Show Less
You do not have access to this content

The Sale of Misattributed Artworks and Antiques at Auction

Anne L. Bandle

The glamour and mystery of the art auction, gathering interested buyers from across the globe, makes it one of the most fascinating marketplaces in existence. ‘Sleepers’, artworks or antiques that have been undervalued and mislabelled due to an expert’s oversight and consequently undersold, appear regularly. This fascinating new book provides the first extensive study of the phenomenon of sleepers through an in-depth analysis of the contractual relationships, liability and remedies that arise in the context of auction sales.
Show Summary Details
This content is available to you

Chapter 1: Notion of “sleeper”

Anne L. Bandle

Upon consignment, an in-house specialist or external expert appraises each lot in order to generate a description for the sale catalogue. In appraising an art object, the expert identifies its attributes, namely its creator or the respective place of origin or discovery, the date or period of creation and provenance.1 The final result of that assessment is expressed in the art object’s attribution. When a sleeper is offered at auction, the expert has failed to correctly determine the valuable attribution of the art object. As a result, the art object is sold for a considerably underestimated price.2

For the purposes of this study, a sleeper is defined as an artwork or antique3 that has been undervalued and mislabelled due to an expert’s oversight and consequently is undersold at auction.4 The auction house’s misattribution is printed in the sale catalogue as well as displayed on its website, communicated to potential clients and to those attending the sale.5 Accordingly, the art object is introduced into the public art market under a wrong label.

In order to qualify as a sleeper for the purpose of this study, three cumulative requirements must be met: first, the sleeper is an artwork or antique;6 secondly, it is sold at auction;7 and thirdly, an auction house is responsible for the incorrect and undervalued attribution.


A misattribution is an erroneous attribution: the error lies in the wrongful identification or appreciation of the object.8 With sleepers, the misattribution must also devalue the art object. Essentially, three types of erroneous and undervalued attributions produce sleepers, and each type is addressed in the following three sections. The first section addresses the attribution of the artwork or antique to a lower-valued creator instead of the actual higher-valued creator. The second section looks at the incorrect dating of the object, which is especially relevant for antiques; and the third section concerns incorrect provenance or ownership history. To be clear, an art object’s physical condition is not an attribute.9 The erroneous identification of one attribute often inevitably leads to a wrong assessment of many of the art object’s other attributes.

1.1  Creatorship Respectively Place of Origin or Discovery

1.1.1  Creatorship

Creatorship links an artist or craftsman to a specific creation. Creators are often distinguishable by their form, which “is the manner or personal style of the creator that determines the characteristics and strategies which comprise his or her visual temperament.”10 Form may change over the creator’s lifetime.

Generally, mistakes may occur during a determination of the consigned object’s artist or craftsman. In fact, sleepers primarily originate because their creator is not accurately identified.11 For example, a sleeper might be attributed to a lower-priced or unknown artist, such as the actual master painter’s studio or a devout follower instead of the master painter himself. Judgment errors regarding the creatorship may occur due to an art object’s poor condition, as sleepers have often been ill-conserved and covered with dust and dirt. When presented to the auction house for consignment, the auction house’s specialist, misled by the neglected appearance of an art object, may have too quickly dismissed it as being created by a low-priced artist or craftsman. In fact, once revealed, many sleepers require substantial cleaning and restoration work.12

More specifically, artworks that raise attribution issues and turn out to be sleepers are often Old Master drawings and paintings. In that event, the sleeper has been attributed to an unknown artist before later being revealed to be by an Old Master painter. Practice shows that misattributed Old Masters works generally date to around the 16th to the 18th century and classify as Baroque Art, such as Sir Anthony Van Dyck (1599–1641), Peter Paul Rubens (1577–1640), William Dobson (1611–1646) and John Riley (1646–1691), or Renaissance Art including Titian (about 1490–1576). Alternatively, sleepers are 17th or 18th century European Art, such as works by Thomas Gainsborough (1727–1788), Rembrandt van Rijn (1606–1669) and William Hogarth (1697–1764). Determining the attribution of Old Masters is particularly challenging, considering that they are often not signed.13 Moreover, authentication remains difficult given the circumstances prevailing at the time of creation, as pupils and assistants may have been working closely with the Master painter.14

Less frequently, sleepers concern sculptures and other kinds of art situated between handicraft and industry, such as carpets, porcelain or crystal. For these types of mediums, the erroneous identification of the craftsman can be as much of a cause for the sleeper as the erroneous assessment of its other attributions, including location of origin, date and provenance. Similar to artworks, the antiques’ misattribution may be induced by their inappropriate conservation and poor condition.

To illustrate, a large redwood relief designed by Sargent Johnson (1888–1967), reputed to be “one of the finest sculptors of the Harlem Renaissance,”15 was originally fixed to the walls of a Berkeley school in California and later dismantled in panels when the school moved.16 Stored in a warehouse, a private dealer found the cracked panels in a plywood bin and bought them for US$164.63 having no knowledge of their real attribution.17 The dealer restored the panels and showed them to different experts, who attributed the work to Johnson. Subsequently, the panels were sold to the Rosenfeld Gallery of New York for US$225,000.18

The concept of creatorship and its significance has evolved over time: “concern with the creator’s hand would seem to be the exception rather than the rule throughout Antiquity, when the name of the artist, alone, stood for certain pictures or statues. Therefore, when copies were made, even at a far later date than the originals, they were given the names of the original creators.”19 Instead of the artist, it was the work of art that mattered and its inherent capacity to trigger an aesthetic reaction from the viewer.20 Shortly after attention shifted from the creation to the creators, concerns of attribution as to creatorship followed.21 Market demand became directed towards the creators who “were no longer linked to an individual feudal lord or monastery, but instead could peddle their skills wherever demand took them.”22 A perceivable differentiation had arisen in recognition of the person of the work’s creator.23 Hence, creators began to sign their works.24

Present society is very much focused on brands: “[b]rands permeate the fabric of our lives – they help construct our identities, our expressions, our desires and our languages.”25 In the art world, creators have become brands. They are more important than anything else, including the quality and aesthetics of the art object itself. Therefore, the creatorship attribute plays a major role in the appreciation of the artwork or antique.26

