American Board of Information Security and Computer Forensics

by John Daab, PhD, MA, RI, CFC

Over the past few years, there has been a spate of fine art incidents in which the works of famous artists have been found in rag shops, attics, or generally just lying around, waiting to be discovered. There was the trucker’s find of a Jackson Pollock (Kennedy, 2006), the rediscovery of a Christie’s attributed German work, now thought to be a Da Vinci (Israely, 2009), and a Pollock from Long Island (Philips, 2007). In all of these discoveries, the self-proclaimed forensic expert who arrived at the conclusion that the works were genuine was untrained and further lacked any credential demonstrating that his expertise was legitimate. More importantly, some of the determinations were dead wrong. In one case, Pat Wertheim, a noted fingerprint expert, found that an alleged print claiming authenticity was in fact a digitally applied one (Hochfield, 2008). In other cases, the individuals verifying authenticity were appraisers. Please note that appraisers are warned by their appraisal organizations to refrain from engaging in any authentication. The organizations note that clients do not want to pay for such a process, that authentication calls lead to litigation, and that appraisers are not trained to perform authentication (Art Appraisers Association of America, 2009). In fact, appraisal organizations offer little or no coursework in art authentication. These episodes of so-called forensic science are making the news and underscoring the weak controls in the science of authenticating objects of art. Instead of having real forensic experts testing for matches between real and fake works of art, we have the forensically untrained taking up the slack (New York University, 2009).

A Brief History of the Fine Art Authentication Process

The process of authenticating objects of art such as paintings, antiques, glass, silver, and rugs follows three distinct operations—scientific analysis of the media for age; provenance or identifying documents supporting ownership; and connoisseurship, when a connoisseur looks at the creative style of the work and makes a final authentication judgment (Museo d’Arte e Scienza, 2009). One could argue that much of the process really does not enter into the domain of forensics since it is more of a subjective arena, often historical. This arena consists of individuals with a background in art history or art collection such as Bernard Berenson (Secrest, 2005) or Eugene Thaw, who would look at a work and pass judgment on its authenticity (Aronson, 2006). This view is not far off the mark. Historically, such individuals have ruled the authentication process; that is, until Thomas Hoving noted that 40% of museum artwork is of questionable authenticity (Hoving, 1997). The former director of the Metropolitan Museum of Art, Hoving based his point on the results of a team of museum experts that examined 55,000 objects of art. The Rembrandt Project came to a similar conclusion following similar methods in the examination of 627 alleged Rembrandt works. Only 267 were deemed authentic (Gash, 1993). The fact that the connoisseur has been supplanted by a forensic/scientific approach marks a significant turning point in the authentication model. It is important to note that current forensic approaches have been involved only minimally, just scratching the surface of a complex process. Fingerprinting, digital analysis, and chaos theory have arisen in the examination of fine art, but standard forensic processes are absent from the analysis or fail to follow up and focus on deeper analysis (Daab, 2009). Fingerprint analysis fails in the most basic approach of securing legitimate exemplar prints for comparison; digital analysis looks to authenticate by choosing one-inch examples of artistic style, forgetting chain-of-custody requirements on how the samples were taken, stored, and protected. Chaos theory, via the use of geometrically indeterminate structures called fractals (think snowflakes, coastlines, tree branches), attempts to generate an empirical approach but is flawed; predictability is logically prohibited from indeterminacy (Taylor, 2002).

In the area of provenance research, document searches are carried out and signatures are traced, but no attempt is made to provide the depth of a questioned document approach. Documents are provided, but little analysis takes place to verify that the signer’s signature or writing matches that of the writer in terms of slant, line following, spelling, use of print versus cursive, and idiomatic use (Daab, Questioned Documents, 2009).

