by Brian R. Williams
Click on images to enlarge.
Prominently displayed in the Frick mansion, in a large vestibule just past the indoor garden, is a painting of a young man on horseback. It is one of the first paintings visitors encounter when visiting the museum, and it is hung beside the monumental 1658 self-portrait of Rembrandt and opposite a few other Rembrandt portraits in the Frick's collection. The painting captures the attention of several visitors passing through the vestibule; a crowd pauses and congregates in front of the canvas and studies the face of the anonymous man riding a bony grey horse through a twilight landscape of mysterious shadows. So much is unknown about this painting. Who is this man? What is his mission? Is he a real person, or an allegorical figure? Where is this strange landscape, dominated by an ominous citadel on a mountain?
The plaque on the painting's frame identifies it as The Polish Rider (1) (right), painted by Rembrandt. But in recent years, a few Rembrandt scholars have expressed their doubts about this attribution. It appears that even what little we thought we knew about his enigmatic painting is being called into question. When doubt was first cast on The Polish Rider by members of the Rembrandt Research Project (2)—a group of seven Rembrandt historians who had taken it upon themselves to publish Rembrandt's complete body of work and declare once and for all which paintings were indisputably Rembrandt and which were not—the Frick stood by its attribution of the painting, and many other artists and historians came to the painting's defense. Indeed, the controversy surrounding the authorship of The Polish Rider has shed light on a larger concern among the community of Rembrandt scholars; namely, who is right and who is wrong when it comes to questions of authenticity? Scientific methods—while crucial in advancing our understanding of the great art and artists of the past—have proven to be inconclusive in many cases of contested authorship. So the question of attribution always falls back on the informed but ultimately subjective eye of the connoisseur. And that is where the experts never seem to agree.
The Polish Rider made its way into Henry Clay Frick's impressive art collection in April 1910 (A. Bailey 4). Frick's buyer, Robert Fry, traveled to Tarnowski Castle in Dzików, Poland and purchased it for Frick. Fry was an English painter and art critic, and was initially reluctant to travel all the way to Poland to visit the remote castle. He didn’t think the castle, which he said was "full of second rate French furniture and 1880s objets d'art" would contain anything of value (A. Bailey 4). But then, he later recalled, "a cord was pulled, a curtain was rolled back, and there, before [my] eyes, was revealed one of the world's masterpieces of painting" (as cited in A. Bailey 4). He paid Count Tarnowski £60,000 for it and shipped it to Frick in New York City, where it has been ever since.
The first documented date for the painting is 1791, when Michal Kazimierz Oginski, the grand hetman of Lithuania, gave it to King Stanislaus Augustus II of Poland (A. Bailey 3). It is unknown where Oginski himself got the painting; he may have bought it while he was traveling in Holland, or he may have gotten it from one of his relatives, several of whom spent time in Holland and may have known and associated with Rembrandt (Schama 601). After King Stanislaus died in 1798, the painting—known at the time by its French title, Cosaque á Cheval—was passed around among the Polish nobility until it ended up in Count Tarnowski's possession, where Western scholars discovered it (A. Bailey 4).
Before arriving in America, The Polish Rider had left Poland only twice: first on a visit to Vienna in 1877 and once more in 1898 for an exhibition of Rembrandt paintings in Amsterdam (A. Bailey 4). A year before the Amsterdam exhibition, famed Rembrandt scholar Abraham Bredius had traveled to Tarnowski Castle to study the painting and confirm it was by Rembrandt, calling it "the fine cavalier" (as cited in A. Bailey 4). Aside from one dissenting opinion in 1906—that of Alfred von Wurzbach—Rembrandt scholars unanimously included The Polish Rider in Rembrandt's oeuvre until 1984, when Josua Bruyn, a member of the Rembrandt Research Project, defied conventional opinion and announced, almost as an aside, in a review of a fellow scholar's book, that he did not believe Rembrandt painted The Polish Rider. Bruyn credited Rembrandt’s student Willem Drost with painting the picture (A. Bailey 3).