Overall, creatorship stands for the artist’s or craftsman’s position as a monopolist. The creator is the sole individual capable of producing an original work. However, creatorship does not necessarily mean that no other creator may copy the work and attain the same quality, nor that the creator’s own hand made the original. Finally, creatorship does not necessarily exclude the involvement of others who might have changed or worked on the art object at a later stage. Imitations of style by pupils, workshops, studios, followers and forgers Determining the creator of an art object may be particularly challenging if the master artist or craftsman inspired skilled pupils or followers trying to imitate his style. For example, many Old Master painters established their own studios of apprentices “who were responsible for the execution of all or part of the painting under the authority of the Master.”27 The Master artist may either have carried out the main part of the work or provided his pupils with a general sketch of the creation that they completed and which the Master would finalise.28 Sometimes the pupils were simply instructed with a general proposal to be executed according to the Master’s vision of art.29

In the sleeper context, one illustrative case concerns a painting consigned to Christie’s London in 1994 and catalogued as “Studio of Titian” with a low estimate price of £8,000.30 The auction house assured the consignors that cleaning the painting would incur an unnecessary expense. Later in 2001, the buyer sold it to the Milanese collector Luigi Koelliker. After the painting was cleaned, it was deemed to have been painted by Titian himself and once formed part of King Charles I’s personal collection. When exhibited in 2004 at the “Age of Titian” exhibition at the Royal Scottish Academy in Edinburgh, the label next to the painting read “Titian and workshop – Salome with the Head of St John the Baptist (1570).”31 Ultimately, Sotheby’s New York offered the painting at auction in January 2009 as painted by Titian with a guide price of US$4–6 million, but failed to attract any bids.32

Moreover, those called “followers” meticulously copied or imitated Master artists or craftsmen by working in the Masters’ style. Generally, the followers were neither associated with the Master nor necessarily contemporary with them.33 For example, the French cabinetmaker André-Charles Boulle (1642–1732), considered to be the preeminent artist in the field of marquetry, had many followers.34 At the time of their creation, the pieces that left Boulle’s atelier were subsequently copied by other marquetry workshops, producing copies of fine quality, such as the “bureaux Mazarin” from the 17th century. Nonetheless, the copies failed to equal Boulle’s oeuvre in terms of perfection and precision.35 Another distinguished copier was the renowned maître-ébéniste Etienne Levasseur (1721–1798), a pupil of one of André-Charles Boulle’s sons, who made later copies of Boulle’s creations.36

It may also occur that an original is taken for a fake,37 copy38 or forgery.39 For example, Sotheby’s sold a piece of furniture described as being crafted “in the manner of Levasseur,” only for it to be attributed a few months later in another auction catalogue as manufactured by Etienne Levasseur himself.40 In another case, experts attributed three works to William Turner (1775–1851), which were previously believed either to be fakes or not fully by the hand of the Master.41 Aided by scientific analysis, the same expert who had de-attributed the paintings revalidated their authenticity.42 Such inaccurate qualifications of an art object’s creatorship often also include wrong dating or provenance attributions. Thus, as a corollary, an art object’s accurate origins or provenance may provide useful information to verify its authenticity.43

At present, difficulties in determining creatorship are caused, on the one hand, by the fact that pupils, followers or forgers were often active around the same period when the Masters’ made their creations, and, on the other hand, by the high accuracy and quality of their imitations.44 For example, Marc Chagall (1887–1985) was asked to testify in court to the authenticity of a series of artworks created by a forger.45 To the prosecutor’s greatest surprise, Chagall declared that he had made one of the works subject to the court proceedings, which the impudent forger adamantly contested.46For more recent imitations, a time lapse is often key to the discovery of their forged nature. Scientific analysis enables experts to determine whether the pigments of a work were in existence during a specific artist’s lifetime. In one case, art forger Wolfgang Beltracchi’s creation “Rotes Bild mit Pferden” was attributed to Heinrich Campendonk (1889–1957) and dated 1914 until scientific analysis revealed the painting contained titanium white, a colour that had not been available in 1914.47 For copies made contemporary to the original artworks or antiques, such technical analysis may not be helpful. Instead, experts must assess whether the art object was accomplished in the original creator’s style. Even so, scientific analysis can help by establishing the creation process of an art object, for instance, in revealing a painting’s under-layers.

Due to the efforts of these various imitators, experts are regularly required to determine whether an artwork was completed by the Master painter himself or by one of his pupils, followers or forgers.48 Where works were accomplished with the help of apprentices, the expert must assess which part of the work the alleged creator has in fact created and whether that contribution suffices to attribute the work to the creator, rather than to his studio, which brings us to the next section.

1.1.2  Involvement of the creator and particular mediums Creator’s hand Throughout history, creators have often referred to the assistants, employees and technicians who effectively produce the works under their guidance.49 Manufactured artworks are generally associated with the name of the directing artist or factory – the relevant feature being the place of its creation and not the name of the employee who actually elaborated the work, as in the event of a “Gallé lamp.”50 Where the work’s creation is divided into many employees, the creator’s name becomes a commercial brand. Instead of certifying that the work was made by the hands of the creator, the brand signature guarantees that it was exclusively made in the specific workshop or factory where originals by that creator are produced.51 The brand signature endorses the art object’s authenticity.

Some Old Master painters, such as Rembrandt van Rijn and Jean-Baptiste-Camille Corot (1796–1875), only signed their pupils’ or friends’ works as a sign of recognition so that the works could be sold with greater ease.52 This phenomenon, called “signatures of convenience,” also appeared in contemporary art, when Andy Warhol (1928–1987) signed a pair of “self-portraits” an art class in Michigan had crafted to conform to his style.53 However, the Andy Warhol authentication board, which is the authority on Warhol’s work,54 denied the works’ authenticity.

Creatorship does not necessarily imply that the specific attributed creator has produced the entire work of art. In light of certain collaborations and other kinds of joint productions, it becomes much more relevant that a work has been created under an artist’s or craftsman’s creative responsibility.55 The creator’s involvement, in terms of weight and extent, during the creative process differs depending on the nature of the art object and the object’s development.56 Particularities of certain mediums The particularities of certain mediums used to create art objects may raise problems in attributing such art to a specific creator. Thanks to new avenues for reproduction in graphic and plastic arts, a large part of the development process can be delegated.57 Given these circumstances of production, the standard of creatorship must be redefined: it does not refer to the actual and material creator of the art object, but to the creative and intellectual initiator of the work or to the creator’s factory or workshop.

