LARB Quarterly Journal larb blog faking galileo

Faking Galileo: From the LARB Quarterly Journal, Spring 2014

| by

The following is a feature article from the newly released LARB Quarterly Journal: Spring 2014 edition. To pick up your copy of the Journal, become a member of the Los Angeles Review of Books at the $11 monthly level or order a copy at or b&


ART FORGERIES have long been the stuff of thrillers, with fake da Vincis or Vermeers fooling connoisseurs, roiling the art world, and moving millions of dollars. We don’t think of ancient books driving such grand forgery, intrigue, and schadenfreude. This is changing thanks in part to a clever forgery of Galileo’s landmark book Sidereus Nuncius, published in Venice in 1610. Arguably one of the most extraordinary scientific publications of all times, Sidereus Nuncius turned Galileo into the brightest new star of Western science. Four centuries later, a faked copy of this book has disarmed a generation of Galileo experts, and raised a host of intriguing questions about the social nature of scholarly authentication, the precariousness of truth, and the revelatory power of fakes.

Galileo Makes a Telescope

The story begins in 1609, when Galileo was a professor at the University of Padua, then under the rule of the Republic of Venice. The 45-year-old Galileo was on his way to toppling Ptolemaic and Aristotelian conceits, revolutionizing the principles of motion, and redefining what physics was and meant. He was also struggling with financial problems, saddled with supporting a lover and three children, with the dowry for a sister’s marriage, and with a hapless brother. With little hope for a salary increase, he borrowed money, gave private lessons, and made instruments to sell.

In July 1609, he heard about a spyglass being made in Holland. Quickly, he figured out the requisite shape and combination of lenses needed to go from two or three orders of magnification to eight. Just one month later he presented his invention to the Senate of Venice as a new military technology; it would, he told the Senate, make it possible to spot enemy ships at sea a full two hours before a naked-eye observer. There was no such thing as our notion of a patent in those days, so Galileo needed to be strategic in order to benefit from his inventions and discoveries. To this end, he offered the telescope as a “gift” to the Senate. It worked. He was rewarded in the form of tenure and a doubling of his salary.

Shortly after, however, spyglasses of ever-increasing power began flooding Europe: Venice had no real advantage over its enemies. Galileo was informed that he would receive no further salary increase and that he was bound to teach in Padua for the rest of his life. Not a fate he welcomed.

Undeterred, he began grinding new and more powerful lenses, while negotiating a position with the Grand Duke of Tuscany. He wanted to move back to Florence, his hometown, and become the court mathematician there. By the end of November, he’d completed a 20-power telescope, which he then, in one of the most famously portentous moves in history, turned to the night sky. What followed is well known in the annals of science: Galileo was able to make astonishing discoveries that transformed natural philosophy and human self-understanding.

Galileo Sees the Moon

The first thing Galileo discovered was that the moon was not smooth and homogeneous, as everyone believed. Instead, it was covered with craters and mountains whose peaks became awash with light when the “terminator” — the line that separates the illuminated and dark parts of the moon — inched forward through the night. Art historians Samuel Edgerton and Horst Bredekamp have written insightfully about how his skills as a draftsman were key to this discovery. Young artists in training during this period were drilled on treatises designed to, in effect, reshape their perception, so that they unthinkingly interpreted certain configurations of two-dimensional light and dark shapes as the surfaces of three-dimensional figures hit by a light source. Galileo’s draftsman eye thus gave him a crucial advantage over other observers, such as Englishman Thomas Harriot, who, a couple of months earlier, had carried out the first recorded telescopic observation of the moon. To Harriot the moon remained smooth and the terminator a fairly clean line. He only saw mountains and craters after he learned of Galileo’s novel description.

The implications of Galileo’s discovery were mindboggling. Aristotelian physics had been based on a fundamental distinction between Earth and the Heavens. Everything on Earth was subject to processes of corruption and change. The Heavens were incorruptible, made of perfectly smooth material, and moved only along circular paths. A pockmarked moon made no sense.