"[The Rembrandt Research Project] were angry young men needing to push away from the judgments of the previous generation of scholars." -- Gary Schwartz
In the midst of the long-running controversy between the Rembrandt Research Project and the academic community at large, Bruyn's remark about the authorship of The Polish Rider was just a potshot in a seemingly endless battle. Still, the reaction of the academic community was instantaneous. "I was stopped in my tracks," recalls Arthur Wheelock, the curator of northern Baroque painting at the National Gallery in Washington, D.C. (as cited in A. Bailey 3). Julius Held, professor emeritus of art history at Barnard College, Columbia University, said, "Of course, there will always be cases where disagreement is possible. However, The Polish Rider is not one of those cases" (as cited in A. Bailey 94). The commotion the members of the Project caused surrounding their views on—as Bruyn put it—"the so-called 'Polish Rider'" notwithstanding, Rembrandt historian Gary Schwartz sees a reactionary streak in the attitude of the Project members as a whole. "[The Rembrandt Research Project] were angry young men needing to push away from the judgments of the previous generation of scholars," he said (as cited in A. Bailey 69).
The Rembrandt Research Project was founded in 1968, with the goal of establishing "an authentic body of work by conscientious examination and by such up-to-date scientific methods of investigation as dendrochronology (to date wood used in panels), chemical analysis of pigments, and X-ray photography" (A. Bailey 8). The group was funded by the Netherlands Organization for the Advancement of Pure Research (3), although the project members themselves were not paid salaries (A. Bailey 8). It was back in 1956, though, when the idea for the project was first discussed. At the time, founding member Bob Haak was participating in curating a Rembrandt exhibition at the Rijksmuseum in honor of the master's 350th birthday. Haak was alarmed at the startling number of paintings they were receiving that had been attributed to Rembrandt:
The paintings poured in, and we tried to put them in chronological order. I was faced with many paintings said to have been painted by Rembrandt at the same period, and it struck me that one man could not have created so many different sorts of pictures at one time. Many, surely, were not by Rembrandt (as cited in A. Bailey 7).
Aware that something needed to be done, Haak addressed his concerns with other scholars. J. G. van Gelder and Josua Bruyn agreed, and the three men decided to establish the Project. They brought in four other historians: J. A. Emmens, S. H. Levie, P. J. J. van Thiel, and young Ernst van de Wetering, the only remaining original member, who at the beginning was a junior researcher. "Bruyn took charge," said Haak. "He is much cleverer than I am" (as cited in Bailey 7). Their goal was to publish a catalogue raisonné of Rembrandt's complete oeuvre that would, once and for all, settle the matter.
What the Rembrandt Research Project was attempting to do had been tried many times before, with varying results. In the 1830s, English art dealer John Smith concluded that Rembrandt had painted no fewer than 614 paintings, although he had not studied many of these paintings in person (A. Bailey 10). The first Dutchman to attempt to publish Rembrandt's catalogue was Cornelius Hofstrede de Groot; in 1913 he increased the total to a whopping 988. After him, John C. Van Dyke at Rutgers College pared the number down to 48—the smallest ever. In 1935, Bredius brought the number back up to a more reasonable 630. Dutch scholar Horst Gerson was a bit more conservative than Bredius. He attributed 420 paintings to Rembrandt in 1968, which was where the number stood when the Rembrandt Research Project began traveling around the world studying paintings in museums as well as in private collections to determine which ones they would include in their catalogue (A. Bailey 10).
Initially, the group tried to be as scientific as possible in their investigation methods and judgments, noting that the methods of previous scholars were haphazard at best. "Gerson threw out many paintings but didn't say why," said Haak (as cited in A. Bailey 10). The members of the Project believed that unbiased analysis of scientific evidence would be the only way to conclude beyond a doubt whether or not a painting was by Rembrandt. Not only did they study the physical make-up of the canvases themselves—through analysis of linen thread counts, chemical analysis of the paint, microscopic cross-sections, X-rays, and photographs—they also scrutinized each painting’s provenance in great detail. "No Old Master has ever been given such concentrated attention," said Christopher Brown, of the National Gallery in London (as cited in A. Bailey 14).
