding hit £1 million an eerie silence fell over the auction room. A little later, one dealer asked for the painting to be removed from its glass display case and studied it closely. ‘It’s definitely a Rembrandt,’ he was heard to tell his client. ‘It’s a bargain at £1.75 million.’ When the price reached £2 million silence fell again, broken only by a few nervous giggles. Then the dealers started bidding again, in increments of £50,000. As the hammer fell at £2,200,000 there was spontaneous applause as everyone present realised they had witnessed a piece of art history.”

One informed observer thought that price too was a bargain. Jan Six, the newly appointed head of Old Masters at Sotheby’s Amsterdam, was quoted as saying, “Nobody pays 2.2 million [pounds] for a follower of Rembrandt. If this was a known Rembrandt and was published in 20 books and had a great provenance it would go for 10 million.” Six had spent a week studying the painting, on behalf of a Sotheby’s relation. He was convinced that it was by Rembrandt himself, not a follower.

Could Jan Six be right that adverse circumstances had cost the vendor eight million pounds? This is not impossible. In the recent past Sotheby’s itself showed how a promising dark horse could be helped to overcome handicaps of that kind – not being in the right books, not in the right collections – and still achieve top prices in the market. On 7 July 2004, a painting that had previously been considered a forgery or late imitation of Vermeer was knocked down by Sotheby’s in London for £16.2 million. If it had been thrown cold into a sale, it would not have fetched much more than one-thousandth of that amount. The success of the sale was due entirely to the superior research and pr campaign of Greg Rubenstein, head of Old Master drawings at Sotheby’s London. Rubenstein spent years gathering, publishing and promoting evidence for Vermeer’s authorship of a painting that on first sight makes a feeble impression. At the beginning of his campaign, the painting was looked down on by virtually everyone in the field. By the time of the sale, enough specialists agreed that the painting is by Vermeer to convince Steve Wynn and an underbidder that it was worth more than £16 million.

What are the chances that a pre-sale campaign on behalf of the Cotswolds copper could have succeeded to the same degree? They are very good, I believe, and easier to accomplish than in the case of the Vermeer. The painting is not feeble-looking at all, and as I argue below, is more likely than not an original by the master. Had a proper dossier on the painting been assembled and distributed, even Moore Allen & Innocent, who are more used to ricks than Rembrandts, would have been able to sell it for 10 million or more.

What kind of art history was it then that the attendees of the sale witnessed? Why was the laughing man not given a fair chance? According to Moore Allen & Innocent, the painting was consigned by a regular customer of the auction house, who told them that it had “languished in the attic” for 50 years. “The vendor had told me that his father had the picture evaluated and the conclusion was that it was not a Rembrandt,” said auctioneer Philip Allwood. “Deciding to do some more research on the painting,” continues the BBC News story on the event, “Mr Allwood spoke to the Rijksmuseum in Amsterdam. Experts there assured him that, while it was of the period of Rembrandt, it was ‘probably not’ painted by him.”

If this is true, it is too bad for the owner and the auctioneer. Too bad as well that they allowed themselves to be discouraged so easily by these negative reactions. The owner, still deterred by his father’s unfavorable experience, went along with an absurdly low estimate. The auctioneer, apparently convinced that there was no point in trying to improve on the judgment of the Rijksmuseum, played down the attribution and the estimate, but he did put the painting on the cover of the catalogue. Had only one serious buyer noticed it and believed it to be by Rembrandt, it would have gone for less than £10,000. Fortunately for the owner, three parties joined in. They kept the news to themselves, so that it sold for less than other recent Rembrandts on the auction block.

 

Why should the experts consulted by the owners’ father and the auctioneer not have seen it coming that there would be serious interest in this painting? The following analysis proceeds from the assumption that they based their opinions on the Corpus of Rembrandt paintings of the Rembrandt Research Project (RRP). Although the Cotswolds copper is not mentioned there, six other paintings are catalogued that are comparable to it in one significant way or another, either in motif or technique. Of these, only one was initially accepted by the RRP as a Rembrandt. Two are rejected outright, two others were originally doubted and only recently reinstated, a fifth is included as a copy of another version of the same composition. If the experts took these judgments seriously, it is no wonder that they responded so discouragingly to the query.

Before turning to those works, we should note that paintings of this description – head-and-shoulders paintings of young men in armor, whether or not laughing – were being painted in the period around 1630 not only by Rembrandt but also by other artists in his immediate environment.