Galileo’s telescope was about to deliver even more shocking news. In the clear sky of January 1610, he pointed it toward Jupiter, and noticed three small stars peculiarly aligned next to it. He recorded their position on a now-famous piece of paper. The following night, he could scarcely believe his eyes: they had moved. And now there were four. A few nights later, Galileo realized that they were not stars but planets orbiting Jupiter as it moved westward against the backdrop of the fixed stars. For the first time ever, someone had observed a celestial body that orbited around something that was not Earth. This was a formidable blow to both the Ptolemaic system and Aristotelian physics, which did not allow for multiple centers of gravity. Galileo’s discoveries spelled the end of conceptions of Earth, and hence of man, as the center of everything.

A 17th-Century Instant Book

In order to benefit from his discoveries, Galileo needed to publish. Fast. He decided to hastily assemble a small book. In just two weeks he had the first part of the manuscript ready, and by January 31, 1610, he was visiting a printer in Venice. He strategically named the newly discovered planets “Cosmica Sidera” (Cosmican Stars) in homage to his would-be new patron Cosimo de’ Medici, Grand Duke of Tuscany. Unlike the telescope, this was a symbolic gift, but one rich in powerful associations. On February 20, Galileo received the news from Florence — the gift had been accepted, albeit with this caveat: the Grand Duke’s new celestial possessions should rather be called “Medicean Stars” to honor the House of Medici and avoid confusion with the term “cosmic stars.”

An unusual wealth of information on the making of this book exists. We know that Galileo gave precise instructions about the order in which the sheets should be printed, because he was still carrying out observations when the production process began and wanted to have the possibility to make last minute insertions. We know that the Grand Duke’s call for a name change must have been a nuisance: by February 20, when the reply arrived, a sheet with the word “Cosmica” had already been printed, so paper slips were applied to correct it. And we know that, while parts of the book were already in press, Galileo realized that he could measure the height of lunar mountains based on their projected shadows; he forthwith added these calculations to the text. The last piece of information to enter the book were the positions of the Medicean Stars on March 2. The dedicatory preface is dated March 12. The final title was Sidereus Nuncius, which means the Sidereal Messenger. The book was completed in the nick of time to be sent to the Frankfurt book fair, the most important fair in Europe.

The printed text contained empty spaces for five etchings of the moon (four images, one repeated), to be added later. These were to be based on Galileo’s own watercolors, the originals of which are on undated sheets known as the Florentine Bifolium (henceforth FB), part of the Galileo papers now at the National Library in Florence. These sheets are among the most fetishized objects in the history of science. They contain stunningly vivid sepia wash studies of the moon in different phases, which have been dated astronomically from November 30, 1609, to January 16, 1610 — in other words they stop when Galileo began frenetically assembling his book.

Four images from the original FB were probably the models for the book’s etchings, via a now-lost intermediate set of drawings. (These may not have survived the procedure). The lunar images in the book are much less nuanced than those in FB, and in fact contain exaggerated features — for instance, a starkly delineated giant crater. Historian of science Mario Biagioli argues that this and other inaccuracies had a precise rhetorical function: Galileo wasn’t so much interested in mapping the moon but rather in the movements and changing appearance of the terminator. In other words, he was interested in topography, not cartography. The unrealistic giant crater then plays the same role as an “inset” does in contemporary scientific illustration.

Harvard astronomer and historian of science Owen Gingerich has called the printing of Sidereus Nuncius “the most extraordinary episode in the history of scientific publishing.”

A $10 Million Book

In 2005, two Italian book traders paid a visit to Richard Lan, who owns Martayan Lan, a prestigious antiquarian bookshop in New York City. They offered to sell a copy of Sidereus Nuncius — a special copy: the title page was signed by Galileo himself. Another selling point: it carried stamps of the library of Prince Federico Cesi, a friend of Galileo and the founder of the Lyncean Academy in Rome, the first scientific academy in the world. The most striking feature of this copy, however, was that the usual etchings were missing. In their stead were five watercolors. Galileo had claimed that 30 copies had gone out un-etched, a dozen of which are known today. Maybe this was one of those copies. Lan showed the book to Gingerich, who at first thought it possible that those drawings were connected to Galileo. Lan figured that if this was so, then the book might well fetch something like $10 million. He bought it for half a million. He then invited the preeminent German art historian Horst Bredekamp to inspect the book. Not only did he authenticate the signature, but also recognized Galileo’s hand in the sketches. But how to make sense of this unusual copy? Bredekamp became convinced it was a sort of proof copy, and that Galileo himself had filled the blank spaces with watercolors. Whoever made the etchings then transported these sketches onto the copperplates. In fact, these watercolors resembled the etchings much more closely than the images in FB. It followed that FB was not their source but probably a later set of images produced in anticipation of a never-realized second edition ofSidereus Nuncius.