One example among many is Andy Warhol’s “Silver Factory,” where art was produced on an assembly line.58 Andy Warhol’s technique used photo-emulsion to burn photographic images to silk screens. His standard procedure was “to choose a film transparency, make some suggestions about colours and how the work should be printed, then leave the rest to the printers.”59 By using assistants and industrial processes, Warhol was able to produce works relatively quickly. Regarding these circumstances of production, scholars debate the extent to which Warhol must have been involved in the production process in order for the work to be attributed to him.60

Another example includes the workshop of the cabinetmaker André-Charles Boulle, which employed a large staff, including Boulle’s three sons.61 Boulle’s atelier was situated in the Louvre Museum as a sign of royal favour.62

Again, the name of the creator or the atelier constitutes a brand signifying that the work was created under the artistic and technical guidance, or under the supervision of, the identified artist or craftsman.

1.1.3  Later changes and restoration work

An additional concern when determining creatorship considers the fact that later artists or craftsmen might have worked on and modified foreign artworks and antiques.63 For example, one artist may have painted over a painting by another artist without that artist’s consent, or a later cabinetmaker may have completed the marquetry of a piece of furniture originally made by another craftsman. The latter case occurred with regard to Boulle’s creations. The cabinetmaker Etienne Levasseur not only copied Boulle’s marquetry work, but also “completed” many pieces that Boulle’s atelier had produced.64

Furthermore, restoration efforts may alter art objects and mislead the authenticator as a result. When restoring an art object, the restorer attempts to reinstate the work as it existed in its original condition. Some art objects are restored in a desire to reveal their full aesthetic beauty by remedying “faded colors, visible seams, or patches of canvas showing through.”65 In those cases, the aesthetic understanding differs from the artist’s conception, or, in other words, people may prefer to see a restored aesthetic work over the one the artist originally created, which has evolved with age. One typical example is the removal of varnish that darkens Old Master paintings.66 In another case, the maître-ébéniste Philippe Claude Montigny (1734–1800) restored Boulle marquetry at a very early point in time, before engaging in the creation of Boulle pastiches based on these restored originals.67

Restoration has proven to be essential in the discovery of sleepers. Indeed, sleepers have been literally uncovered thanks to the removal of a second layer of paint a later artist had applied. Technical means of examination, including x-ray, infrared and ultraviolet analysis, can greatly help to detect whether a work of art has been repaired or has sub-layers of paint. Art objects can undergo forensic scientific inquiry, which may “provide valuable information visible only to sophisticated analytical tools such as electron microscopes.”68 To illustrate, the art dealer and “sleeper sleuth” Philip Mould bought the painting “Portrait of Queen Henrietta Maria” for £8,750, including the buyer’s premium, which Christie’s had catalogued as being “after Sir Anthony van Dyck”69 (1599–1641). An x-ray examination of the work suggested that parts of the portrait had been overpainted and enlarged during the 18th century, covering an incomplete studio version of Queen Henrietta Maria as St Catherine.70 The director of the Ashmolean Museum and authority on van Dyck,71 Christopher Brown, subsequently attributed the face of the portrait to van Dyck.72 Creatorship can be an essential attribute of an artwork or antique. Identification can pose considerable challenges to the authenticator in light of the circumstances surrounding the creation of the art object, including the creation process, and due to the skilled imitations, alterations and restorations of the original.

1.1.4  Location of origin or discovery

Generally, objects whose creator is unknown, or irrelevant, are likely to be attributed to their location of origin, place of discovery or to an indigenous group.73 Hence, the location of origin or discovery plays a major role for ethnological and archaeological cultural property.74

A property’s location of origin may imply a qualification of quality, should the place refer to a group of artisans and their technique, such as a “Kirman carpet.” Similarly, the location of discovery of cultural property, such as an excavation site, may be a qualification as to the property’s rarity. Both implications have an impact on the cultural property’s value. Their correct geographic identification is therefore essential.

Misjudgment as to the location of origin or discovery has given rise to sleepers on several occasions. For example, a local auction house in Augsburg, Germany, considerably undersold a carpet it failed to identify as originating from the mid-17th century in Kirman, in southeast Persia. The carpet was offered at auction with an estimate price of €900. In a subsequent sale at Christie’s London, experts rectified the carpet’s attribution and enhanced its provenance, advertising that the carpet had once been owned by the Countess of Béhague (1870–1939), who had amassed an impressive collection of Iranian art.75 The change in attribution dramatically impacted the estimate price, which Christie’s set at £200,000–300,000. The price was justified considering that “[b]y the 17th century, when the carpet was made, Kirman’s designers were at their most inventive and their weaving techniques of a sophistication not seen in other parts of the Persian empire.”76 At auction, two interested buyers bid against each other until the lot reached a final hammer price of £6.2 million, including commissions and taxes.77 While the carpet’s provenance certainly contributed to its greater appreciation at auction, it was ultimately the Kirman origin, indicating a superior weaving quality, that became decisive for the carpet’s increase in value.

For certain types of cultural property, creatorship draws on a community, not a single artist or craftsman. As a consequence, it identifies the property’s geographic and communitarian origins, rather than a specific person’s name.78

1.2  Date or Period of Creation

Generally, artworks can be attributed to an exact date and antiques to a specific period of creation.79 The artwork’s date of creation may be easily determined if recorded by the author on the work itself, whereas if not, specialised experts must be consulted.80 It may be held that the older the work, the more difficult is its precise dating. Consequently, antiques and artworks of a certain age are more likely to be misattributed and undervalued than other art objects.

