There is bad news for museums and private collectors of works of art by Rembrandt. Up until the mid 90's it was thought that Rembrandt produced many works of art. Through the years the number of paintings thought to have been created by him kept growing in number. In the 1830's when John Smith first cataloging his works he came up with the figure of 200 works of art by Rembrandt. Around the 1930s Wilhelm von Bode produced a catalog noting the number of authentic Rembrandt's to be 377. Later, Wilhelm Valentiner raised that figure up to 606. In 1935 this figure was raised again to 630 by Abraham Breius. (1)

These figures in the 1990s were proven invalid. The Dutch government assembled a group of experts who were dubbed the Rembrandt Research Project. It was their function to determine which Rembrandt's could be directly attributed to him, and which ones were either copies, or fakes. The Dutch government paid them to travel around the world so that they could evaluate every painting they could find thought to be a Rembrandt. The result? According to their determinations the figure was not 630, not 606, but a scant 280!

How did they arrive at these figures? In fact, how does any expert determine if a Rembrandt is indeed a Rembrandt? That is the subject of this story.

The first thing to keep in mind is that there is a difference between a copy and a forgery. A forgery is an attempt by anyone to intentionally deceive. A copy, at least for this story, means a work of art that was an authorized reproduction, or the creation of any facsimile without the intent of deception. They are evasive and difficult to prove. The former is often easy to detect with the technology available to the scientist today.

Often scientists are called on to determine if a work of art is a forgery. Some of the many tools available to the art detective is the scanning electron microscope, ultraviolet and infrared light, the x-ray, and mass and emission spectroscopy. There are many areas where the forger can falter...the paint, canvas, wood, and aging process. (2)

There needs to be only one inconsistency to prove a fake is a fake. For example, as new scientific processes were developed throughout history, advances in organic and inorganic chemistry occurred, thus there was an improvement in the quality of pigments and varnishes over the years. In 1814 a shade of emerald green was created in Schweinfurt, Germany. Not surprising, it was called Schweinfurt Green. If a suspected Rembrandt from 1635 is being evaluated for its authenticity, and Shcweinfurt green is discovered, the work is obviously a fake because the green was invented 179 years after the work of art was supposed to have been painted. All it takes is one single fault, as I said, and it is proof positive the work is bad. (3)

Copies, on the other hand, are another story. They are the reason that there has been so much confusion over Rembrandt paintings over the years. In the seventeenth century workshop it was not only expected, but required that the student copy the master. In order to have been accepted, the art student was forced to paint in their masters style. They needed to emulate the teachers style. In Rembrandts workshop, his students were exceptionally good, or Rembrandt was a great teacher (or both). The authenticity of a painting from that workshop generally caused great confusion. In Urecht in 1651 a decree was issued stating that masters were forbidden "to keep or employ any person as disciples or painting for them if they work in another manor or sign their own name." So many works of the era were painted by very good students which they signed with Rembrandts name. This confusion was pushed even further by Rembrandt himself, or at least that is as it seems. It is believed that when he sold paintings signed in his name to art dealers, there was no provenance saying whether the work was actually done by him or a student. (4)

In cases like this, science does little good. The paint composition was the same, as were the wood panels and the canvases. The only scientific techniques that are useful are the x-ray (used to analyze the artists brushstroke, and varying levels of paint), and autoradiography (used to look at the underdrawings). So where does this leave the mystified art historian on the authenticity question?

It boils down to evaluating the technique that the artist used.

Although evasive to the unknowing eye, Rembrandt's technique was so refined that even his best students couldn't copy it exactly. There are many areas that historians evaluate.

First, the historian must place the work within a particular phase of his career. Rembrandt's earlier work is far removed from that of his later creations. It's as if there were two different artists.