Bredekamp’s interpretations thus implied a significant revision of the story of the making of the book, of Galileo’s working style, and of the dating of Galileo’s observations. These results were included in Galilei der Künstler, a detailed and lavishly illustrated 500-page book that Bredekamp published in 2007. William Shea, Galileo Chair of History of Science at the University of Padua, saluted the discovery of the Martayan Lan copy (henceforth ML) and declared himself convinced that the watercolors were indeed made by Galileo.

In 2008, the book was sent to Berlin, where Bredekamp had assembled an international team of experts to scrutinize ML and reveal its secrets. In 2011, the final results of their research appeared as Galileo’s O, a two-volume set. One can safely say that no other bookmaking process, aside from those of Gutenberg’s Bible and Shakespeare’s First Folio, has received a comparable degree of attention.


The first volume of Galileo’s O showcases the team’s interdisciplinary research, and is devoted to comparing ML to other known copies ofSidereus Nuncius. In one essay, a curator at the National Academy of Lynceans (the current incarnation of the ancient academy) identifies the stamp in ML as one that was used in the private library of Prince Cesi. Galileo, she argues, had probably given the book to Cesi during his visit to Rome in 1612. This essay contains tokens of Lyncean imagery and a photo of the ML stamp, but, strikingly, does not compare it to other known specimens of the same stamp. In another essay, Bredekamp himself takes on the issue of the signature on the title page. He argues that it is very similar to one found in a contemporary dedication and exhaustively analyzes it letter by letter. His conclusion: “Its authenticity is beyond doubt.” Bredekamp also compares ML’s watercolors to those of FB and again emphasizes similarities, recognizing Galileo’s hand in both sets.

The central section of the book presents high-quality images of each and every page of ML, while in the subsequent sections ML’s paper, ink, letterpress, page layout, and watermarks are analyzed, compared to those of other known copies, and declared authentic. Again, however, images of key elements such as the gauffered page edges, the headband, or the sewing structure are missing. No doubt ML presents some idiosyncrasies, concede the authors, both in its materials and in its printing. But this is hardly damning evidence: ML was, after all, a proof copy.

The second volume is comprised of a long essay by Paul Needham, Scheide Librarian at Princeton University, and formerly head of the Morgan Library in New York, on the composition and printing of Sidereus Nuncius. It should be compulsory reading for all those curious about the making of early modern books. Needham is interested primarily in the book as a material object and so focuses on the complex relations between Galileo’s draft, his sketches, and the printed text. The in-depth analysis of each material component along with the available technology enables readers to understand the book as an artifact constituted by the interplay between the author’s conceptual priorities, his rhetorical strategies, and the material possibilities and limitations of early 17th-century printing technology. ML is not central to Needham’s narrative, but he is convinced that it is authentic, and his overall reconstruction supports Bredekamp’s interpretation of ML as a proof copy. Needham also agrees that ML’s sketches are much closer to the etchings than are those in FB, and hence must be their original source. This is an important point, and one of Bredekamp’s central claims: the ML images functioned as preparatory sketches for the etchings. This means that they had to be transposed onto the copperplates, and survive the process unscathed. How could this happen? Bredekamp offers a theory: the watercolors were placed face down onto an inked surface, their key features pressed lightly to pick up the ink. This ink was then transferred onto the etching plate. This practice, though not described by any contemporary source, was confirmed by microscopic traces of black ink on top of the watercolors.

The two elegant volumes provided authoritative authentication of ML, and also, just as importantly, a revision of the history of the most sensational scientific publication of all time.


