In his memoir, Eric Hebborn tells a story from his early years working for a painting restorer as he underwent his initiation into the underworld of art. A dealer with a gentlemanly manner approached him and his employer with a canvas to be worked on. The man announced, “I’ve made an exciting discovery, what do you think of this Vandevelde? [sic]” Not knowing what to say, Hebborn simply seconded his employer’s reply that he found it very interesting. The dealer then admitted that the painting would need extensive restoration, and mentioned his hope that after it was cleaned the signature would be visible. And he provided photographs of Van de Velde’s works he thought would be helpful in understanding the present painting. The whole conversation went on with an air of deadpan innocence as if in an on-stage comedy where the audience, looking from behind Hebborn, would see just what he saw—an entirely blank canvas, and an ideal one for the project at hand in that it dated to the seventeenth century and had been carefully scraped down to the base layer of gesso. It was an Old Master waiting for the restorer’s full treatment.43 Obviously, the painting Hebborn made light of (it was proclaimed to be an “original” Van de Velde) was a fake. What, though, if the restoration was to a painting made by Van de Velde himself, but where much of the paint was missing and the blank areas needed to be filled in? What status would this work hold?

When an artwork that has deteriorated or been damaged undergoes restoration, an alteration occurs in what remains of the original work. A question looms as to how much alteration is acceptable. Is there a limit not to be surpassed, a point after which authenticity has been unacceptably compromised? The issue here is a variation on a metaphysical puzzler about change and permanence, which is sometimes exemplified in terms of replacing the parts of an automobile. If it has been repaired, perhaps on several occasions, so that most of the original parts have been replaced, is it the same automobile as before? Art forgers capitalize on the gray area involved in answering this sort of question. A strategy they sometimes employ is to find a damaged piece by a recognized artist that can be bought cheaply and then “restore” it by repainting key portions. In the process, the image may be changed to make it more interesting—an old woman becomes young and attractive, a house appears in a landscape, or a cat in an interior scene. With skilled brushwork, an old reject becomes vitalized and is given a much higher price tag. Other times, the intention in restoration is not to create a deceptively false image but to legitimately renew what has gone missing so viewers see the image as it originally looked. In this case, the artwork has been altered without attempted fraud, yet with extensive restoration its authenticity comes into question.

Part I discussed changing views over time about the amount and type of restoration considered acceptable by professional restorers as well as by art critics. And it was noted that although today standard treatment consists of less intervention than it did historically, that perspective is often disregarded in favor of significant alteration for improved visual effect. How, then, can it be determined whether an artwork has retained its original status after restoration or has been altered in such a way as to be classified either as a fake or as unintentionally rendered inauthentic? While it may seem that there should be a point of differentiation, how can it be established? Experts of different types can be consulted who give answers in various forms.

In the legal realm, statutes are silent on this issue, although on rare occasions case law has spoken to it. In 1981, the US Tax Court decided the case of Monaghan v. Commissioner over the tax write-off claimed for donating a seventeenth-century Spanish portrait to a museum. Two expert witnesses determined the painting to be inauthentic on connoisseurial grounds, but one of them, after examining the restoration the work had undergone, also found it problematic that only 40 percent of the original paint remained on the canvas.44 In 1989, the Tax Court heard the case of Ferrari v. Commissioner in which the value for charitable contribution of twenty-one Pre-Columbian art objects was challenged. Many of the works had extensive restoration. Here, an expert witness stated her opinion that a work with more than 50 percent restoration was of doubtful authenticity, while the court declared,

At some point, excessive restoration takes a piece of this art out of the category of an original. . . . it would be misleading to sell as pre-Columbian an art object which consists of less than 25 percent original material. That dividing line may in fact be too low without full disclosure to the customer.45

On the other hand is the case of De Balkany v. Christie’s (1997), decided by the High Court in England, in which Christie’s was found liable when a client claimed to have been sold an inauthentic painting by Egon Schiele. The entirety of the painting’s surface except for 6 percent was overpainted, including the artist’s initials E and S in the lower right and left corners (his typical signature). With the signature misrepresented by the seller as having been applied by Schiele, the work was declared a forgery. The 94 percent overpainted surface was not a disqualifying factor as long as it adhered faithfully to what appeared on the canvas originally and the original brushwork was in fact done by Schiele. In the words of the court, which accepted the defense’s argument,

if the original picture had been by Schiele then nothing that was done by another hand could be said to make it a forgery . . . even if the painting had been 100% overpainted by someone other than Schiele . . . it would have been proper for Christie’s to attribute the work to Schiele if he had designed the picture and the overpainter had reproduced Schiele’s colours.46

In this situation the judgment was liberal to the point of being unconcerned with the amount of surface area that was restored, while strict about changes to what the surface originally looked like. What was not established was how much deviation from the original look, even if unintentional, this strictness allows.