Should experts fail to determine the exact date of an object consigned for auction, the date is roughly estimated and attributed to a broader time period, such as “early 17th century” or “Baroque era,”81 which is often the case for furniture and unsigned, old paintings whose creator may not be clearly established. Moreover, ethnological and archaeological cultural property is typically very approximately dated by reference to a certain era or style, such as “Roman,” “Corinthian” and “Etruscan” or to a specific century.82

Auction houses have become cautious and tend not to include an attribution specifying the date of an artwork. Instead, they indicate the years of birth and death of the artist, and if he signed his work with a date, such as “signed and dated ‘THMANS:1683’ (‘THM’ linked, lower right)”83 or “signed and dated on the lower right: O. PILNY. 1897.”84 Other artworks and antiques are even less clearly dated in the sale catalogue by referencing the name of a monarch, such as “Louis XV”85 or “George III,”86 or simply by approximations like “circa 1640.”87 References to a dynasty are likely to have an ambivalent character as they directly relate to the period of creation and indirectly imply that the art object was crafted in the country where the monarch had reigned.88

Dating the object is part of its attribution and may have a significant impact on its value. Especially for porcelain, crystal, furniture, jewellery and silverware, determining the correct historical context may be essential for the appreciation of the art object in both expressive and economic terms. For example, should a sale catalogue wrongly describe a framed mirror as originating from the 19th century, when it was in fact manufactured by an Augsburg goldsmith in the 17th century, it is very likely to be sold at a fraction of its market value. An error in the object’s dating is often linked to gaps in its provenance, as information on the primary owners can be of assistance when establishing the art object’s historical context.89

The importance of properly dating consigned property can be illustrated by an impressive discrepancy in estimate prices regarding a certain jug. The regional auction house Lawrence auctioneers, in Somerset, England, first offered the jug at auction in January 2008 as a “19th century French claret jug” with an estimated price range between £100 and £200. The lot was finally sold to an anonymous buyer for £220,000. In October 2008, following the sale’s cancellation, the original owner consigned the jug at Christie’s, which described it as “a carved rock crystal ewer made for the court of the Fatimid rulers of Cairo in the late 10th or early 11th century.”90 The jug sold for its new estimate price of £3 million.91 Clearly, neither the owner nor the regional auction house spotted that the crystal ewer originated from the early high Middle Ages and was hence of considerable value.92

As another example, a family in the State of New York purchased a finely potted bowl at a yard sale for US$3.93 The family brought the bowl to Sotheby’s for an appraisal, which identified it as “[a] rare and important ‘Ding’ bowl from the Northern Song Dynasty” (960–1279) with an estimated price of US$200,000–300,000.94 According to the sale catalogue, only one other piece was known, an almost identical piece in terms of form, size and decoration, held by the British Museum in London.95 The bowl was sold to a London art dealer for US$2.225 million, including the buyer’s premium.96 The connection to the Song Dynasty was decisive in unveiling the sleeper. Despite establishing its correct attribution, the auction house had considerably underestimated the market interest for the bowl.

Determining the age of an artwork or antique is an essential part of its identification. It supplies the work’s historical context and allows scholarship or, more generally, the viewer to distinguish the piece from more modern copies.

1.3  Provenance

An art object’s provenance, i.e. “the chain of ownership from the original artist to the present owner,”97 is one tool used by experts to attribute an art object.98 Information on the provenance of artworks and antiques may be found in books and catalogues, collector’s lists, pictures and other archives. Again, the art object’s age is of relevance, as “the older the work, the less continuity is going to be found in the provenance, except for the most famous works owned by the most prominent collectors.”99 Provenance has an impact on the art object’s identity as it contains the art object’s journey since creation. Hence, provenance places the art object within its historical context, but it does so with regard to its entire chronology of existence, unlike a date or period, which is limited to the art object’s moment of creation.

Although provenance generally does not include any physical alterations to the art object other than changes due to its age or use, it shapes the art object in its historicity, expression and exclusivity. For example, artworks and antiques once part of a royal collection have a provenance that adds distinction. Two identical objects will be fundamentally different because of their distinct provenance. Hence, sale catalogues include information such as “Ségal Collection, Basel”100 or “from the Fatimid Royal Treasury”101 relevant to the art object.

In some cases, experts have ignored the valuable provenance of a sleeper, which is then sold along with an incorrect attribution. The omission of, or the failure to indicate, a previous owner’s identity in a lot’s description also often implies that it has not been correctly dated or authenticated. Misattributions related to a wrong provenance are thus often accompanied by errors as to the creator of the art object and/or its dating.

The provenance of the previously mentioned Kirman carpet was attributed by the German auction house to “Persian Gallery, Antique, … Collectible.”102 When consigned with Christie’s a few months later, the Islamic Department examined the carpet and established that it had once formed part of the renowned collection of Martine Marie Pol, Countess of Béhague, based upon the carpet’s design and its inclusion in an established Persian art advocate’s survey.103 Christie’s sale catalogue entailed a two-page analysis and the carpet’s attribution as “Kirman Vase Carpet, South East Persia, Mid-17th Century.”104

The false provenance of a painting bought at auction was central to the plaintiff’s allegations in Foxley v Sotheby’s.105 Foxley, the purchaser of a painting from Sotheby’s, claimed that the auction house had knowingly replaced the name of an equivocal, but disreputable, dealer in its catalogue with “Private Collection, Paris” to avoid “placing a cloud over the painting.”106 The plaintiff argued that he would not have purchased the painting had he known of the dealer’s presence in the provenance.107

Auction catalogues generally only include prestigious provenances, not a full list of previous owners. Also, cryptic provenances, like “Private Collection, Paris,” in catalogue descriptions are quite frequent. They guarantee the privacy of the consignor’s identity. Sometimes, they may also conceal the fact that the auction house is missing conclusive information as to the object’s precise identity and history.108 Notwithstanding the risks of misuse, provenance is an essential part of an art object’s attribution. It evidences all past owners and locations of an art object since creation.

1.4  Correlation Between Attributes

Generally, pieces of information as to creatorship, origin or discovery, date or period, and provenance come together to form a complete attribution. Attributes are usually connected and complement each other. For instance, a painting’s date of creation correlates with the lifetime of the artist and forms the starting point of the provenance chain. As a consequence, an erroneous attribution generally implies that all other attributes have been wrongly determined. As an example, in Thomson v Christie Manson & Woods Ltd and Others,109 Taylor Thomson sued Christie’s for misrepresenting a pair of porphyry and gilt-bronze urns she purchased at auction in 1994 for almost £2 million – dramatically higher than the pre-sale estimate of £400,000–600,000.110 The sale catalogue described the urns as “A pair of Louis XV porphyry and gilt-bronze two handled vases designed by the architect and ornamentalist Ennemond-Alexandre Petitot (1727–1801) circa 1760” and owned by the seventh Marquess of Cholmondeley before 1921.111 Following the sale, rumours surfaced that the urns might be mid-19th century imitations, which would reduce their value to only £25,000.112 Christie’s subsequently initiated metallurgical investigations, which confirmed the rumours. As a result, it was established that all attributes of the vases had been misidentified: their date, designer-creator and provenance.

