The Weary Herakles

A Roman marble of the Weary Herakles type, carved between AD 170 and 192, that spent thirty years in two pieces on two continents. The lower half was excavated at Perge in southern Turkey in 1980 by Jale Inan, the first Turkish woman to lead excavations, who had been hearing rumours that something important had been stolen from the site. The upper half surfaced in the United States the following year and was bought in 1981 by the Museum of Fine Arts, Boston, in a fifty-fifty partnership with the New York collectors Leon Levy and Shelby White, on terms that gave the museum full ownership when Levy died.

The two halves fit. That has been demonstrable since 1992, when a plaster cast of the legs was flown from Antalya to Boston after two earlier attempts had failed. The museum did not accept the conclusion. Its argument was that the Weary Herakles was a common type, that the break might be ancient, and that the top could therefore have left Anatolia long before any Turkish law forbade it. Its head of public relations said in 1994 that the museum did not acknowledge Turkey's claim to ownership and that there had never been any evidence that the statue was stolen.

No case was ever filed. What moved the object was not a court but a change in Turkish policy - from 2010 Turkey withheld loans from institutions holding contested objects, and in May 2011 Germany's Pergamon Museum returned a Hittite sphinx after Turkey threatened the excavation permits of German archaeologists - and the death of a co-owner, which left the museum free to negotiate without a donor in the room. The MFA announced the transfer on 23 September 2011, as part of a broader cultural cooperation agreement. The statue was made whole in Antalya. Jale Inan had died in 2001.

The break is across the hips, and the break is the case

Photographs of the statue as it now stands show a horizontal fault line running across the upper thighs and hips, visible on both legs and through the drapery of the lion skin. Above it is marble that spent thirty years in Boston. Below it is marble that spent the same thirty years in Antalya.

Every other question in this case is downstream of that line. If the break happened in antiquity, the upper half could have travelled at any point in the last two thousand years, including centuries in which no law prohibited it, and Turkey's claim collapses. If the break is recent - made by whoever took the top of a statue out of an excavation site and left the legs behind - then the object was removed from Turkey after Turkish law had vested ownership of antiquities in the state, and it is stolen property.

The two halves fit. That has been demonstrable since 1992. What the fit does not by itself establish is when they were parted, and that gap between a physical fact and a legal conclusion is where the argument lived for nineteen more years.

Perge, 1980: an excavation that found half a statue

Jale Inan had been running the excavations at Perge, in Pamphylia on the south coast of Turkey, since 1967. In 1980 her team uncovered the lower half of a marble Weary Herakles.

She was already uneasy. There were rumours at the dig that something important had been stolen from the site, and suspicion had fallen on a workman. What nobody worked out at the time was that the missing thing and the found thing were the same object, cut in two.

Perge is not a marginal site and Inan was not a marginal archaeologist. She was the first Turkish woman to work in the discipline, trained in Berlin under Gerhart Rodenwaldt, and completed her doctorate in 1943 by carrying her thesis and notes into an air-raid bunker while the city was bombed. She had helped found the chair of classical archaeology at Istanbul University in 1946. The sculpture from her excavations at Perge and Side is the reason the Antalya Museum had to be enlarged twice.

The lower half went into that museum, where it stayed on its own for thirty-one years.

A museum, two collectors, and a gift that had to wait for a death

In 1981 the upper half appeared in the United States, bought from a German dealer. The Museum of Fine Arts in Boston acquired it in a fifty per cent partnership with Leon Levy and Shelby White, the New York collectors whose antiquities collection became one of the most scrutinised in America. The following year it was exhibited in a temporary show at the Metropolitan Museum in New York.

The terms of that partnership matter more than they look. The museum would take full ownership only when Levy died. Until then, any decision about the object required the agreement of living donors who had paid for half of it.

This is a structural feature of American museum collecting that almost never appears in restitution coverage, and it has consequences. A fractional gift ties an institution's hands: the museum cannot deaccession, transfer or surrender an object it only half owns, and any discussion about doing so is a conversation with a benefactor about their money and their judgement. When Levy died in 2003 and the museum held the object outright by 2004, it reopened discussions with Turkey. Reporting at the time made the connection explicit - the negotiation was easier once a donor no longer had to be in the room.

