Kouros

An over life-size marble youth in the Archaic Greek manner, standing with the left foot forward and the arms at the sides, bought by the J. Paul Getty Museum in 1985 for around ten million dollars. Its label at the Getty Villa read: Greek, about 530 B.C., or modern forgery. There is no other label like it in a major museum, and it is the most honest sentence in this record.

The statue arrived in Los Angeles in seven pieces with a Swiss provenance going back to 1930, supported by letters. The letters were fabricated: one supposedly written in 1952 carries a postcode that did not exist until 1972, and another of 1955 refers to a bank account that was not opened until 1963. The scientific case for antiquity rested on de-dolomitisation, the leaching of magnesium from the surface of a dolomitic marble, argued by the geologist Stanley Margolis to require centuries and natural conditions. In the early 1990s Miriam Kastner produced the same alteration in a laboratory.

Nineteen invited specialists met in Athens in May 1992 and could not agree. The sculpture came off view in 2018. Nobody has ever named an ancient workshop for it, and nobody has ever proved a modern one.

The label is the object

For most of the time it was on show, the wall text beside this statue at the Getty Villa read: Greek, about 530 B.C., or modern forgery.

Museums do not write labels like that. A label is an institutional assertion, and the whole apparatus of curatorship exists to make one possible. This one refuses. It sets out two propositions separated by two and a half thousand years and declines to prefer either, and it stayed up for years while the museum kept the object on public view underneath it.

Whatever else is true about this sculpture, that sentence is the most valuable thing about it, and it is worth being clear about what it costs an institution to write. It concedes that a purchase of around ten million dollars may have bought nothing; that the museum's own experts cannot settle the question; and that the visitor is being handed an unresolved problem rather than a fact. It is also, on the evidence, correct.

What the thing is

A kouros is a standing nude male figure, an Archaic Greek type used both as a dedication in sanctuaries and as a grave monument. The standard pose is fixed: left foot forward, arms at the sides, looking straight ahead. The figure is carved in from four faces of a rectangular block and retains the block's general shape, and Archaic sculptors reduced anatomy and musculature to decorative patterning across the marble surface. The type is not a portrait of any individual; it is the idea of youth.

This one is over life-size, about 206 centimetres high and 54.6 across, in dolomitic marble. X-ray diffraction gave a composition of roughly 88 per cent dolomite and 12 per cent calcite, and the quarry source was identified as the island of Thasos with about ninety per cent probability.

Dolomitic marble matters to the case in a way ordinary calcitic marble would not, because the whole scientific argument for the statue's antiquity depends on a chemical alteration that only dolomite undergoes.

Seven pieces and a Basel dealer

The sculpture entered the Getty's orbit in September 1983, when the antiquities curator Jiri Frel took it on loan for approval from the dealer Gianfranco Becchina, based in Basel. It arrived in seven pieces.

With it came documents. They traced the statue to the Geneva collection of a Dr Jean Lauffenberger, who was said to have acquired it from a Greek dealer in 1930, and included correspondence from the early 1950s. One letter, dated 1952, purported to be from Ernst Langlotz, then the leading authority on Archaic Greek sculpture, remarking on the statue's resemblance to the kouros from Anavyssos in Athens, National Archaeological Museum 3851. Another, of 1955, referred to repairs carried out by a restorer named as A. E. Bigenwald.

A provenance beginning in 1930 was not incidental. It placed the statue in a private collection before the modern law of cultural property, and it supplied the one thing an unexcavated antiquity cannot supply for itself: a history.

The three dates that destroyed the paperwork

The letters were forgeries, and they were caught by dates rather than by handwriting.

The postcode on the 1952 Langlotz letter did not come into existence until 1972. A bank account referred to in a 1955 letter was not opened until 1963. A document cannot cite something that does not yet exist, and no amount of expertise in Archaic sculpture is required to see it; the check is clerical.

That is the most transferable lesson in this file. Forged provenance is generally not detected by testing the object. It is detected by testing the paper against the calendar: postcodes, telephone formats, letterheads, company registrations, bank accounts, the death dates of the people who supposedly wrote and received things. Whoever assembled this dossier knew a great deal about Greek sculpture and not enough about the Deutsche Bundespost.

The Getty's own later investigation established that the letters constituting the provenance were fabrications. What that establishes about the statue is precisely nothing. A forged provenance proves that somebody wanted the object to have a history it did not have; it does not prove the object is modern. The two questions are separate, and running them together is the commonest error made about this case.

A trustee resigns

Federico Zeri, the Italian art historian and a founding member of the museum's board of trustees, argued that the kouros was a forgery and should not be bought. He was overruled. He left the board in 1984.

The acquisition was approved in January 1985. The reported price differs between careful sources: the Trafficking Culture research project at the University of Glasgow gives 9.5 million dollars, and other accounts give 10 million. This record does not reconcile them; the discrepancy is itself information about how little of this transaction was ever made public.

The statue was first exhibited in October 1986 and publicly announced in 1987 by Marion True, who by then held the antiquities curatorship. Marion True and Jiri Frel are both matters of public record in this chain and appear here only in the roles the published sources give them.

The science that seemed to settle it

The case for antiquity was geological and it was, on its face, powerful.