   

Rembrandt (as I believe)
Self-portrait in gorget, laughing
Signed RHL
Ca. 1629-30
Oil on copper, 23.7 x 17 cm.
Sale Cirencester, 26 October 2007, lot 377

Jan Lievens
Portrait of Rembrandt van Rijn in gorget

Signed I.L. Ca 1630
Oil on panel, 57 x 44.7 cm.
Amsterdam, Rijksmuseum (on loan)

Jan Lievens, a contemporary and close associate of Rembrandt’s in Leiden, portrayed Rembrandt himself in a gorget. However, the conception of the subject is so different, Lievens’s Rembrandt is so much more detached and dignified, that we are not inclined to think of Lievens in the first place as author of the new painting.

   

Rembrandt
Self-portrait in gorget, laughing
Signed RHL
Ca. 1629-30
Oil on copper, 23.7 x 17 cm.
Sale Cirencester, 26 October 2007, lot 377

Isaac de Jouderville
Young man in gorget, laughing
Ca. 1629-30
Oil on panel, 52 x 49 cm.
The Hague, Bredius Museum

A closer match in terms of composition is to be found in a painting in the Bredius Museum in The Hague attributed to Rembrandt’s early pupil Isaac de Jouderville. A comparison between the two reveals, to my eye, a considerable difference in quality. The Cotswolds copper is marked by subtle transitions in delicate brushwork, compared with the broader swabbing in the painting in The Hague and the resultant exaggerated effects. The painter of the copper succeeds in conveying an insouciant, transient laugh. By comparison, the Jouderville figure has a forced and rigid look. Nonetheless, the two compositions are amazingly similar. One possible explanation is that de Jouderville was painting a portrait of himself in a gorget, laughing, in imitation of his master’s example.

   

Upper left: Rembrandt
Self-portrait in gorget, laughing

Signed RHL
Ca. 1629-30
Oil on copper, 23.7 x 17 cm.
Sale Cirencester, 26 October 2007, lot 377

Upper right: Rembrandt
Laughing man in gorget
Signed Rt.
Ca. 1629-30
Oil on copper, 15.4 x 12.2 cm.
The Hague, Mauritshuis

Below: Jan Gillisz. van Vliet after Rembrandt
Laughing man in gorget
Signed JG Vliet fec. RHL inventor
Ca. 1634
Etching, 22.6 x 19 cm.
Haarlem, Teylers Museum

More satisfying comparisons are to be found in the six works referred to above. With a Laughing man in gorget in the Mauritshuis, also on copper, the new work shares not only the tilt of the head and the open mouth with teeth, which we also see in the Jouderville, but also the specific form of the gorget and the reflections on it, as well as the long line of the right cheek against the long lock of hair. (The Corpus of Rembrandt paintings, vol. 1, p. 229, calls this a “cadenette”; the French Wikipedia terms it a hair style favored by grenadiers and hussars, named for Honoré d’Albert, seigneur de Cadenet.) The strong modeling of the facial forms is also more pronounced in these two works than in either of the above.

The powerful resemblances between these two well-wrought paintings on copper of military men laughing justifies to my mind the assumption that they were both conceived and possibly executed by the same artist. What artist may that have been? In the judgment of the Mauritshuis, which I second, the artist is Rembrandt. This attribution is not only a conclusion of present-day connoisseurship. It is also based on the existence of a print by Jan Gillisz. van Vliet, a close associate of Rembrandt’s, reproducing the Mauritshuis painting (in reverse, as usual), with the inscription RHL inventor. Since 1634, the attribution to Rembrandt has held sway. This however did not dissuade the RRP in 1982 from casting doubt on Rembrandt’s authorship, for vague stylistic and qualitative reasons. If the experts consulted on the painting in the attic followed the conclusion of the Corpus, the resemblance would sooner have given them reason to doubt rather than embrace the attribution to Rembrandt. The RRP places the Mauritshuis painting in its B category, with the tortuous title: “Paintings Rembrandt’s authorship of which cannot be positively either accepted or rejected.” (In vol. 4 of the Corpus, p. 627, Ernst van de Wetering writes “We no longer doubt the work’s authenticity.)