Nineteen of the thirty years this statue spent apart were years in which the museum could not have returned it alone even if it had wanted to.

Three casts, and the one that fitted

Turkey's evidence was never a document, because there is no document. It was a shape.

Three separate attempts were made to prove that the two pieces were parts of one statue using plaster casts. The first two were unproductive. In 1992 a cast of the lower half was transported from Antalya to Boston, and this one satisfied those who saw it - Turkish officials, scholars, and Thomas Hoving, the former director of the Metropolitan Museum - that the halves were from the same object.

A join is the strongest kind of provenance evidence there is, and it is worth being clear why. It is not stylistic, so connoisseurship cannot argue with it. It is not testimonial, so nobody's credibility is in issue. Two broken surfaces either interlock at every point or they do not, and the probability that two independently broken marbles of the same type match along an irregular fracture is negligible. The Euphronios krater case turned on matching Polaroids; the Cimabue diptych reconstruction turned on worm tunnels bored through joined panels. This is the same class of evidence and it is the best available in the field.

A 3D modelling study of the statue was presented at the Turkish-German Geodetic Days in Berlin in 2006, digitising the halves rather than casting them. By then the physical question had been closed for fourteen years.

The museum's answer: the type is common and the break may be ancient

The Museum of Fine Arts did not concede. Its arguments were of two kinds and both are worth stating precisely, because they are the arguments any holding institution will make.

The first was about abundance. Cornelius C. Vermeule III, the museum's department head, pointed out that Weary Herakles statues were made in quantity in the Roman period, so that fragments of many different examples exist. If there are dozens of these torsos, the fact that one of them resembles the top of a Turkish pair proves nothing on its own. As a general proposition this is true, and it is the reason the cast test was necessary.

The second was about time. In 1994 the museum's director of public relations, Robert P. Mitchell, told the Los Angeles Times that the museum did not acknowledge Turkey's claim to ownership, that there had never been any evidence that the statue was stolen and that allegations to that effect were entirely unsupported. He went on: the break between the top and bottom halves of the statue appears to be an ancient one, such that the top half could have been removed long ago from the territory that is now known as Turkey. He added that if settlement discussions failed, a lawsuit would be necessary, and that such a case would raise significant questions.

The museum's public line in 2011 was still that the piece could have been found at any time since the Italian Renaissance.

None of this was ever tested. Turkey never sued. The proposition that the break was ancient was never subjected to examination by an independent body, and it was never withdrawn either.

What changed was policy, not proof

The evidence in 2011 was the evidence of 1992. What had changed was the cost of saying no.

From 2010 Turkey took a harder line, refusing loans to institutions holding objects it considered contested. The Art Newspaper reported in early 2012 that two British museums had recently been denied loans. In May 2011, four months before the Boston agreement, the Pergamon Museum in Berlin agreed to return a Hittite sphinx from Bogazkoy after the Turkish culture minister threatened to bar German archaeologists from excavating in the country. Ertugrul Gunay, the minister, said Turkey would fight in the same way for all its other artefacts.

These are the two levers a source country actually holds over a universal museum, and neither is a court. The first is loans: a museum that cannot borrow cannot mount exhibitions, and a country with as much material as Turkey is a lender nobody can afford to lose. The second is excavation permits, which are what keep foreign university departments and their funding alive.

Turkey did not win this object by proving anything new. It made the refusal expensive.

September 2011: an agreement, and what it is not

On 23 September 2011 the Museum of Fine Arts announced that it and the Turkish Republic had reached an agreement for the transfer of the top half of the Weary Herakles to Turkey. It formed part of a wider cultural cooperation agreement, and it was reported at the time that the reunited statue would come back to Boston on a short-term loan.

Six distinctions matter here, and coverage of restitution routinely collapses them.