There were still doubters. Gingerich himself, who had seen ML in 2005, had been growing skeptical about the drawings. In an article in 2009, he offered an extensive critique of Bredekamp’s “astonishing but dubious claim.” Bredekamp’s new reconstructions of events neglected the astronomical constraints — i.e., when certain phases of the moon were visible. They also made little practical sense: allegedly Galileo had sketched his lunar images on a sheet now lost, then copied them onto an early copy of Sidereus Nuncius as it came out of the press, and then had these images traced and transferred onto copperplates for the etchings. But why not trace the original drawings? Why delay the preparation of the etchings by weeks? Gingerich thought that the best way to make sense of these incongruities was to conclude that the ML drawings were a forgery.

Furthermore, he pointed at problems with the lunar representations. In one of the ML images, the orientation of the moon is turned sideways, so that the “up” of the image is on the right and the terminator appears as a horizontal line. There is no good reason to alter the representation in such a confusing way — unless someone was copying the images from FB; for reasons of space, Galileo had to turn an image sideways while making clear where the “up” was through the position of its number. That detail, Gingerich reasoned, was lost on the forger, who clearly did not understand the astronomical meaning of the image. Gingerich pointed at other smaller details that could only have been derived from FB, and so concluded that the copier had used both FB and the etchings as sources. He also noticed that the circle of the moon in one image of ML was smaller than in the corresponding etching: how could it then be the original source? Finally, as different sections of Sidereus Nuncius had been printed at different times during the six-week period of its making, how could this alleged proof copy have been assembled in advance?

Gingerich’s damning article came out in Galilaeana, a specialized journal. Bredekamp’s team does not seem to have taken it seriously. William Shea, the aforementioned Galileo Chair at the University of Padua, replied directly to Gingerich in the same journal in 2010, arguing that Bredekamp’s book of 2007 had already provided all the answers to these questions, which he reproduces. Some of Shea’s arguments, however, are rather puzzling. For example, in response to the sideways image problem, Shea — following Bredekamp — argued that the alleged forger would perforce have to be skilled and knowledgeable and so would be unlikely to miss its meaning. As for why ML was not discovered before? Well, it was bound together with a later and rather unremarkable edition of Galileo’s works. Apparently, the mysterious provenance of the book — Lan had only disclosed that it came from Argentina — did not raise suspicions.

Shea made the valid point that Bredekamp has been providing, for many years, invaluable and enlightening accounts of the close relationship between art and science during the scientific revolution. Shea clearly saw the authentication of ML as another stage in this fascinating research project. Indeed, if anything, he was awed by the way Bredekamp and Needham had brought together the history of art, the history of the book, and the history of science in these two volumes. “I am sanguine enough,” he wrote, “to believe that even I can appreciate the hand of Galileo at work in [ML].”

Authenticating ML was for him and his fellow ML advocates the culmination of a two-decades-long, exhilarating intellectual adventure.

A Fresh Pair of Eyes

In the summer of 2011, Nick Wilding, an assistant professor at Georgia State University, received a copy of Galileo’s O for review. Wilding was working on his first book, a study of how Galileo’s network of patrons and friends had shaped the making of Sidereus Nuncius. He found the idea of a 17th-century proof copy bizarre; as far as he knew, it did not fit with contemporary bookmaking practices. Also, he was aware of rumors about dubious library stamps that had surfaced in the antiquarian market in 2005 and 2006. He obtained copies of genuine stamps of Cesi’s personal library (a lynx surrounded by an inscription between two lines) and realized that they all shared a minor detail: one of the lines that circled the lynx was broken. In the ML stamp the line was continuous. All other dubious copies shared the same feature. Wilding also verified that there was no trace of a copy of Sidereus Nuncius in Cesi’s library inventory. John Heilbron, a Berkeley historian of science and a biographer of Galileo, also commented on this absence. Heilbron did not hide his skepticism about ML, especially as Cesi’s stamps had been seen on other recently discovered unrecorded books. “Perhaps they did not keep good records,” he commented wryly.