   

 

 

Rembrandt
Self-portrait in a cap, laughing
Signed and dated RHL 1630
Etching, 5 x 4.4 cm.
Bartsch 316, third state of six
Amsterdam, Rijksmuseum

Rembrandt
Self-portrait in gorget, laughing
Signed RHL
Ca. 1629-30
Oil on copper, 23.7 x 17 cm.
Sale Cirencester, 26 October 2007, lot 377

Rembrandt (as I believe)
Self-portrait, laughing

Signed RHL
Ca. 1629-30
Oil on panel, 41.2 x 33.8 cm.
Amsterdam, Rijksmuseum

Even worse for the status of the Cotswolds copper is its more general resemblance to a panel in the Rijksmuseum itself. Until 1982, this painting of a young man laughing hung in the galleries as a self-portrait. In that year, the first volume of the Corpus of Rembrandt paintings appeared, which relegated the Laughing Rembrandt to the C class: “Paintings Rembrandt’s authorship of which cannot be accepted,” and it was taken off to storage. This painting too is related, though not as directly as the Laughing man in gorget, to a print. In this case to the Rembrandt etching Self-portrait in a cap, laughing. The RRP not only fails to see Rembrandt’s hand in the Rijksmuseum panel but also his face. They see in it “traditional features of the rather feeble-minded fool type.” Although Arie Wallert of the Rijksmuseum is inclined to give the painting back to Rembrandt, the RRP is not.

 

Rembrandt, Self-portrait in gorget
Ca. 1629
Panel, 38 x 31 cm.
Nürnberg, Germanisches Nationalmuseum

After Rembrandt
Rembrandt in a gorget
Ca. 1629
Panel, 37.9 x 28.9 cm.
The Hague, Mauritshuis

If there is reluctance to admit that Rembrandt painted himself laughing, no such doubt exists concerning his self-portraits in military getup. The closest example of this type is a self-portrait in gorget in Nürnberg. The pose and mode are more restrained than in the copper, but the coiffure, with its cadenette, closely resembles that of the Cotswolds figure.

       

Rembrandt
Detail of Self-portrait in gorget, laughing
Signed RHL
Ca. 1629-30
Oil on copper, 23.7 x 17 cm.
Sale Cirencester, 26 October 2007, lot 377

Rembrandt
Detail of Self-portrait in gorget
Ca. 1629
Panel, 38 x 31 cm.
Nürnberg, Germanisches Nationalmuseum

Rembrandt
Detail in reverse of Laughing man in gorget
Signed Rt.
Ca. 1629-30
Oil on copper, 15.4 x 12.2 cm.
The Hague, Mauritshuis

Here too, as in the laughing soldier in the Mauritshuis, I am struck by the way in which the reflections are placed on the gorget – the squarish block on the upper collar, the diluted light on the expanse of the surface, the narrow arc of light on the lower ridge and the highlighted rivet heads. These are stylistic particulars that do not have to be that way. See the very different treatment of light on the gorgets in the paintings by Lievens and Jouderville above.

 

Needless to say, similarity of motif is no proof of identical authorship. The Corpus of Rembrandt paintings published as an original a very faithful copy in the Mauritshuis after the painting in Nürnberg. Even so, the similarity of the new discovery to the paintings in The Hague and Nürnberg justifies at the least the hypothesis that Rembrandt was the intellectual author of the new painting. That in itself should be reason enough, considering the stakes involved, to consider with great care the possibility that he executed it as well.

   

Rembrandt
Self-portrait
Signed RHL 1630
Oil on copper, 15 x 12.2 cm.
Stockholm, Nationalmuseum

Rembrandt
Old woman praying
Signed R
Ca. 1629
Oil on copper, 15.5 x 12.2 cm.
Salzburg, Residenzgalerie

Rembrandt
Soldiers at a fireplace
Signed (with some reservation on the reading) RHL 1628
Oil on copper, 22.1 x 17.1 cm. (possibly cut down on the left)
Tokyo, Bridgestone Museum of Art

Rembrandt or Rembrandt workshop
A man reading by candlelight
Unsigned, undated. Ca. 1628-29
Oil on copper, 13.9 x 13.9 cm.
Milwaukee, collection of Alfred and Isabel Bader