One would not expect forensic examiners to be involved in connoisseurship, but there is no reason to assume that the subjective realm of denoting a match between style, content, and subject matter cannot be developed into a more objective, scientific endeavor. There is no reason why the points on a fingerprint establishing a match cannot be a fitting model for connoisseur matches; for instance, this painting is authentic because the connoisseur points of style, content, subject matter etc., match the points of the exemplar work. This is not to reach out for an unequivocal proof of match, but one based on descriptive data rather than unsupported feelings or presuppositions.

The Current Fine Art Authentication Paradigm

Scientific Analysis of the Medium

The current method of ascertaining authenticity in a given piece of art begins with the scientific age analysis of the media. Here, the particular medium or material used to create the object is examined to determine age and makeup. Forensic scientists will look at wood panels, paint pigments, underlayments, interior structures, chemical sources, canvases, and stretchers upon which a work is applied or built. The logic of science in the analysis is grounded in the notion that if scientific fact establishes that the medium does not support the time period of a work’s creation, the work is not authentic and no further testing is necessary. If science finds that a panel of wood supporting a Da Vinci work is incompatible with the known time of the work’s creation, it follows that the work could not have been done by Da Vinci. If the paint pigment was created in 1945 and Da Vinci created the work in the 1500s, it goes without saying that unless Da Vinci traveled around in a time machine, he could not possibly have painted it. Although science may discount or provide evidence that a work was a fake, scientific analysis may not establish that a work is real. If the components of a work have been corroborated as stemming from a certain period, that in itself does not necessarily mean that the work is authentic.

Some sample tests of artistic materials consist of:

Infrared reflectography focuses on the layers of painting and the sketch made by the use of carbon black or charcoal by the artist. The different layers or alterations of the work are known as pentimenti. The alterations reveal the extent of restoration, whether or not it is a copy, and the different materials used. The more pentimenti present, the more likely it is that the work is the original (Art Institute of Chicago, n.d.).

Wood’s Light seeks to identify the different types of fluorescent matter in a work. Fluorescent matter emits radiation, and the quantitative levels emitted link the work to a particular time (Museo d’Arte e Scienza, 2009).

Microscopic analysis allows the viewer to see beneath the surface and identify the particles within the structure of the work. Particles or minute matter comprising the work have birth and death times; materials are created at a particular point in time and fade out of circulation from disuse, exhaustion (supply no longer available), or the implementation of regulations outlawing their use. The pigment Verona Green was exhausted in 1930, yellow iron oxide (Mars yellow) was created in the lab sometime in the 1920s, and the use of lead in paint was outlawed for the most part in the 1980s due to negative health consequences. The type of pigment used in a work tells us in what time frame the work was created or restored (Derbyshire, 1999).

Carbon 14 (C-14) tests the age of a given compound by focusing on the quantity of C-14 material left as the compound decays over time. It is less reliable as a predictor of ages closer to the present and ages beyond 60,000 years. The accuracy and reliability of C-14 increases with age, up to about 60,000 years, and like a bell curve, begins to decrease. Less accuracy is associated with materials made early and after 60,000 years; greater accuracy is associated with materials found in an area midpoint in the curve. The maximum deviation or inaccuracy occurs for materials 600-700 years old (Museo d’Arte e Scienza, 2009).

Some general principles relevant to a scientific approach to keep in mind:

Analysis may determine an age range but not a precise date.

The age range may discount an object of art from a given period of time.

Age range determination of a medium does not guarantee the art is authentic.

Scientific determination of age range is shaped by the technologies in the process of testing, and as such may not be totally accurate.

The logic of testing entails that if the age does not corroborate the time of creation, further analysis is unnecessary.

There are fraudulent methods available to overcome testing for material age.

The sample methods noted above are only some of many methods of testing.

New scientific methods developed must be treated with suspicion unless they are accepted by the scientific community.

Recently created objects may not be suitable for testing since testing may not be able to determine age.