With legal opinions varying considerably as to how much restoration it takes to nullify authenticity, what does philosophy have to say? Here the discussion moves away from percentages to an even greater disparity of viewpoints. At one pole is the position proclaimed in the nineteenth century by John Ruskin that no restoration is acceptable because the original object will never exist again. Earlier in the century, artist Francisco de Goya announced a similar stance:

Whenever one touches a painting under the pretext of conservation, one always destroys it; and even the authors themselves, coming back to life, would not be able to retouch them perfectly because of the yellowish tone which they acquire with Time, who, as the sages have observed, is also a painter, because it is not easy to retain the fleeting and momentary impulse of the imagination and concert which are found in the initial creation.47

Not even the original artist, we are told, can mitigate the natural aging process. Even if no human cause degrades an artwork, nature does so over time. A created work carries its history, which is bound at its origin and cannot be reembodied.

Philosopher Mark Sagoff presents a more recent and detailed explanation of the importance of an object’s history for its authenticity. Art restoration, he notes, is like medical prosthetics in that it saves the appearance of an original but changes the substance. Referring to the analogy of an automobile, Sagoff notes that replacing its parts is different than renovating an artwork because the artwork was created by a particular artist at a particular time as unique. The features of the artwork are historical and cannot be identical to anything else, even something that has the same appearance. Appreciation of the artwork invests in this special quality, which is why originals are valued much more highly than copies made of them and damaged works are restored rather than simply replaced. The authenticity we prize in an artwork is tied to its unique history in such a way that to alter it through restoration, even while offering a pleasing appearance, corrupts its historical being.48

At the opposite pole from what Sagoff describes is a view that also keys on the history of an artwork but sees the moment of its creation as a mechanism for change rather than a completed event. Martin Heidegger is sometimes cited as a primary adherent.49 “Being” is regarded as always in flux, which applies not only to living beings but also to nonliving objects, including artworks. The history of an artwork is dynamic in that its state at creation undergoes development as its life extends into the future, and it is expected to be different over time. Change is not seen as a loss of originality, although new characteristics will be introduced that were not present at creation. An artwork’s authenticity, then, evolves rather than remaining static. Reaction to viewing an artwork should respect this factor, along with decisions about what physical treatment is warranted. Here is an invitation for liberal restoration, and a possible justification for the claim in the De Balkany case that an unlimited amount of overpainting is acceptable.

Both of these poles are untenable. Adhering to either one points toward the loss of art objects in order to respect the nature of their authenticity. Without intervention, an object in physical decline may collapse into rubble or remain in existence but be unrecognizable. With intervention, we are told by one view, authenticity is destroyed. By the other view, the expectation of change for any artwork suggests that intervention is not a liability but to be welcomed. If a sculpture is badly damaged, it can live on by having missing parts replaced, and its history has not been compromised but has been assisted. But when all of the parts have been replaced, what remains is merely a reproduction that was constructed gradually. The alternatives are having no object at all or having an object void of any original material and workmanship. When put in these practical terms, neither pole has many adherents, with most parties accepting some sort of middle ground. And over this broad terrain, there are various and conflicting answers for determining how much restoration is acceptable.

One approach to narrowing the middle ground is to apply the principle of parsimony and allow for intervention in a minimal way. Rafael De Clercq contends that “restoration is to make as few alterations as possible while aiming to return those properties that the artist intended the work to have, and which at some point after completion it actually had.”50 Still, a judgment is called for not only about the purpose of the artist, which with many works is something that will never be known, but most importantly, based on a chosen metaphysics of change applied to the historical quality of artworks. Sagoff describes two fundamentally different types of restoration. The more conservative type, “purism,” aligns with his own metaphysical preference and includes cleaning along with reattaching remnants from a work’s original creation, such as an arm that had broken off from a sculpture. To go further is considered deceptive to viewers and to destroy authenticity. “Integral” restoration is less concerned about changing a work’s basic nature, and goes further than purism by replacing what was lost from the original with new materials and workmanship, such as creating a new arm for a statue and applying paint to a canvas where paint is missing or weakened.51