The Rembrandt Research Project would publish its findings, along with supporting evidence, and conclusions in a multi-volume catalogue titled, A Corpus of Rembrandt Paintings. The team published Volume I of the Corpus, cataloguing Rembrandt's oeuvre from 1625 to 1631, in 1982. Volume II came in 1986, chronicling Rembrandt's work up to 1634; Volume III, in 1990, covered everything from 1634 to 1642. In each volume, the Project assigned the paintings to one of three categories: the paintings they indisputably attributed to Rembrandt were placed in Category 'A,' Category 'B' was for the works of questionable authorship, and Category 'C' paintings were determined by the Project members not to be by Rembrandt. The group intentionally tried to be a definitive as possible, placing the majority of the works in either the A or C Categories: in the first three volumes of the Corpus, there are 144 Category A paintings, 120 Category C paintings, and in the B Category—the questionable paintings—the Project assigned only twelve paintings. In defense of this assertiveness, the project members included this note in Volume I:
This can be seen as an indication that there has been an urge to express firm opinions. In this respect, this book is in the tradition of oeuvre catalogues that present a solid body of accepted works and just as solid a body of rejected paintings, in a situation where in fact there is always room for discussion and reconsideration.
Many scholars believe the Rembrandt Research Project's methodology is exactly backwards. Wheelock, Brown, Walter Liedtke at the Metropolitan Museum of Art, and Peter Sutton from the Museum of Fine Arts in Boston, to name a few, all feel that the A and C Categories in the Corpus are too rigid, and the B Category should be larger. Liedtke explains: "The 'A' pictures should be absolutely indisputable works; the 'C' pictures should be indubitably deficient; and the 'B' paintings should be numerous, with a lease in limbo, where they can be protected by penetrating if inconclusive scholarship from pungent press reports or embarrassing moments in an auction house" (as cited in A. Bailey 65).
This rigid categorization was not the only concern among Rembrandt historians. "The Corpus, for all its merits, has a false aura of objectivity," said David Freedberg in Anthony Bailey's book about the controversy. "It reflects the team's supposition that it can reduce everything to a science. But the Rembrandt Research Project in the end remains subjective, like previous connoisseurs" (66). Data can be interpreted in different ways, depending on who is analyzing it. Furthermore, the Rembrandt Research Project has used scientific evidence inconsistently. They have not used the same level of scrutiny for every painting they have studied, says Maryan Ainsworth, of the Metropolitan Museum of Art. She points out that a painting in a museum, for example, will have accompanying photographs and X-rays the Project would use to determine authorship, whereas a painting in a private collection would be judged on visual analysis alone (A. Bailey 66). And in some cases, the Project has rejected any scientific evidence that does not corroborate their opinion. Ainsworth and Wheelock also believe that the Rembrandt Research Project members have not given enough consideration to comparing Rembrandt's etchings and drawings to his paintings. They believe that by comparing brushstrokes or under drawings viewed via X-rays with pencil and engraving marks in Rembrandt's prints and sketches, one could make as strong a case for or against an attribution than by relying solely on scientific evidence alone (A. Bailey 66-67).
What evidence did the Project collect that led them to doubt the attribution of The Polish Rider to Rembrandt? Bob Haak and Josua Bruyn were the two members who studied the painting, in 1969, along with two other Rembrandts in the Frick collection (A. Bailey 9). The museum permitted them to use the upstairs offices that are closed to the public to examine the paintings, which were removed from their frames. "It was an experience to see the picture off the wall, out of the frame, and away from the slightly sanctifying atmosphere a gallery gives it," said Bruyn. "You could look at it in more of a vacuum. In such circumstances a lesser picture will lose much of its authority, with its weaknesses laid bare. A good picture may gain" (as cited in A. Bailey 9).