This was not a court judgment: no case was filed, in any jurisdiction, at any point in nineteen years. It was not a forfeiture: no state seized anything. It was not a settlement of litigation, because there was no litigation to settle. It was not a criminal proceeding and nobody was charged. It was not an admission that the museum had bought stolen property, and the museum did not make one; asked why it had moved, its deputy director Katherine Getchell said it was only in the last couple of years that Turkey had presented photographs and other evidence of looting from the site. And it was not unilateral generosity, because the museum received something - the cooperation agreement, and the prospect of loans from a country that had stopped lending.

It was a negotiated transfer of title by a private institution, made when the balance of advantage changed. That is a real and honourable outcome, and describing it as an admission would misstate what the museum said.

The woman who found the legs died ten years too early

Jale Inan died in Istanbul on 26 February 2001, aged eighty-seven. The upper half of the statue she had spent two decades pursuing came home ten years and seven months later.

Her work on this statue is described as having formed the basis on which the return was argued, and the description is fair: she found the lower half, she connected the rumour of a theft to the object, and she drove the casting tests. But the Herakles is not her largest contribution to this subject.

In 1967, the year she took charge of Perge, a group of Roman bronzes appeared on the American market with a murky provenance and vague claims of a south-west Anatolian origin. She began in 1973 to try to establish where they had actually come from. In 1990 she excavated at Bubon and identified a Sebasteion, an imperial cult temple, as the building those bronzes had stood in, publishing the case in 1994. That identification is the foundation of the seizures of Bubon bronzes from American museums that are still going on three decades later.

The pattern of her career is the inverse of the usual restitution story. Ordinarily a site is known and an object appears in a museum. In the Bubon case the objects were known and the site was not, and she went and found it.

A museum that displayed its own dispute

For thirty-one years the Antalya Museum showed the lower half of the statue on its own. That is an unusual object to exhibit: a pair of marble legs, cut off at the hip.

The museum did not disguise it. The legs were displayed in recent years alongside a photograph of the upper half in Boston, and next to a collage of newspaper cuttings describing the dispute with the Museum of Fine Arts.

Whatever else that is, it is a piece of museology worth noticing. Most institutions put a tombstone label under an object and leave the argument out of the gallery. Antalya turned the absence into the exhibit, and made every visitor a witness to a claim. The best evidence of the effect is that the display is remembered and described by American reporters who came to write about the Boston end of the story.

The type, and why abundance was a defence

The Weary Herakles shows the hero at rest, leaning on his club with the lion skin draped over it, exhausted after the labours. The composition descends from a bronze of about 320 BC attributed to Lysippos of Sikyon, court sculptor to Alexander, and its most famous descendant is the colossal Farnese Hercules signed by Glykon of Athens, found in the Baths of Caracalla and now in Naples.

Roman patrons bought this figure repeatedly for baths, gymnasia and palaestrae, where a resting Herakles is a joke and a compliment at once: the strongest man who ever lived, worn out, standing where men come to exercise. The Perge statue is one of a large family, made between AD 170 and 192.

That ordinariness is the point at which art history entered the legal argument. Because the type is common, the museum could say that a torso in Boston and legs in Antalya were not necessarily one object - and it was right that the resemblance alone proved nothing. Rarity would have made the claim easy. Abundance made a physical test necessary, and it took twelve years and three casts to run it.

Where the statue is in August 2026

The statue is not on public view, and a record that stops in 2011 would tell a reader the opposite.

The Antalya Museum closed in July 2025 and its holdings were moved into temporary storage on the site. A seismic analysis carried out in 2020 was cited as showing that the 1970s building was at high risk in an earthquake; the replacement is planned at around twenty thousand square metres and a cost of about sixty-five million dollars, tripling storage and modernising conservation. Demolition began in September 2025 over the objections of academics and professional bodies who wanted the building reinforced rather than demolished, noted that it was the first Turkish museum to be built from an architectural competition, held in 1964, and asked for the seismic report to be published. As at May 2026 the site was still under reconstruction.

So the object that took thirty-one years to be made whole has been in a crate for most of the last year. Nothing in the sources consulted says where, and this record does not guess.