Overall, it may be argued that each attribute has an equal stake in the identification of an artwork or antique. None should enjoy priority, as they constitute together the attribution of the object. If one attribute cannot be established due to a lack of conclusive information, the other attributes play an even greater role in the identification of the art object. For example, an anonymous painting from the Baroque period may still be attributed to a certain period or date and possess a detailed provenance.113 Additionally, if one attribute is clearly determinable, but contradicts another established attribute, the first attribute may not outweigh the second. Instead, the expert must attempt to overcome any contradiction.

To conclude, each art object is shaped and identified by its three main attributions: creatorship or location of origin or discovery, date or period of creation, and provenance. In the event of a sleeper, at least one – if not all – of these attributes has been misidentified. The next section assesses how sleepers become identified in the market.


In practice, the less valuable an art object appears to be, the less extended the auction house’s effort is to assess its attribution.114 The auction house’s initial appreciation of the art object is therefore essential for its appropriate attribution, especially for the discovery of a sleeper. Potentially, each time a sleeper is traded on the market it can be discovered, since it may be exposed to public scrutiny and require an expert’s authentication. Thus, one must ask at what specific moment do artworks or antiques become qualified as sleepers? Section 2.1 examines the defining action leading to the identification of a sleeper. Section 2.2 compares the sleeper phenomenon to another authentication issue: fakes and forgeries.

2.1  Defining Action

When imagining a sleeper, one might think of a treasure that remains hidden until its discovery. This assumption may certainly be accurate in most cases. Sleepers generally appear on the market following a long period of time during which they have not been traded and were stored away by private individuals. Commonly, these sleepers have not been included in any sale register or catalogue as a result, which would document their real attribution. The action of the sleeper’s discovery, however, is not physical but intellectual. Physically, the sleeper is still the same object – it does not necessarily require restoration or cleaning work. Instead, an art object becomes a sleeper due to the benefit of scholarly hindsight. It will only be qualified as such upon the revelation of its veritable attribution. Until that specific moment, the art object is merely an artwork or antique with a lower-valued attribution believed to be accurate nonetheless. For example, a painting is held, traded or loaned as the work of the artist to which it is attributed. A sleeper can only retroactively be declared as such from the moment the misattribution is revealed. As a result, the designation of an art object as a sleeper applies to the period the art object was misattributed until the time of its discovery.

A sleeper comes into being through the actions of people around it: it results from a consensus of scholarship to re-attribute the work of art. After careful research, one or more experts may assert that the artwork under examination has been assigned the wrong creator, respective origin, date or provenance and, consequently, may rectify one or several of its attributes. If the art market accepts the new attribution, which is most likely if made by the relevant leading expert (“authority”115), the art object will be traded under that new attribution. Identifying a sleeper is more complicated than the simple detection of a mistake in attribution. It implicates, on the one hand, the proper assessment of a work’s correct creatorship, origin, provenance and date, and, on the other hand, the scholars and stakeholders of the art market’s approval of the identified attribution. The latter requirement is of vital importance to the new attribution’s establishment in the art market.

In essence, the identification of a sleeper is only a change of its appellation and does not require any physical change of the art object itself. In other words, the painting will still depict the same portrait, have the identical brushwork and be made out of the exact same material. The defining action impacts the non-tangible aspects of the artwork, such as its appreciation or its quality perceived by scholarship or more generally by the viewer. This is due to the nature of attribution itself, which acts as a “label” parameter that does not necessarily physically transform the sleeper. In summary, the defining action transforming an art object into a sleeper is the establishment of its correct and much more valuable attribution as well as the market’s acceptance of that new attribution. To this point, it is not the art object that is “sleeping,” but rather the scholars and the art market that have failed to correctly identify the sleeper.

This revelation, meaning the realisation that the art object has been held and traded as a sleeper, enables all interested parties to ascertain their resulting loss. From that moment on, the consignor may turn against the auction house or the buyer.

2.2  Comparison to Fakes and Forgeries

To a certain extent, sleepers are the negative reflection of counterfeits: both are held as something that they are not. The following section analyses the correlation and dissimilarities between sleepers and counterfeits.

2.2.1  Correlation and similarities

The misidentification of an art object’s attribution is much more common with regard to fakes and forgeries. In their essence, counterfeits can be equated to sleepers since they all have been subject to a false attribution of creatorship, location of origin or discovery. Both sleepers and counterfeits carry an incorrect label. An art object will only be designated as a fake or a forgery once the deception is discovered. In other words, as long as the deception remains unidentified, the art object is believed to be an authentic work from the given artist or craftsman, or from a specific place of origin or discovery. Something authentic is something properly described. Accordingly, the authenticity’s characterisation is relevant not only to define a counterfeit, but also a sleeper.

A further correlation between sleepers and counterfeits entails the impact of the re-attribution in terms of value and subjective appreciation.116 While sleepers dramatically rise in esteem and value once correctly attributed, unmasked forgeries become despised objects after losing the art market’s consideration.117 Exceptionally, some counterfeit works may, despite the lack of authenticity, still fascinate because of the forgers’ ability to deceive and the quality of their technique.

Unlike forgeries, fakes cannot be compared to an original exact replica. Given that fakes are works “in the style of the artist,”118 they will only be detected when compared to the complete artist’s or craftsman’s oeuvre.119 No existing copy exists to which experts can refer in establishing the sleeper’s accurate attribution. Like fakes, sleepers generally also undergo an examination that attempts to compare the sleeper’s style, technique, form, subjects, colours, and so on with the oeuvre of a particular creator who could have potentially made the sleeper.