There is one more important way in which the experts could have been thrown off track if they were following the line of the RRP. The newly discovered painting has a singular technical feature that is at the center of discussion. It is painted not on the usual supports of panel or canvas, but on copper. In volume 1 of the Corpus, five paintings on copper are treated. Four of the five were accepted as Rembrandts by earlier scholars. All but one were catalogued as Rembrandts by Horst Gerson in 1968. In 1982, however, the RRP accepted only one as a Rembrandt, an Old woman praying in Salzburg. Two others, the Laughing man in gorget in the Mauritshuis and a Self-portrait in Stockholm, are in the B category, and another two, an unidentified historical scene in Tokyo and A man writing by candlelight in the collection of Alfred and Isabel Bader, are rejected outright. The rejected painting in Tokyo, which is dated 162_ – probably 1628 – comes closest to the Cotswolds painting in size and technique. Three of the five works have a layer of gold ground under the pigment which is lacking in the Tokyo and Bader paintings as well, apparently, as in the Cotswolds copper. (See below.) [The paragraph above was revised on 26 February 2012 to take account of the Man writing by candlelight, which was inadvertently omitted previously.] In volume 4 of the Corpus (2005), the attribution to Rembrandt of the Stockholm and Hague coppers is accepted by the present project leader, Ernst van de Wetering. However, the Tokyo painting is not mentioned, despite the publication in 1989 by “The Bridgestone Painting Research Group” of an outstanding book-length discussion of the painting and its attribution. [20 February 2008: Van de Wetering did respond to the Bridgestone publication. In the exhibition catalogue The mystery of the young Rembrandt, Rembrandthuis 2002, nr. 58, he and Marieke de Winkel attribute the painting to Gerard Dou.] The confusion sown by the RRP on the issue of the paintings on copper may therefore also have affected opinion on the attribution of the new discovery, contributing toward the opinion that it is “probably not” by Rembrandt.
For the first time in decades, a serious sum was paid for a Rembrandt that did not come with the imprimatur of the RRP. Indeed, it was paid for a painting whose attribution to Rembrandt went against the grain of RRP attributions. Acceptance of the new painting should lead to a wholesale reconsideration of all the works to which it is related, distributed across the A, B and C categories of the Corpus. It is to be hoped that this experience will break the stranglehold of the Rembrandt Research Project on attributions of Rembrandt paintings and that a greater diversity of opinion will be listened to and respected.

The initial reactions by Rembrandt specialists to the Cotswolds painting were nearly all marked by caution. The most common response was that the expert in question had to see the painting first with his own eyes before pronouncing an opinion on it. With all respect, this is more a pious declaration than a reflection of how connoisseurs usually make up their minds. Not a month in the life of a Rembrandt connoisseur goes by without their writing a mail to the owner of a purported Rembrandt from whom they have received a terrible amateur photo, a mail saying that the owner’s fond hope of possessing a priceless Rembrandt original is baseless and that there is no point in making the journey the owner proposes to examine the original. The same applies to new discoveries that carry the field from day one. A photograph is good enough, if the attribution is considered an open and shut case and if the alpha authority has expressed a positive opinion.
One reaction quoted in the press was richer than this. Martin Bijl, a restorer with long experience, expressed two reservations standing in the way of a quick judgment in favor of Rembrandt’s authorship of the painting. In the images he saw he was unable to detect any sign of a gold-leaf ground. As we have seen, this is also the case of the Tokyo copper of the same dimensions. To those who reject the attribution of that painting to Rembrandt, its resemblance in this respect to the new painting will weaken the case for giving the latter to Rembrandt. Those who believe the Tokyo painting was painted by Rembrandt, as I do, will see in the technical correspondences a substantiation of the Rembrandt attribution, despite Bijl’s proviso.

   

 

Rembrandt
Self-portrait in gorget
, laughing, detail of monogram
Signed RHL
Ca. 1629-30
Oil on copper, 23.7 x 17 cm.
Sale Cirencester, 26 October 2007, lot 377

[In 2013 bought by the J.Paul Getty Museum from the London dealers Hazlitt, Gooden & Fox.]

Rembrandt
Bust of an old man with turban, detail of monogram
Signed RHL
Oil on panel, 26.5 x 20 cm.
Aetas Aurea Foundation

Bijl also has reservations about the authenticity of the monogram. He told the Dutch daily NRC Handelsblad that it does not seem to have been painted wet-in-wet, in the same narrow time frame as the composition itself. This is a serious point that can only be decided by microscopic examination of a paint sample, something the new owner may not want to permit. However, as far as the form of the monogram goes, it surely comes very close to that of such a contemporary monogram as that on a head of roughly the same size from the same period now in the Aetas Aurea collection.

The value of opinions such as those advanced here is relative. Real proof of Rembrandt’s authorship of an undocumented painting is unattainable. What I have done in this essay is to line up some of the variables involved and demonstrate, as I believe, that they point in the direction of Rembrandt’s authorship of the new painting. The captions, provided with the names of artists and identification of subjects, are to be read as matters of opinion.