Wilding decided to contact Gingerich. He learned that, between 2005 and 2006, Gingerich had inspected three suspicious copies of Le operazioni del compasso geometrico e militare. This is a very rare small book published by Galileo in 1606, of which only about 25 copies survive today. That three new copies should surface in a year was odd indeed. Gingerich was unconvinced by the watermarks and by the depth of the printing, which was unusual given the machinery available in the early 17th century. Wilding also discovered that, in 2006, Frank Mowery, an expert paper conservator at the Folger Shakespeare Library, had inspected two of these copies ofCompasso for a private collector and declared them to be forgeries. Mowery had noted a strange distortion of a couple of words in the text that could never happen with letterpress but that is typical of digital reproductions. Most likely, he reasoned, someone had scanned the text of Compasso, produced a photopolymer plate for each page, and printed it. This is done by placing a negative image of the page over the light sensitive plate surface, and letting the light harden the parts that are not protected by the black of the film. The rest is washed away, leaving a relief surface from which to print. While this technology has been around since the 1980s, it is now widely accessible and relatively inexpensive. This piece of information was key for Wilding, who began seeing analogies between the forged copies of Compasso and ML. He also became aware that the forged books could be traced to Marino Massimo De Caro, one of the traders who had brought ML to Lan in 2005.

Looking for further hints, Wilding noticed that a word on ML’s title page contained a strange mistake: it read “pepiodis” rather than “periodis.” No other copy he had seen contained that mistake. A copy offered by Sotheby’s in 2005, also provided by De Caro, and with the lunar images supplied, according to the catalog, “in facsimile,” had the same misprint. There was a further problem: the letters p and i touched each other in a way that is impossible to produce with letterpress — but that can happen with digital reproductions. Wilding also noticed a significant inked blotch that deformed a letter at the bottom of the title page, something that was missing in other copies, but not Sotheby’s. He realized that the same deformation could be seen in a 1964 facsimile edition of Sidereus Nuncius, where a brownish area on the paper of the original copy had turned black in the reproduction process. In June 2012, Wilding contacted Needham and asked him to check this mark in ML and also the faint signs of inking that appeared at the margins of some pages. In a copy produced with letterpress printing, they appear only as superficial traces known as “shoulder ink.” In making a polymer plate, however, any dark sign on the original page becomes the equivalent of a character and, if it’s not cleaned out, will be inked and bite deeply into the page. After several weeks and many conversations with Wilding, and after comparing ML with a genuine copy from Columbia University, Needham became convinced that ML was an outright forgery.

Meanwhile, in Italy

In those same weeks, in an apparently unrelated series of events, the Italian police were putting together the pieces of a complex investigation that had revealed a massive theft of antique books. They had discovered that thousands of rare volumes had been removed from the baroque Biblioteca dei Girolamini in Naples to be sold at auctions internationally. The traces of their provenance had been erased and the original catalogs mutilated, making it extremely difficult to identify and recover the stolen volumes. The evidence implicated the then-director of the Biblioteca dei Girolamini, Marino Massimo De Caro. During the course of the investigation, De Caro admitted to having coordinated the production of a few fake antique books; however, the police did not pursue this lead, their energy being entirely concentrated on the stunning Girolamini theft and the criminal network that had made it possible.

In March 2013, De Caro was sentenced to seven years in prison, later modified to house arrest. In December 2013, The New Yorker published an excellent article on this case by Nicholas Schmidle, who had traveled to Italy to interview De Caro. De Caro admitted to Schmidle that he had forged ML with the collaboration of a few specialists from Italy and Argentina. Proudly, De Caro went into a detailed description of the making of the artifact, including the technique of baking the printed pages in an oven to age them. A street artist was commissioned to fake the watercolors of the moon, claimed De Caro.

One should handle De Caro’s confession with care. The people he mentioned have denied having anything to do with the forgeries. Also, there are gaps in his story, and, interestingly, he doesn’t take responsibility for all the Galileo forgeries that have emerged thus far. At this point, it is unclear what role he played in the making of such forgeries, and the team of specialists that created them remains at large.

Unconscious Collaborators

Back in June 2012, Bredekamp had met Wilding’s assertion that ML was a forgery with incredulity and skepticism. He declared himself deeply offended. After all, he had been guaranteed that its paper and binding were authentic. The evidence brought to him by Needham, however, was overwhelming. In October 2012, part of the team of experts that had authored Galileo’s O met once again in Berlin, with the task of reverse engineering what was now seen as a very well-crafted forgery. The team, which included some new specialists, carried out a second round of analyses, whose results were published early in 2014. Hence we arrive, patient reader, to the volume presently under review.