In practice, restoration commonly involves both purist and integral principles. An often-cited example by philosophers is the work done on Michelangelo’s Pietà after it was damaged in 1972 by a crazed viewer wielding a hammer. Numerous pieces were found that had broken away from Mary’s arm and face. They were reassembled with adhesive, and areas where pieces were missing were filled with replacement material fashioned from marble dust. The materials and process respected the call for detectable and reversible repairs that has been voiced by professional associations in recent decades, so the added materials were not permanent and were visible under ultraviolet light.52 The sculpture also was cleaned using soap and water, resulting in a lighter patina and removal of shadowing around the face from an accumulation of dirt that was thought to add depth to the eyes. Critics argued over whether cleaning resulted in an improvement of the work by revealing its original state or detracted from an improved state it had reached due to aging and that Michelangelo may or may not have anticipated.53

The restoration process for the Pietà used Michelangelo’s own materials as well as integrated substitutions. Another case that followed suit, although in a different fashion, is the treatment of Rembrandt’s painting Danaë (see Figures 2.2a and 2.2b) in the Hermitage Museum, which was badly damaged in 1985 when a vandal slashed the canvas with a knife and splashed acid on its surface. Restoration took twelve years of painstaking work, with authenticity paramount on the minds of the restorers. Despite pressure from government officials to overpaint and put a prize exhibit back on view quickly, they settled on inpainting blank areas around patches of existing original paint in the fashion of a mosaic. As a museum staff member explained,

We were very careful with original pigments. Any repainting means dissonance. . . . Some parts are 100 percent Rembrandt, some are 50 percent Rembrandt, and some had to be redone. . . . What the visitor sees is not “the original,” and we would never put it forward as such. But the spirit of Rembrandt is intact.54

The goal was achieved of presenting the painting in a way that viewers can recognize the workmanship of the artist but also realize that it is now a weaker version of its former self. When seen from a distance of ten or fifteen feet, evidence of the repair fades, while in a close-up observation problematic areas can be distinguished.

The question of how much restoration is too much—so that authenticity has been compromised—is subject to differing views among museum officials when restoring works of art and to various positions among philosophers, as well as the absence of a definitive answer in the courts. What constitutes in-authenticity as a result of restoration has variable interpretations. Even without intentional deception on the part of the restorer à la Hebborn, Jef Van der Veken, Lothar Malskat, and other figures in the history of art forgery, unintended


Figure 2.2a. Danaë by Rembrandt, 1636, oil on canvas, 185 × 203 cm, after the canvas was damaged in 1985 and prior to restoration. Reuters/Alamy Stock Photo


Figure 2.2b. Danaë by Rembrandt, after restoration. Alamy Stock Photo

deception may occur. But there is not a zero-sum formula or line to be crossed that determines when that happens. Still, without clear-cut guidelines to follow, commercial interests in the evaluation of art objects make financial determinations of their worth that take into account changes made by restoration. Two of the court cases cited previously provide examples.

In the Monaghan case, the value claimed for the donated painting was $80,000, supported by documentation from an appraiser who stated that it was in good condition. Although its status as inauthentic was determined later on connoisseurial grounds rather than overrestoration, the discovery that only 40 percent of the original paint remained on the canvas devalued the work monetarily. The experts who declared the painting was inauthentic estimated its fair market value to be $3,000 at the time it was listed as a charitable deduction. One of the experts said this was the proper value even if the painting were an original, given the extensive restoration that had occurred. Similarly in the Ferrari case, when during the trial the works in question were said to have undergone restoration not recognized at the time they were declared as a charitable deduction, their monetary worth dropped. The most notable change was for a Jaina figurine originally appraised at $18,000, which the professional who had done the appraisal now estimated at $500 after recognizing its considerable restoration. With other objects, he revised his estimates to a lesser extent, including a Mayan cylinder base originally listed at $30,000 lowered to between $15,000 and $20,000. Restoration, then, clearly can reduce the commercial value of artworks, whether or not their authenticity is in question. This downward revaluation suggests a perspective in which authenticity is seen as lost in degrees, with the extent of the loss being an individual matter for each object in question that is determined by expert opinions.

If you find an error or have any questions, please email us at Thank you!