Aside from my understanding of the motif, technique and attribution of the new painting, my personal reaction to it is important to me as well. Although I have only seen low-resolution images, I am completely captivated by it, like Jan Six, Otto Naumann and Johnny van Haeften. I see a lively and vivid presentation of Rembrandt at the start of his imagined military phase, a highly intriguing aspect of his self-portraiture. He wears a remarkably free expression on his face, coming across less guardedly, less posed, more vulnerable, than in nearly any other of his self-portraits. The laugh is infectious and inviting. The painting is marked by an amiability that I am touched to find in Rembrandt.
Finally, my judgment of the quality of the painting, another subjective response, is highly positive. The face and body have a marked plasticity that provides the figure with space and light and presence in three dimensions. The outline of the head and shoulders is full of variation without declining into arbitrariness. The touch has the quality that I have called “continuous modulation”: minute, sensitive transitions millimeter for millimeter across the entire surface. The coloring of the face looks somewhat off; I hope that this will respond to gentle cleaning by the restorer to whom it is entrusted by the buyer. I hope too that the buyer will see fit to lend the painting to a public museum as soon as feasible. I am looking forward to meeting this new Rembrandt with a winning laugh on his face.


© Gary Schwartz 2007. Posted on the Schwartzlist on 11 November 2007. and on ArtsFuse on 17 November 2007. Published in abbreviated form in German (“Der Würgegriff des RRP muss gelockert werden”) in the Frankfurter Allgemeine Zeitung, 17 November 2007, p. 46. With kind thanks to the underbidder at Cirencester on 26 October, Otto Naumann.

 

A good photograph of the painting was released by the Rembrandt House in June 2008, when the painting was shown there in a small exhibition at which I was able, to my delight, to meet the laughing Rembrandt. The encounter convinced me that my initial judgment was correct. The exhibition was accompanied by a long article by Ernst van de Wetering in the Kroniek van het Rembrandthuis 2007 (published in 2008) in which no mention is made of the previous publications on the Schwartzlist, The Arts Fuse or FAZ.

Self-portrait of Rembrandt, laughing. Owner unidentified

A somewhat embarrassing coda. The Cotswolds Rembrandt was not at all unknown until 2007. Ernst van de Wetering sums up the situation well in his article in the Kroniek van het Rembrandthuis. pp. 19-20.:
“The notion of ‘discovery’ is perhaps better avoided here. It would be more accurate to say that it ‘finally turned up’. Although the painting unexpectedly was recognized as a Rembrandt at an English provincial auction in October 2007 … and was immediately hailed in the press as a ‘sensational find, the existence of the work had already been recorded – indirectly – in the Rembrandt literature for a long time.

From the addendum to Kurt Bauch’s book on the painting of Rembrandt, 1966. Scanned from the copy in my library that I did not consult when writing Schwartzlist 285.

Several art historians knew it in the form of a reproductive print by the Flemish engraver Lambertus Antonius Claessens (1763-1834). Claessens, incidentally, regarded the painting as a work by Frans Hals. In 1966 the German Rembrandt expert Kurt Bauch (who specialized in the early Rembrandt) listed the painting on which this print was based under ‘works by Rembrandt that have survived only in copies or reproductions’. Bauch realized this as early as 1933, when he wrote in his book Die Kunst des jungen Rembrandt: ‘It is clear from the type and the composition that Claessens’s print is based on a work by Rembrandt. It is essentially impossible to derive any criteria for an accurate dating from the apparently freely executed print.’ We now know that the print is strikingly faithful. As we shall see, however, Bauch’s observation is very understandable.

Claessens’s print had been identified as a copy after a Rembrandt painting earlier still. In 1905, in his Iconographia Batava volume II, Moes described it in a section of his book devoted to portraits and self-portraits by Rembrandt as: ‘By Rembrandt c. 1629, laughing, bareheaded, with gorget (L.A. Claessens sc. As “Le Rieur” by Fr. Hals)’. Hofstede de Groot in 1915 adopted Moes’s attribution.” [The above coda added on 27 February 2012.]


It has been a while since the last Schwartzlist mail. Travels of course, but they never prevented me in the past from filing my columns on time. This time they have been aggravated by more serious circumstances. The past two months have seen the final sickbed and last week the death of my father-in-law, Mathieu Hendriks of Maastricht and Utrecht.

Tomorrow we are off on another visit to London and St. Louis, where I will be speaking at Washington University on 14 November and at the Pulitzer Foundation for the Arts on the 15th. In London, on 12 November, we will be attending a dinner at the Arts Club where my book on Rembrandt is one of the six titles nominated for the Banister Fletcher Award for the most deserving book on art or architecture of the past year.

Reactions to Gary Schwartz.