With counterfeits and sleepers, an attribution is more likely to be challenged the more frequently the art object is traded on the market, as each transaction will subject the art object to critical assessment by experts and the public eye.

2.2.2  Distinction

Unlike sleepers, fakes and forgeries are negative concepts. They refer “not to any specific characteristics that a work possesses, but to those characteristics that the work fails to possess.”120 A counterfeit lacks pedigree in terms of originality, provenance and age. The creator simply tried to reproduce a work or style of the artist or craftsman and fake its historical origins, without adding a personal creative component. A sleeper, on the other hand, is usually not an original considered to be a counterfeit. Instead, a sleeper has an inaccurate attribution that previously described a less valuable creator.

A counterfeit pretends to be something it is not, but a sleeper makes no such pretension. The sleeper exists entirely as a result of the scholar’s and the market’s failure to detect its real attribution. For forgeries and fakes, the misattribution derives from what the work depicts, caused by its physical resemblance to other works from a valuable creator. Forgers “penetrate into the domain of aesthetics”121 by attempting to create a work that carries the same aura as an original. A sleeper, on the other hand, is not misleading because it resembles the oeuvre of another creator. Instead, the mistake results from insufficient research on the art object and the failure to recognise that it has the quality, age and provenance of a work with a similar attribution.

Furthermore, sleepers do not involve intentional deception, whereas forgeries have been purposely created with fraudulent intent or are subject to a fraudulent use.122 The followers, pupils and other artists who created works and aspired to achieve the quality and technique of Master artists differ from forgers in that they lacked intent to deceive during creation. The intent to defraud generally exists where there is an intent to make a profit from the fraud. Consequently, counterfeits are purposely sold as genuine works, leading to greatly overvalued prices. Sleepers do not involve any such intent; their correct attribution remains unknown, which results in their sale at a highly undervalued amount. Only the final purchasers – should they recognise the sleeper – might have an interest in keeping the secret until the deal is completed.

Studying comparisons between sleepers and counterfeits helps us to understand why the law and auction houses treat them differently and demand different approaches. Chapter 2 examines the process of attribution of artworks and antiques, assesses why attributions are important, and explains how they can be manipulated by auction houses.