The two men also interviewed William Suhr, the Frick's conservator, who likely has spent more time with The Polish Rider than any other living person. Suhr had removed the painting's old varnish in 1950, cleaned it, and restored some of the lost pigment and repaired some damage to the canvas before re-varnishing it (A. Bailey 51-52). Suhr also had the painting X-rayed. The painting was re-varnished again in 1969, the year Haak and Bruyn examined it. The museum mailed the X-rays and close-up photos of the paintings to Haak and Bruyn after they returned to Amsterdam. Who does Suhr think painted The Polish Rider? "A characteristic bit of Rembrandt technique is the area just behind and above the fire at the pool," he observes. "Similar areas can be seen in the  Self-Portrait (left) in the sleeves, the Rijksmuseum Betrayal of St. Peter (4), and others" (as cited in A. Bailey 51). Suhr is arguably more intimately knowledgeable of every square inch of The Polish Rider than Haak and Bruyn, but they do not agree with his attribution. "Pictorially, The Polish Rider seems to be to spring from a temperature different from that of the 1658 Self-Portrait," said Bruyn (as cited in A. Bailey 83). Haak was more direct in his assessment:
All over, the painting is a little too weak. It's not at all balanced. I don't see Rembrandt's brushstrokes. There's the expression of the face, the impossible leg, the very strange horse. And then the whole background—well, it could be unfinished, but some tiny details in it led us to conclude that it was finished or nearly so, and in that case the painting is not by Rembrandt (as cited in A. Bailey 83).
Ernst van de Wetering, the youngest member of the Project, also weighed in:
In Rembrandt, things always have a certain weight, a certain 'body-ness' or stability. Here, with The Polish Rider, everything is too slim. The entire painting is shaky. The proportions of the figure are strange. There is a lack of relationship between the figure and the background—and I am convinced Rembrandt always got his backgrounds right first of all. There is a great deal of detail that isn't incorporated in the whole scene. And the lighting is curious—it breaks up the unity of the painting. All this makes me doubt the picture, makes me wonder if another mentality was behind it (as cited in A. Bailey 84-85).
It should be noted here that the Project members, who relied so heavily on analyzing scientific data before reaching conclusions, have reached an opinion regarding The Polish Rider using subjective visual analysis alone. Furthermore, their greatest argument against attributing the painting to Rembrandt is their belief that it is too technically inferior to have been created by the hands of the master. With The Polish Rider and other paintings, the Project has demanded—their critics complain—an impossibly high standard for any painting to be considered to be, without a doubt, a Rembrandt. A high level of technical consistency is rare, even for a master like Rembrandt; the fact that he was constantly adapting his style throughout his career further complicates matters. Once again, the Project's critics believe they are going about this in the wrong way. "Quality is not an argument for authenticity," argues Sutton (as cited in A. Bailey 74). Wheelock agrees: "Some clumsiness may make it authentic rather than not authentic" (as cited in A. Bailey 74). Nigel Konstam, an English sculptor and Rembrandt aficionado, wrote to Bruyn and Ernst van de Wetering:
I find Rembrandt hugely more variable than your team will allow. I can give very many instances where the quality of the work on a single sheet, all clearly by Rembrandt himself, varies from exquisite to poor, by any standards. In spite of this fluctuating quality, I find my Rembrandt not only is a more prolific artist but is more interesting and has far more to teach us, perhaps because he is more human (and, incidentally, closer to the artist revealed by the documents) (as cited in A. Bailey 75).
The Project's rather dispassionate approach to art history and connoisseurship has also been noted by Julius Held, who believes he would be disqualified from membership in the Rembrandt Research Project because he allows his feelings to be a factor in his evaluation of paintings like The Polish Rider. Other scholars agree. Gary Schwartz explains that the reason he was compelled to write about Rembrandt was a need "to find a way to deal with this emotion that came out of me in front of the work" (as cited in A. Bailey 97). David Freedberg says, simply, of The Polish Rider: "It continues to move me (as cited in A. Bailey 98).