Stanley Margolis, a geology professor at the University of California, Davis, examined the marble and found that the surface had undergone de-dolomitisation: the magnesium had leached out of the dolomite, leaving a crust of calcite. His conclusion was that this alteration could occur only over the course of many centuries and only under natural conditions.

That is exactly the kind of argument that ends an authenticity dispute, because it is not a judgement about style. It is a claim about a chemical process with a known rate, applied to a surface that a sculptor cannot fake by carving. If it holds, the statue was in the ground or exposed to weather for a very long time and cannot be a modern carving.

It did not hold.

The same crust, made in a laboratory

In the early 1990s Miriam Kastner induced de-dolomitisation artificially in the laboratory. The procedure was not simple, and nobody has claimed that a forger casually did it. But the argument Margolis had made was not that artificial de-dolomitisation is difficult. It was that it is impossible, and that the absence of magnesium is therefore a reliable indicator of age. Once the alteration had been produced by human agency in a laboratory, that inference was gone.

This is the central mechanism of the record and it deserves to be stated in general terms, because it recurs across the whole field. A scientific authentication argument has two halves: the measurement, and the claim that no process other than age could produce the measurement. The measurement is usually robust. The second half is a negative claim about the entire space of possible processes, and negative claims of that kind are only as good as the current state of knowledge. Margolis's measurement was never challenged. What was refuted was the sentence that gave it evidential force.

The consequence is that the geology no longer decides anything in either direction. De-dolomitisation on this statue is consistent with two and a half millennia of burial. It is also consistent with a laboratory. A test that is consistent with both hypotheses is not evidence for either.

The torso in Basel

In 1990 Jeffrey Spier published a study in the Burlington Magazine under the title Blinded by Science. He had found a kouros torso which is a certain forgery and which shows notable technical similarities to the Getty statue. The proposal was that both had come out of a workshop in Rome, and possibly out of the same block of marble.

The Getty bought the torso for study purposes, which is a rational and underused institutional act. Having acquired it, the museum could compare the two directly, and the comparison cut against Spier's reading: the torso is made of identical dolomitic marble, but the sculpting techniques are dramatically different, including the use of power tools that the Getty statue does not show.

So the comparison object established that a modern forger did carve dolomitic marble of this kind, and did not establish that this modern forger carved this statue. Like everything else in the file, it narrows nothing.

Athens, 25 to 27 May 1992

The Getty then did something no museum was obliged to do. It shipped the disputed statue to Greece and convened a colloquium in Athens from 25 to 27 May 1992, at which nineteen invited specialists, scientists and connoisseurs together, examined it and argued.

They did not reach a consensus. The proceedings were published the following year under the editorship of Angeliki Kokkou, which means the disagreement itself is on the record in a citable form rather than being reported second-hand.

A remark attributed to one American museum curator at the conference, quoted in The Economist of 20 June 1992, is the summary the case deserves: the moral is, never ever buy a piece without a provenance.

That is the finding. Not that the statue is fake, and not that the science failed, but that an object with no findspot and no verifiable history cannot be resolved after the fact by any means available, and that the resolution had to happen at the point of purchase or not at all.

Off view since 2018

Following a reinstallation of the Getty Villa, the kouros came off public display in April 2018 and went into storage, viewable by appointment. The museum's director, Timothy Potts, is reported to have said that it was removed because of its disputed authenticity.

Removing it from the galleries is not the same act as deaccessioning it or as issuing a finding. It is a decision about what a gallery of authentic Archaic sculpture should contain, and it can be read either as a quiet concession that the object is modern or as ordinary curatorial hygiene about an unresolved object. The sources consulted here do not record a formal change to the catalogue entry, and the object retains its accession number, 85.AA.40, and the photograph the Getty publishes of it under an open licence.

Later claims about who made it circulate in the literature. Ralph Frammolino and Jason Felch reported in 2011 that a man named Fernando Onore had sold the statue to an intermediary in Calabria, who sold it on to Becchina for one hundred thousand dollars. That is a published account of a claim; no finding has been made, and the same source states that whether Becchina knew the object's status is unknown. It is recorded here because it is in the published record and for no stronger reason.

What this object is actually useful for

It is the best available demonstration that scientific authentication is asymmetric.

Science can exclude. Thermoluminescence excluded a modern date for the Metropolitan Museum's bronze horse; radiocarbon excluded the Renaissance for the Berlin Flora bust. Those are strong results because they are negative: they rule a period out, and a rule-out survives improvements in technique.

Science struggles to confirm. To use a measurement as proof of age, somebody has to assert that nothing but age could produce it, and that assertion is a claim about everything anyone might ever be able to do. Margolis's de-dolomitisation argument was of the confirming kind, and it lasted until one laboratory did the thing it said could not be done.

The second lesson is about sequence. The provenance documents were checked and found false after the purchase. Every one of the disqualifying facts, the postcode and the bank account, was available before it, at the cost of a few enquiries, and none of them required an opinion about Archaic sculpture. The Getty spent around ten million dollars and then thirty years of scientific and scholarly effort on a question that a clerk with a directory could have opened in an afternoon.

And the third is the label. Faced with an object it could not resolve, the museum printed the alternatives on the wall and left them there. That was the right thing to do, and it is the reason this record can be written at all.