1 For example, Christie’s Conditions of Sale stipulate the following: “Our description of any lot in the catalogue, any condition report and any other statement made by us (whether orally or in writing) about any lot, including about its nature or condition, artist, period, materials, approximate dimensions or provenance are our opinion and not to be relied upon as a statement of fact” (para. A.1 (b)); cf. Jáuregui, “Rembrandt Portraits: Economic Negligence in Art Attribution,” 1950; Spencer, “Introduction,” xi.
2 See Davis and Ludlam, “Sleepers: Whose Side is the Law On?” 35 (defining the issue by “the seller of a sleeper has failed to appreciate the true value of the item he or she is putting up for auction”).
3 Antiques that have been sold as sleepers include furniture, carpets and porcelain.
4 Generally, each lot’s description is accompanied by high and low estimate values that serve as an indication of the hammer price that the auctioneer expects the lot to achieve. The estimated values, in particular, are not a guaranteed price range for the lot or its market value. See, for instance, Bonhams, Conditions of Sale for Sellers UK, para.1.3.1; Harvey and Meisel, Auctions Law and Practice, 11.
5 See Meisel, “Auctioneers and Misdescriptions,” 1036.
6 For the purposes of this study, artworks are tangible and movable objects created or fashioned by humans and embody some creative expression. Antiques may be defined as old commodities, ornaments or crafted objects, which are collected for their beauty, rarity or curiosity and therefore traded on the market; see Katz, “Sachmängel beim Kauf von Kunstgegenständen und Antiquitäten,” 17. In this study, artworks and antiques are jointly called “art objects.” Other categories of cultural heritage, such as human remains, “evade ownership in a traditional sense” and are therefore difficult to subject to the attribution debate; Flessas, “The Repatriation Debate and the Discourse of the Commons,” 2.
7 In this study, the term “auction” is used as a reference to the voluntary and public sale of artworks and antiques to the highest bidder by public competition.
8 Misattributions must be distinguished from mistakes in writing, such as typographical errors, that involve the simple omission, transposition or duplication of a minor number of characters in the printed description of an object held in catalogues. Unlike misattributions, such errors do not result from a judgmental analysis of the object and, therefore, do not reflect an erroneous appreciation of the object, but are purely of a formal nature.
9 At auction, the physical condition of the lot is described in a so-called condition report, which is simply a detailed statement of the physical condition of an artwork at the time it is brought into the auction house. In particular, it describes eventual damages to the work and what restoration work has been or should be done. See Bowen, “Condition Reports for Artworks.”
10 O’Connor, “Authenticating the Attribution of Art,” 10. For improved readability, individuals are sometimes referred to in this report using solely the masculine form.
11 Many examples from practice show that an attribution to an incorrect artist is possibly the principal error leading to undersold consigned property.
12 See Mould, Sleuth – The Amazing Quest for Lost Art Treasures, 103; Melikian, “The Art of Collecting: The European Fine Art Fair,” S1.
13 See Annesley, “Attributing Old Master Drawings,” 82.
14 See below, Section
15 Pogash, “Berkeley’s Artwork Loss Is a Museum’s Gain.”
16 See Pogash, “Berkeley’s Artwork Loss Is a Museum’s Gain.”
17 See Pogash, “Berkeley’s Artwork Loss Is a Museum’s Gain.”
18 See Pogash, “Berkeley’s Artwork Loss Is a Museum’s Gain.”
19 Jáuregui, “Rembrandt Portraits,” 1953 (citing Mansfield, “Connoisseurship and the Methodology of the Rembrandt Research Project,” 175).
20 See Jáuregui, “Rembrandt Portraits,” 1953.
21 See Jáuregui, “Rembrandt Portraits,” 1953.
22 Jáuregui, “Rembrandt Portraits,” 1954.
23 See Jáuregui, “Rembrandt Portraits,” 1954.
24 See Jáuregui, “Rembrandt Portraits,” 1954.
25 Katyal, “Between Semiotic Democracy and Disobedience,” 53.
26 See Chapter 2, Section 2.3.
27 Reeves, “The Rights and Risks of Experts in French and American Courts,” 19.
28 See Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 107 (n. 113).
29 See Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 107 (n. 113).
30David Seton Pollok-Morris Dickson and Susan Marjorie Glencorse Priestley v Christie Manson & Woods Ltd (unreported, out-of-court settlement); see for instance, Brown, “Couple Settle after ‘Their’ Titian was Sold for a Song,” Daily Mail Reporter, “Christie’s Auction House Sold £4million Titian Painting for Just £8,000”; Moore, “Christie’s Sued for Misidentifying Titian Painting Worth Millions.”
31 According to the museum curator, the work includes “a substantial amount of Titian’s own work, and this seemed to be corroborated by evidence of quite significant changes to the composition revealed by X-radiography,” Brown, “Couple Settle after ‘Their’ Titian was Sold for a Song.”
32 Sotheby’s Sale, “Important Old Master Paintings, Including European Works of Art,” January 29–30, 2009, New York, Sale n. 08516, Lot n. 33.
33 See Art Libraries Society of North America, “Attribution Qualifiers for Artists’ Names.”
34 On the particularities of the Boulle marquetry, see Miessen, “Le meuble ‘Boulle’ à travers les âges.”
35 See Miessen, “Le meuble ‘Boulle’ à travers les âges.”
36 See Miessen, “Le meuble ‘Boulle’ à travers les âges.”
37 A fake is a work “in the style of the artist” (Karlen, “Fakes, Forgeries, and Expert Opinions,” 21, fn. 2) that is “made or altered, with intent to deceive, in a manner that it appears to have an authorship, origin, date, age, period, culture or source which it does not in fact possess” (Olsburgh, Authenticity in the Art Market, 1–2).
38 A copy is a faithful reproduction of an original created by another artist or craftsman, without any fraudulent intent. Honest copies may, however, be subject of a fraudulent use. See Karlen, “Fakes, Forgeries, and Expert Opinions,” 21, fn. 2; Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 118 (n. 128) and 125 (n. 131–132); Dutton, “Authenticity in Art,” 259.
39 Forgeries are exact “replicas of genuine pieces that are either deliberately created to deceive or else innocently created but later passed off as original works of the famous artist”; Karlen, “Fakes, Forgeries, and Expert Opinions,” 21, fn. 3.
40 Noce, Descente aux enchères, 143.
41 See Kennedy, “Three ‘Fake’ J.M.W. Turner Paintings Authenticated.”
42 See Kennedy, “Three ‘Fake’ J.M.W. Turner Paintings Authenticated.”
43 See Simmons, “Provenance and Auction Houses,” 85 (stating that “if it was in a 19th century collection then it is not a modern copy”); see Section 1.3.
44 See Lequette-de Kervenoaël, L’authenticité des œuvres d’art, 120 (n. 128–31).
45 See Grampp, Pricing the Priceless, 126.
46 See Grampp, Pricing the Priceless, 126. Nicolas Poussin (1594–1665) was also copied early on by artists such as Angelo Caroselli (1585–1652) whose accuracy deceived even the authors of the reproduced works, see Lequette-de Kervenoaël, L’authenticité des œuvres d’art, 121 (n. 128–31).
47Trasteco Ltd v Kunsthaus Lempertz KG, LG Köln, 28 September 2012, 2 O 457/08. On the Beltracchi fraud, see Koldehoff and Timm, Falsche Bilder Echtes Geld.
48 See Reeves, “The Rights and Risks of Experts in French and American Courts,” 19.