It is conceived as a third volume of Galileo’s O, an unplanned addition to the original two-volume set. A slim book, just 100 pages, it is clothed in a neutral light gray cover. The key point is made clear by its laconic title: A Galileo Forgery. The introduction mentions the correspondence between Needham and Wilding in May and June 2012 and admits that the very idea that ML might be a forgery was “unimaginable” for the members of the team. However, they had been forced to return once again to this object by the “logical rigidity of the newly presented facts” and their awareness that a number of forged Galileo books had been identified. The short chapters that follow contain expert assessments of ML’s printing, paper, and binding; a comparison to the forged copy of Compasso; the results of new scientific tests on its materials; and some concluding remarks.

Needham’s chapter is thoughtful and informative. Describing the forged printing, he shows examples of digital retouching, including capital letters that appear damaged and then “healed” in sequences that do not make typographic sense — a “not in this universe situation.” The sewing of the volume also looks suspicious this time around: ML shows a different sewing pattern from the text that is bound with it. As for the signature that Bredekamp had authenticated, Needham discovers here that in fact it is very similar to Galileo’s genuine signature, but from a much later period. The forger’s models were the signatures Galileo produced for the documents of the Inquisition during the 1633 trial in Rome, whose images had been often reproduced. The anachronism is now evident: “It is not the signature of the forty-six-year-old man to whose eyes the gates of the heavens had recently begun to swing open, but of the sixty-nine-year-old, aged and ill and threatened with torture.

As for the paper of ML, it “remained something of a mystery.” In their chapter, the specialists who had formerly pronounced it genuine now present some important quibbles. The watermarks are similar to genuine ones, but how similar? They now see a number of oddities separating ML from other copies of Sidereus Nuncius. Even the paper stock looks significantly different. They did not think this was suspicious in the first round of tests, as it made sense, they argue, that a lower quality paper had been used for a proof copy. Also, they “had not considered fiber samples [of ML] because the originality of the paper had not been disputed.” In short, they had not thoroughly analyzed the paper, as they had taken for granted the authenticity of ML. In this second round of tests, evidence of the forgery emerges in incontrovertible fashion in the form of cotton linters, which began to be used in papermaking only in the late 18th century.

Plenty of other evidence emerges about the watermarks and procedures used to produce the forged paper sheets. A number of features that were at first considered “odd but convincingly authentic” are reinterpreted as indicators of forgery. For example, the strange combination of thin and thick sheets, which was believed to be functional to provide the book with certain structural characteristics, is now seen as a way to achieve the correct watermark sequence. Or, again, the dark lumps noticed in the paper, which were at first consistent with the “low quality interpretation,” have now become “inconsistent with seventeenth-century papermaking practice.” “The forgers,” the authors comment, “went overboard in faking some of these historic marks.” Similarly, the depth of the printing, which was so convincingly authentic in 2005, now looks patently anachronistic.

Many telling passages in this third volume address what is described as a gestalt shift of sorts: the brownish soiling of the pages, for example, is now “a sure sign of intent to deceive.” The soiling is well made but is not where it should be on the page if it were the result of historic handling. Its patterns on the pages are now “strange.” It is as if the authors’ “belief in the authenticity of the book convinced [them] that what [they] saw was natural.” Once their eyes were “awakened,” they “saw what was unnoticed before.” In a short chapter, the parchment cover and its decoration are confirmed to match the date of the text bound with ML: 1655. The construction of the binding, however, now reveals the insertion into the bookblock of an element that was not originally part of it. In particular, “the change of thread and sewing technique” between ML and the comparison text “indicates beyond reasonable doubt that [ML] has been inserted into a pre-existing bookblock.” Many other features that “do not fit with historical practice” are further elucidated in this manner. A chapter is devoted to a fourth forged copy of the Compasso, discovered by Wilding in Padua, which presents similar characteristics to ML but is cruder in its execution and therefore can give useful indications about forging techniques.