"Quality is not an argument for authenticity." -- Peter Sutton
Despite the scholarly community coming to the defense of The Polish Rider, the Rembrandt Research Project's doubts on its authenticity nevertheless appear to have colored their perception of the painting. Freedberg admits that he has "looked at it more closely because of the RRP" even though he says it continues to move him. "I suppose I'm now inclined against its being by Rembrandt," he confesses (as cited in A. Bailey 88). Egbert Haverkamp-Begemann of New York University's Institute of Fine Arts says, "I see weaknesses now that I didn't see before" (as cited in A. Bailey 88). Schwartz says that, when he takes a closer, more critical look at the painting’s idiosyncrasies, "one…discovers that features of the painting which were always considered marks of its greatness as long as it was a Rembrandt—a certain vagueness, a poetic suggestiveness, and ambiguity of meaning—turn to signs of inferior artistry when they are attributed to Drost" (as cited in A. Bailey 88).
But if Rembrandt scholars at large were finally coming around to accepting the Project's final verdict on The Polish Rider in print, they were in for yet another surprise. In 1993, the Project members wrote a letter to The Burlington Magazine, announcing that everyone except Van de Wetering was retiring. In the letter, the members wrote that Van de Wetering called for a "change of approach" but despite receiving a "sympathetic hearing from the other team members, [they] failed to generate the enthusiasm necessary for a converted change of course" (as cited in A. Bailey 116). So Van de Wetering would forge ahead on his own, assisted by "a small staff, and with the collaboration of specialists from various disciplines" (as cited in A. Bailey 116).
In his new role as the sole Project member, Ernst van de Wetering (left) seemingly took the controversies of the previous decades to heart, and announced some crucial changes to the way the Project would analyze and publish its findings. First, he would instigate a discussion of attribution issues with colleagues before publishing subsequent volumes of the Corpus. He would also make considerable changes to the 'A', 'B', and 'C' classifications, expanding the 'B' Category and steering away from such hard-lined classifications from previous volumes. Furthermore, Volume IV of the Corpus, published in 2005, fifteen years after the previous volume, was not arranged chronologically, but was instead a complete collection of Rembrandt's self-portraits. If scholars believed that Volume IV would at long last feature the official publication and analysis of The Polish Rider, they would have to wait even longer.
In the meantime, Van de Wetering's colleagues were publishing their own opinions of The Polish Rider's authorship. In Gary Schwartz's biography The Rembrandt Book, he does not include a reproduction of the painting and only mentions it in reference to the "public debate" surrounding the painting, but he does conclude that The Polish Rider is "probably by Rembrandt" (369). Simon Schama, in Rembrandt's Eyes, offers a more comprehensive analysis of the painting. He believes the painting features a member of the Oginski family, possibly Marcyan, who was studying in Leiden at the time the picture was painted (601). He also believes the painting is Rembrandt's. Schama also mentions Van de Wetering's revelation at a symposium in Melbourne in October 1997 that he believed it was in fact painted by Rembrandt (n44, 721).
So, after years of speculation and conjecture, does Van de Wetering attribute The Polish Rider to Rembrandt? Yes, he does. But, in his book Rembrandt: The Painter at Work, he adds, "In my view, based on stylistic arguments, [The Polish Rider] may in some details have been finished by later hands" (207). His conclusion is that Rembrandt started the painting, finished some details such as the rider's face, costume and weaponry, but another artist embellished some areas at a later date (209). He compares the backgrounds and rather unfinished nature of The Polish Rider to a similarly executed Rembrandt of the same time period, Moses Breaking the Tablets (right). He also sees in Moses areas that appear to be worked by another artist.
The Polish Rider didn't appear in the Corpus, though, until Volume V. Published at last in 2011, Volume V consisted of small-scale history paintings, further deviating from the chronological organization of the first three volumes. It was also in 2011 that Ernst van de Wetering announced his own retirement from the Rembrandt Research Project effectively shutting down the entire operation (M. Bailey). Van de Wetering simply does not have the energy to catalogue the last eighty or so Rembrandt paintings that have yet to be included in the Corpus (M. Bailey). So, as a compromise, Van de Wetering "now plans to publish a final, summary volume reproducing all the 320 paintings that he believes are by the master's hand. For the 240 already catalogued, there will be brief entries and references to the earlier volumes. For the 80 uncatalogued paintings, there will be slightly longer entries," writes Martin Bailey of The Art Newspaper.