Another area that helps determine if it is authentic or not, is how well the light contrasts to the dark. This technique Rembrandt borrowed from Caravaggio. Take for example The Anatomy of Dr. Tulp (1632). The men all are dressed in black, the back and fore grounds are also very dark. The expressive powers Rembrandt gave them is a key factor in determining authentication. Their heads are strikingly against the dark background, and their facial features seem to be accented by their ruffles. The body, lying there in its pale, morbid yellow skin, sits within a sea of darkness accentuating death. Later in his life his backgrounds often took on a profound richness in color, as in Aristotle Contemplating the Bust of Homer (1653). Whatever secret he used, his apprentices rarely duplicated successfully.

Consider detailed looks at Rembrandt techniques that his students found difficult to copy. His rendition of older people was never idealized. If they were old, he made them look old, unlike other artists of his era that cleaned up subjects wrinkles and other perfections. The realism of the wrinkling he portrayed was difficult, at best, to imitate. If ever you get close to a Rembrandt, look hard at the intricacy.

Consider his shading. Since the way he creates faces it is often the brightest part of his paintings, shadows there are easiest to compare with the shadows of other artists. If you have the opportunity to look, his shadows don't end in a sharp line. Rather they tend to gradually blend into the skin so there is no delineation between beginning and end.

Another comparison can be made in the way he painted his eyes. Again there is a big difference in how the details are presented. Rembrandt's presentation of eyes are very soft. Most important, observe the lighting. The iris' pick up the light in a very soft way. There are no sharp, contrasting lines. As with his shadows one color blends into the other. In Rembrandt's eyes the peaks of the wrinkles surrounding them sometimes shine, and the water around the eyelids sparkle in such a way that we think it is possible to wipe the water away with a handkerchief. (5)

Science can take another step in working with the art historian. At times x-rays are used to determine patterns in the brush strokes. The x-ray is able to reduce the brushstrokes to intricate patterns. (6) It would seem, then, that different artists create different patterns, thus leaving their own signature in another form. The way this is done is that the x-ray picks up the differing pigments in different ways. Lead white will show in one particular way on an x-ray because of its high concentration of lead. Cobalt blue will have different concentrations of cobalt, and so, be picked up differently on the film. The x-ray can "see" the differing layers of pigment and each artist applies paint in their own style. So it is possible to determine who painted what. This is not necessarily successful every time, but it is just another tool available to art detectives. Artists approach different paintings in differing ways, so this is not a foolproof method, but merely an aid.

The problem of determining Rembrandt attribution facing curators today boils down to many factors. It is not possible to determine attribution through pigment analysis because many Rembrandt copies came out of his workshop by apprentices using the exact same material. Rembrandt did not have a 100% unique brushstroke, but one that is often recognizable. The signature is of little use as his students were ordered to sign his name. We must rely on the skill and knowledge of an array of people, curators, scientists, and art scholars. It is frequently the intricate details that answer the question. The Rembrandt Research Project has made leagues of progress in determining which painting is attributed to whom. Thanks to them, original Rembrandts are much more scarce now, much to the disappointment of museums and collectors. The Metropolitan Museum of Art was hit the hardest. They once thought they held 42 Rembrandt originals, and now that figure stands between only 16 and 21, which is bad news for them and everyone who once believed they were genuine.



REFERENCES (Links return to article)

(1) Adams, Henry. "Rembrandt or not Rembrandt (determining paintings by Rembrandt)." Smithsonian Magazine, 12(1995): Page 4 *
(2) Dornberg, John. Artists who fake fine art have met their match -- in the laboratory 2(1992) Smithsonian Magazine: Page 62
(3) McCrone, Walter C. "Artful Dodgers: Virtuosos of Art Forgery Meet the Masters of Scientific Detection." The Sciences Jan. & Feb. 2001: 34
(4) Christopher Brown, Jan Kelch, & Pieter van Theil, Rembrandt: the Master & His Workshop, Yale University Press 1991: Pages 62-3
(5) Brown, 25. Observations from detailed pictures
(6) Adams. Pages 6-10 *

* Note: This Smithsonian article came as a text only version from Infotrac. The pages indicated refer to the Infotrac article, not from the original in the magazine.