Yet another chapter is devoted to the results of nondestructive instrumental analysis — X-ray fluorescence, infrared spectroscopy, and confocal microscopy. Tests using these methods were carried out on a comparison text as well — namely, a genuine copy of Sidereus Nuncius. We know expertise can fail, but surely not the heavy artillery of scientific authentication? Ironically, the results presented here make for the most inconclusive chapter of the book. After pages of graphs and data, one realizes that none of it proves that ML is a forgery. Paper and ink are certainly different, but they could both date to the 17th century. No anachronistic material is detected. (Cotton lint can only be detected by analyzing paper fiber samples.) The author limply concludes by pointing to the general limitations of material analysis methods — whether destructive or nondestructive: “The forgers could also have employed materials from the proper era to create the forgery.”

Bredekamp’s own chapter is short and contains less than compelling comments on the psychology of forgers, a topic that has already generated a vast and important literature. He also dwells on these forgers’ allegedly astonishing abilities, arguing that they must have inspected multiple copies of Sidereus Nuncius in order to come up with the exact print errors that one would expect to find in the proof copy. But the fact is that if the forger used a couple of facsimile editions to produce the polymer plates, as indeed seems likely, then such considerations are hardly relevant. Here, as in his final remarks, Bredekamp insists on the exceptional nature of this forgery and the astonishing investment of time and resources that must have gone into it. Only this extraordinary sophistication, he insists, can explain why the team of experts authenticated ML in 2005.

Needham, by contrast, is not at all convinced that this was a unique, almost perfect forgery. Instead, he is astonished by his own blindness. Armed with a simple magnifying glass, an expert should, he declares, have spotted straight away that this work had not been printed from movable type. “Fatal flaws abound” in ML. The credibility of ML, he argues, has less to do with the cleverness of the forgers than with “[his] own failure to consider all possibilities” inherent in the typographic anomalies he had noticed. In fact, he suggests labeling his own failure as an “unconscious collaboration in forgery.” During his first inspection of ML — and knowing that Bredekamp had authenticated the drawings — Needham noticed anomalies and differences between ML and other known copies, but he “formed a fantasy”: that ML was a collection of proof printings of Sidereus Nuncius. He did not even consider the obvious alternative: that ML was a forgery. He only saw ML “truly” and with “better schooled” eyes four years later, “on a brilliant day in late spring,” thanks to the “persistent skepticism” of Nick Wilding.

Fake Matters

Some reviewers of the first two volumes of Galileo’s O, aware of the rumors about the authenticity of ML, argued that the controversy was hardly relevant for the historian of science. I disagree. It is true that those volumes contain much scholarship that remains unaffected by ML’s shift in status, and that Horst Bredekamp’s fundamental contributions on the intersection of science and art are unaffected by the ML attribution. The affair, however, raises broader issues about the way we deal with credibility, evidence, and proof in our own research.

Needham, for instance, now sees the very features that had convinced him of ML’s authenticity as clear signs of an intent to deceive. Until recently, forged printed texts have been relatively uncommon, and most of them have been produced with techniques that did not give depth to the typographic characters. One of the reasons ML rang true was the way its characters bit the paper. But the forger exaggerated this feature: the printing was deeper than it should have been in a 17th-century book. Similarly, the soiling on the pages and the impurities in the paper were overdone. These features were designed to meet the expectations of an expert assessor. In other words, only a forgery could have seemed so authentic. The ML’s many typographic anomalies were chalked up to ML being a proof copy, and thus they ended up reinforcing the perception of its authenticity. When the international team of experts met to analyze ML in 2008, they already shared a collective representation of ML as a collection of proof printings, one that contained drawings that had been authoritatively authenticated. The evidence that they encountered seemed consistent with this representation.