49 See Guggenheim, “Exit the Deciders: What Will the Recent Disbanding of Authentication Committees Mean for the Art World?”
50 See Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 162 (n. 172).
51 See Graw, Der große Preis − Kunst zwischen Markt und Celebrity Kultur, 29.
52 Such signatures are named to be “of convenience.” Jean-Baptiste Corot (1796–1875) and Amadeo Modigliani (1884–1920) have applied signatures of convenience to artworks by their artist-friends. Salvador Dalí (1904–1989) signed blank sheets of paper, which could be completed by other artists and sold as originals. Rembrandt (1606–1669) signed and retouched his pupils’ works that he judged to have particularly succeeded. See Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 187–188 (n. 232); Goepfert, Haftungsprobleme im Kunst- und Auktionshandel, 27.
53 See Polsky, I Sold Andy Warhol. (too soon), 134.
54 On the phenomenon of expert “authorities” in the art market, see Chapter 2, Section 2.2.1.
55 See Von Brühl, Marktmacht von Kunstexperten als Rechtsproblem, 11.
56 See Von Brühl, Marktmacht von Kunstexperten als Rechtsproblem, 11.
57 See Von Brühl, Marktmacht von Kunstexperten als Rechtsproblem, 11.
58 See Design Party Website, “La Factory d’Andy Warhol: Entre décadence et génie.”
59 Kaufman, “Challenge to the Andy Warhol Authentication Board,” 4.
60 See Kaufman, “Challenge to the Andy Warhol Authentication Board,” 4.
61 “L’histoire d’André Charles Boulle,” Jean Bourdette’s Workshop.
62 “L’histoire d’André Charles Boulle,” Jean Bourdette’s Workshop.
63 See Karlen, “Fakes, Forgeries, and Expert Opinions,” 21, fn. 2. It may then be necessary to proceed to extensive restoration works in order to uncover the original, i.e. first layer of paint. See Mould, Sleuth – The Amazing Quest for Lost Art Treasures, 103.
64 See Miessen, “Le meuble ‘Boulle’ à travers les âges.”
65 Grampp, Pricing the Priceless, 134.
66 Grampp, Pricing the Priceless, 134.
67 See Miessen, “Le meuble ‘Boulle’ à travers les âges.”
68 Flam, “Defending the Integrity of an Artist’s Life’s Work.”
69 Christie’s Sale, “Interiors.” January 24, 2012, London South Kensington, Sale n. 4102, Lot n. 261; “Van Dyck: What Lies Beneath,” Fake or Fortune? For an illustration of the sleeper, see Appendix 3.
70 “Van Dyck: What Lies Beneath,” Fake or Fortune?
71 On “authorities” see Chapter 2, Section 2.2.1.
72 “Van Dyck: What Lies Beneath,” Fake or Fortune?
73 See Duret-Robert, “L’authenticité des œuvres d’art dans la pratique du marché de l’art,” 29.
74 See Skripsky, Die Online-Kunstauktion, 39.
75 Christie’s Sale, “Oriental Rugs and Carpets,” April 15, 2010, London King Street, Sale n. 7845, Lot n. 100.
76 “Rug Rave,” The Economist.
77 See Truscheit, “Der fliegende Teppich.”
78 For the sake of simplicity, the following chapters will not specifically mention creatorship and location of origin or discovery, but simply refer to creatorship. However, this does not imply that the analysis in this study is not concerned with certain cultural property that has no creator.
79 See Skripsky, Die Online-Kunstauktion, 38.
80 See Skripsky, Die Online-Kunstauktion, 38.
81 Koller Auktionen Sale, “Möbel, Pendulen, Bronzen, Spiegel, Tapisserien und Dekorative Kunst,” March 29, 2012, Zurich, Lot n. 1052; see also Skripsky, Die Online-Kunstauktion, 38.
82 Koller Auktionen Sale, “Möbel, Pendulen, Bronzen, Spiegel, Tapisserien und Dekorative Kunst,” March 29, 2012, Zurich, Lot n. 1005.
83 Christie’s Sale, “Old Masters & British Paintings,” May 4, 2012, London South Kensington, Lot n. 111.
84 Koller Auktionen Sale, “Gemälde des 19. Jahrhunderts,” March 30, 2012, Zurich, Lot n. 3256 (transl. author).
85 Koller Auktionen Sale, “Möbel, Pendulen, Bronzen, Spiegel, Tapisserien und Dekorative Kunst,” March 29, 2012, Zurich, Lot n. 1053.
86 Christie’s Sale, “Important English, Continental, and American Silver,” May 17, 2012, New York, Lot n. 104.
87 Christie’s Sale, “Important English, Continental, and American Silver,” May 17, 2012, New York, Lot n. 69.
88 Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 161 (n. 170).
89 See Simmons, “Provenance and Auction Houses,” 85.
90 Christie’s Sale, “Art of the Islamic and Indian Worlds.” London, October 7, 2008.
91 See Bailey and Harris, “Islamic Sleeper Priced at £100 in January to be Auctioned for £3 million in October.”
92 See Davis and Ludlam, “A Sleeper in the Attic,” New Law Journal, November 28, 2008, 1661–1662.
93 See Reuters, “Chinese Bowl Found at Garage Sale Fetches Over $2.2 Million at Auction.”
94 Sotheby’s Sale, “Fine Chinese Ceramics & Works of Art,” March 19–20, 2013, New York, Sale n. 08974, Lot n. 94.
95 Sotheby’s Sale, “Fine Chinese Ceramics & Works of Art,” March 19–20, 2013, New York, Sale n. 08974, Lot n. 94.
96 Sotheby’s Sale, “Fine Chinese Ceramics & Works of Art,” March 19–20, 2013, New York, Sale n. 08974, Lot n. 94; Reuters, “Chinese Bowl Found at Garage Sale Fetches Over $2.2 Million at Auction.”
97 Butt, “Authenticity Disputes in the Art World,” 73 (referring to Greenberg Gallery, Inc. v Bauman and Entwistle, 817 F.Supp. 167, 173 (D.C. 1993), aff’d, 1994 U.S. App. LEXIS 27175 [D.C. Cir. 1994].
98 See O’Connor, “Authenticating the Attribution of Art,” 17.
99 O’Connor, “Authenticating the Attribution of Art,” 17.
100 Koller Auktionen Sale, “Möbel, Pendulen, Tapisserien, Bronzen und Dekorative Kunst,” December 6, 2012, Lot n. 1001.
101 Christie’s Sale, “Art of the Islamic and Indian Worlds,” October 7, 2008, London King Street, Sale n. 7615.
102 Auktionshaus Georg Rehm Sale, “229. Kunstauktion,” October 9, 2009, Lot n. 2035. See also Section 1.1.4.
103 See Pope, A Survey of Persian Art from Prehistoric Times to the Present.
104 Christie’s Sale, “Oriental Rugs and Carpets,” April 15, 2010, London King Street, Sale n. 7845, Lot n. 100. The original consignor took the regional auction house to court but was unsuccessful in obtaining damages based on the misattribution; LG Augsburg, 22 O 3163/10, January 27, 2012.
105Foxley v Sotheby’s Inc., 893 F.Supp. 1224 (1995 U.S. Dist. Lexis 5332).
106Foxley v Sotheby’s Inc., 893 F.Supp. 1224, at 1232.
107Foxley v Sotheby’s Inc., 893 F.Supp. 1224, at 1232.
108 See Chapter 2, Section 3.3.1.
109Thomson v Christie Manson & Woods Ltd and Others [2005] EWCA Civ. 555 (12 May 2005). See also Hudson, “A Tale of Two Urns,” 307–316; Olsburgh, Authenticity in the Art Market, 37 et seq.
110 [2005] EWCA Civ. 555., para. 1.
111 [2005] EWCA Civ. 555, para. 1. The catalogue entry is reproduced in Vyas, “Is There an Expert in the House?” 426 et seqq.
112 [2005] EWCA Civ. 555, para. 2.
113 Cf. Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 160 (n. 170).
114 See Skripsky, Die Online-Kunstauktion, 38.
115 On the phenomenon of authorities in the art market, see Chapter 2, Section 2.2.1.
116 Analysed in greater detail in Chapter 2, Section 2.3.
117 See Friedländer, “On Forgeries,” 41.
118 Karlen, “Fakes, Forgeries, and Expert Opinions,” 21, fn. 2.
119 See Gerlach, Die Haftung für fehlerhafte Kunstexpertisen, 12; Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 197; cf. “Faux” in Rey-Debove and Rey, Le Nouveau Petit Robert de la langue française, 1760.
120 Clark, “The Perfect Fake,” 10.
121 Friedländer, “On Forgeries,” 39.
122 See Karlen, “Fakes, Forgeries, and Expert Opinions,” 21, fn. 2; Lequette-de-Kervenoaël, L’authenticité des œuvres d’art, 118 (n. 128) and 125 (n. 131–132); Dutton, “Authenticity in Art,” 259.