It is a somewhat anti-climactic end to the story. It would be easy at the end to hoist one final criticism at the Rembrandt Research Project for their audacity to think that they were the self-appointed final authorities on issues of authenticity, because, despite their adherence to the scientific process, they fell back into the role of the connoisseur when inconclusive evidence left questions unanswered. It would also be easy to criticize the Project's presumptuous attitude that they would be able to reveal every truth, solve every riddle, and declare once and for all that they have looked upon and judged every existing Rembrandt painting in the world. But their colleagues owe them more than that. The Project attempted a monumental task. One could argue that it was downright impossible, but in their attempt, the members of the project have advanced Rembrandt scholarship to a degree practically unthinkable before they started. They also forced their colleagues and the public at large to take a second, closer look at many paintings that have previously been taken for granted. Before the Rembrandt Research Project, The Polish Rider was considered to be a Rembrandt in the same way that it was considered a masterpiece: simply because it had always been. But that is not honest scholarship. In that respect, the Rembrandt Research Project started a much-needed dialogue. The Polish Rider, for the time being, has withstood the scrutiny, and at the very least, its admirers can say that all the attention, examination and analysis simply confirmed what they had already believed to be true.
 The Rembrandt Research Project members believe that there are between 250 and 300 surviving Rembrandt paintings, and they appear to be trying to pare down Gerson's number to fit this estimation (Bailey 10). But this action has also been criticized by the academic community. "A fundamental flaw in the RRP approach may appear in the team's assumptions about the size of Rembrandt's oeuvre," writes Bailey (67). Professor Ad van der Woude at Wageningen Agricultural University in the Netherlands, Bailey cites, has estimated that there may be as many as 600 existing Rembrandts. He reached this number by assuming a Dutch artist like Rembrandt would have painted approximately 25 pictures in a year over the course of his 45-year career, and then subtracting a substantial number that my have been destroyed or lost (67).
 These curators all had Rembrandts in their collections deattributed by the Rembrandt Research Project.
 The project members were also criticized for not notifying collectors before publishing. In one case, a painting at auction, titled A Bearded Man in an Archway, was placed in the C Category by the project just before the sale, saying it was "from Rembrandt's workshop but was not executed by the artist himself." As a result, the opening bid for the painting went from $10 million to $800,000 (Bailey 79).
 Julius Held cites Horst Gerson's untrained eye in analyzing X-ray imagery. "His occasional references to X-ray evidence, flawed as they are by an incomplete understanding of how to read an X-ray image," writes Held, of Gerson, "were concessions to the growing trust in technology as a substitute for decisions inevitably based on subjective elements" (as cited in Bailey 39).
 For example, Bailey writes: "The Project is in dispute with the Metropolitan Museum in New York about the authenticity of two portraits that, partly on the basis of autoradiographs, the Metropolitan is sure are by Rembrandt, and, partly on the basis of such Morellian details as lace, the Project is sure are not" (40).
 After the deaths of J.A. Emmens in 1971 and J. G. van Gelder in 1980, the remaining members who announced their retirement in 1993 were Haak, Bruyn, Levie, and van Thiel.
Bailey, Anthony. Responses to Rembrandt: Who Painted The Polish Rider? A Controversy Considered. New York: Timken Publishers, Inc., 1994. Print.
Bailey, Martin. "Rembrandt Research Project ended." The Art Newspaper. 24 Feb. 2011. Web.
Van de Wetering, Ernst. Rembrandt: The Painter at Work. Revised Edition. Berkeley & Los Angeles: University of California Press, 2009. Print.
Schama, Simon. Rembrandt's Eyes. New York: Random House, 1999. Print.
Schwartz, Gary. The Rembrandt Book. New York: Abrams, 2006. Print.
1, 2, 3, 4: Retrieved September 4, 2011.