It would be naive to think that all we have here is a straightforward example of scientific malpractice. While there were certainly weaknesses in the initial expert assessment of ML, there was nothing pathological about it. If we have learned anything from the past 50 years in the history and sociology of science, it is that the hermeneutic circle that I have sketchily described here is the way in which knowledge, including scientific knowledge of the most sophisticated kind, is routinely produced and validated. Authenticating an artifact is a particular case of knowing the properties of an object, be it natural or artificial. Such knowledge is best understood as the outcome of a social process, one that involves trust as well as forms of inductive and deductive reasoning. This reasoning is itself social — our perception of what is relevant in any particular situation is channeled by the culture in which we operate. Much in this story has to do with similarity relationships: we came across instances of the same stamp, the same watermark, the same signature, which ended up not being the same at all. The point is that no two perceptible objects are ever identical: we have to learn which traits are important, and what can be ignored. We have to train our eye and our judgment, and we can only do so in coordination with other practitioners. One implication of this view is that there is no uncontested way to split empirical evidence from the knowledge of the observer. Knowledge, no matter how technical, always bears the marks of its historical contingency.

Forgery, and the professional detection of forgery, are practices that captivate our imagination and reveal much about our sense of the past. The historical and anthropological literature shows that authenticity and forgery are themselves contingent cultural constructions that have changed enormously over time and space. This does not imply that discriminating between the two is a futile exercise. The entire history of Western scholarship, as shown by the Princeton historian Anthony Grafton, is shaped by the long-lasting struggle of critic and forger, a quest for truth that has transformed our perception of the past. Today, in the fine arts, fakes are attracting growing interest precisely because they tell us so much about the way artists are understood. A fake da Vinci from the 19th century, for example, looks irredeemably romantic to our eyes, to the point that it is difficult to see how an expert could possibly have mistaken it for the real thing. What does the ML forgery tell us about the way we see Galileo and the making of early modern science? This would be a study of a fake against the intentions of those who produced it, in a mode pioneered by historian Carlo Ginzburg, explorer of the thick relations among the true, the fictional, and the false — the pretense that advertises itself as true. While well equipped to study the constructive dimension of knowledge, historians and sociologists of science seem less inclined to explore the implications of these insights for their own research.


There are exceptions. A few years ago, Ken Alder — a historian of science at Northwestern University — presented to various academic audiences a newly discovered letter written by 19th-century forger Denis Vrain-Lucas. This notorious forger had managed to sell to Michel Chasles, a well-known mathematician and member of the Institut de France, thousands of fake letters, many of them from the protagonists of the scientific revolution. Chasles aimed to use the letters to demonstrate the priority of French science over that of its neighboring nations at a time of great anxiety about the perceived decline of the French scientific tradition. It was Pascal and not Newton, he then knew, who had first formulated the universal law of gravitation. Chasles was ready to put up with massive problems of consistency and chronology — in addition to the fact that all these letters, including those from antiquity, were written in French. The letter discovered by Alder had been composed by Vrain-Lucas during his long stay in prison. In it he declares his sincere admiration for Chasles and praises his fight for the honor of French science. He is unable to understand, however, why Chasles would be so ungrateful. After all, he insists, wasn’t I always providing you, week after week, the very evidence you needed? Didn’t we work together on this project? How could you not see that what I was selling you for a few francs was just a projection of your desires?

As you might have guessed, Alder had made it up. Through the voice of the pseudo-Vrain-Lucas, he was inviting a reflexive exercise: how do we handle evidence and proof in our own work, and in that of our respected colleagues? How do our expectations impact our interpretations? How do we monitor our own projections? Bredekamp refers to this in the final — unexpected and illuminating — sentence of the book: “I have learnt, in a bitter way, what I knew before, but not in this concrete sense: that phenomena can be looked at from different perspectives and that from different viewpoints they tell completely different narratives.”

Bredekamp’s first instinctive reaction to the possibility that ML might be a fake had been one of incredulity and despair: if that turned out to be the case, he wrote to Wilding, the “history of science could close its doors.”

As unsettling and frustrating as the ML story might be, there is no need to despair. One can imagine that historians of science — and indeed whoever works with printed sources — will gradually accommodate the presence of forgery, joining other scholarly communities who have been dealing with it for centuries. As for the more fundamental issue that all narratives, including our own, contain constructive elements, this awareness is not fatal to truth. It is fatal to the illusion that truth can be other than a precarious and contestable achievement.


Massimo Mazzotti teaches history of science in the Department of History at UC Berkeley, where he is the Director of the Center for Science, Technology, Medicine, and Society.

Leave a Reply