Provenance

If scientific analysis of the materials of the objects of art determines that the work is from a certain time frame, research may take place concerning the past ownership or possession of the object. Provenance analysis focuses on documents surrounding the work in terms of where it has been. This examination looks at sales slips, taxes paid, letters of possession, estate transfers, loans, and liens on the object, exhibition history, catalogue raisonne, custom duties paid, and title searches (Harvard, 2008). In the best of all worlds, all this information would be available from the time the work went from the artist to those who later possessed it. This rarely happens. In the most recent case of the new Da Vinci, very little documentation is available to demonstrate where the work has been over the last 500 years or that it existed sometime in the past as a Da Vinci. If in fact documentation existed, questioned document analysis by a forensic document expert rarely takes place. Yes, exemplars for signatures exist, but establishing that the slant, use of idioms, spelling, configuration of letters, and grammar are not used to match documents. After signature, provenance documents seem to be assumed as accurate. Hoving (1997) and Salisbury (2009) make a point of noting that provenance documentation is easily forged. Again, in the absence of a trained record examiner, one could easily assume the opposite: assume inauthentic and then analyze via accepted questioned document standards (University of California, n.d.).

Provenance results, if positive, will turn the examination over to the next stage: connoisseurship.

Connoisseurship

Having passed stage one and stage two, the object of art enters into an examination of connoisseurship. At this stage, art historians, curators, museum directors, and art experts will be called upon to render an opinion that the object falls into a corroborative situation whereby the culture, technology, style, content, subject matter, shade, color, and even size match with the exemplar provided. This stage is subjective, intuitive, and ephemeral. The connoisseur’s call has no standard, or process forging a warranted conclusion. Most opinions come down to solipsist responses ranging from “you just know it,” Muller (2003); “it sings,” Peterson (2002), or “I know it because I was raised among master objects of art,” Aronson (2006). This last connoisseur forgets that more than likely the surrounding art was of questionable authenticity (Hoving, 1997).

One may argue that although scientific analysis and provenance have issues—particularly scientific ones such as changes in technology, standardization skepticism, and theoretical assumptions—connoisseurship leads the way in authentication problems. Judgments based on subjectivity and intuitions often become unraveled as criticisms emerge and a deeper analysis takes place. “It is authentic because I say so” fails scholarly standards and often results in courts throwing the proverbial book at the issuing connoisseur. The significant problems attached to connoisseurship have produced an array of scientific approaches to the field of connoisseurship, such as digital analysis of style, fingerprinting, and chaos theory or fractalization of style analysis. Digital analysis and fingerprinting have problems in that logical considerations befall the first and exemplar and chain of custody issues apply to the second. Chaos theory and the use of fractals is way off the mark, since indeterminate theoretical structures serve only to weaken and contradict a predictable outcome. To assume that the artist Jackson Pollack, a known drug abuser, planned indeterminism in his work seems beyond the ken of intelligent discourse. The lack of empirical structure to present approaches grounded in history leaves connoisseurship open to myriad issues associated with verification, operationalism, falsifiability, lack of testing standards, and a grounding in subjectivity. The historical principle of allowing the connoisseur the final say in the authentication call weakens the current process. This principle is slowly being supplanted by a more objective quantitative grounding.

In summary, the present authentication process consists of scientific analysis, provenance research or providing documentation of ownership through time, and the final touch of establishing that the object was from the hand of the artist by corroborating style, technology, content, and subject matter and so on, with an exemplar of the artist.

A Case Study of the Authentication Process: A Composite Approach History

John Drewe and John Myatt were able to fool many collectors in England with a plenitude of art fakes. Myatt created the fake art, while Drewe provided the documents to establish that the fake works had proper and unassailable provenance. Through the manufacture of bogus catalogs, document insertions into historical archives at the Tate museum, and even a stamped attribution from a religious order, Drewe convinced his dupes that the works were authentic and backed by accredited documentation. An unhappy ex-wife and a connection to a killing set the stage for a lengthy investigation involving museum directors, archivists, buyers, and assistants, auction houses, and connoisseurs, ultimately leading to the arrest and conviction of both Myatt and Drewe for art fraud. The following composite consists of a combination of the fakes examined and recorded in the work of Salisbury and Sujo (2009).