Tomb Cover

A complete loom-woven funeral pall, 210 by 102 centimetres, cut to the shape of a body with a circle for the head, its Kufic bands asking forgiveness for a named dead man and labelling which part of the corpse each band covers. Cleveland bought it in 1954 as a masterpiece of tenth-century Buyid weaving, reproduced it in four museum handbooks, and lent it to Paris in 1961, to the Hayward Gallery in 1976 and to a four-city American tour in 1982.

An epigraphic study in 1989 raised such doubts about the whole group of so-called Buyid silks that sixteen were radiocarbon dated by accelerator mass spectrometry. This one came back at 1640 to 1950 with 95 per cent probability. The museum did not quietly withdraw it. It changed the date line on its own public catalogue record to the radiocarbon range and changed the culture line to read in the style of the Buyid period, and it did the same across roughly thirty accessions.

Two of those objects, a tunic bought in 1982 and a caftan bought in 1985, date to after 1950, identifiable because they carry radioactive isotopes from atmospheric nuclear testing. One piece in the group, given to the museum in 1975 and reportedly from the same necropolis, tested at 899 to 1160 and keeps its medieval attribution. That single survivor is why this is a story about evidence rather than about a scandal.

An object made to be buried with a named man

This is not a fragment. It is a complete loom piece, 210.2 by 102.5 centimetres, woven in one operation to a shape: a circle at the top to cover the head, and below it a long tapering form for the body. The ground is dark brown and the only ornament is Kufic writing reserved in beige on bands.

The text is doing two different jobs. Around the circle at the top it names the deceased and asks forgiveness for him: forgiveness from the Merciful, the Compassionate, to the disobedient servant Ali bin Muhammad bin Shahrazad bin Rastawayh al-Isbahani. Further inscriptions down the length identify the part of the body each band lies over. A funeral pall that labels its own anatomy is an unusual object, and the museum's own description says the text was undoubtedly composed for the decoration of a funeral pall.

Hold that thought about how much this object tells you. It will matter.

What the museum's catalogue says today

The tombstone on the Cleveland Museum of Art's public record now reads: Tomb Cover, 1640 to 1950, radiocarbon date, 95 per cent probability. Iraq or Iran, in the style of the Buyid period, 945 to 1055.

Two things are worth pausing on. The first is that a museum has printed a scientific date range in the slot where a date of manufacture normally goes, which is an unusually candid piece of cataloguing. The second is the range itself. Its upper end, 1950, is four years before Cleveland bought the object. On the published evidence there is a real probability that this pall was younger than the curator who acquired it.

How the Buyid silks came to exist as a category

The group to which this object belongs is a large body of complex drawloom silks attributed to the patronage of the Buyid dynasty, which ruled in Iran and Iraq in the tenth and eleventh centuries. In Sheila Blair's summary they have been the subject of a long and vitriolic debate ever since some fragments were uncovered at Rayy in 1925.

The mechanism that followed is documented and simple. The few textiles genuinely excavated at Rayy sold for relatively high prices in the 1930s. Forgers responded by producing more, and often larger or complete, pieces. The market that made the category also filled it. The 1931 Exhibition of Persian Art in London is the event usually credited with making Persian art popular and expensive, and the timing of the supply follows the timing of the demand exactly.

Almost every one of these objects arrives in a museum with the same story: reportedly found at Bibi Shahr Banu, the necropolis at Rayy. Cleveland's own provenance lines say so in those words, hedged in the museum's own voice as originally reported. A findspot repeated by dealers across three decades and documented by nobody is not a provenance. It is a sales line, and this archive treats it as one.

The inscriptions gave them away before the physics did

The order of the investigation is important and is usually told backwards. Radiocarbon did not raise the suspicion. Epigraphy did. A study of the inscriptions carried out in 1989 raised such grave doubts about the authenticity of the group that sixteen textiles were selected for radiocarbon analysis using an accelerator mass spectrometer. The results were published by Sheila S. Blair, Jonathan M. Bloom and Anne E. Wardwell in Ars Orientalis in 1992.

The testing confirmed what the epigraphy had suggested. Only a handful of the textiles proved medieval, mainly but not exclusively those associated with the 1924 and 1925 finds at Rayy and carrying inscriptions that name known historical figures. Most are forgeries produced in the early twentieth century, just after the initial finds.

Blair's characterisation of the worst of them is the sentence that best explains the object on this page: some, including those with very strange iconography and over-informative inscriptions, were made after 1950. An over-informative inscription is one that answers questions a real object would leave open. A pall that names its occupant in full with four generations of patronymic and then labels which band covers which limb is answering a great many questions.

The bomb pulse

The pieces made after 1950 are identifiable by a specific mechanism: they contain radioactive isotopes produced by nuclear explosions and tests. Atmospheric testing from the early 1950s raised the level of carbon-14 in the atmosphere sharply, and anything that grew, or any silkworm that ate leaves that grew, after that point carries the signature. It is a physical timestamp that cannot be forged backwards.

Two objects in Cleveland's own group carry it. A tunic with a lion killing a khilin in octagons, accessioned in 1982, is dated after 1950. A caftan accessioned in 1985 is dated after 1950. In both cases the museum bought a garment that was, at the outside, thirty-five years old and in all probability younger, as a thousand-year-old survival. Neither of those two records is released as CC0, so neither can be shown here; the dates are in the museum's public catalogue.

The one that passed

A record that only lists the failures would be propaganda. So here is the control.

Fragments with winged lions in roundels, given to the museum by Paul and Marguerite Mallon in 1975, accession 1975.45, carry the same reported findspot: Bibi Shahr Banu. Their radiocarbon date is 899 to 1160 at 95 per cent probability. The museum's culture line for that object does not say in the style of. It says Iran, Buyid or Seljuk period, and it is the only object discussed on this page that is still catalogued as what it was bought as.

That result does two things at once. It shows that the method can return a medieval answer, so the negative results are not an artefact of the technique. And it shows that the same dealer network, and the same reported necropolis, handled both real and false material, which is why the group could not be sorted by provenance, by eye, or by reputation. It had to be sorted by measurement.

The catalogue history is the attribution history

What makes this object worth a record rather than a footnote is that its false attribution is fully documented at every stage, in print, by the institution that held it.

It was reproduced in the Cleveland Museum of Art Handbook of 1958 as catalogue number 676, and again in the handbooks of 1966 and 1969 at page 215 and of 1978 at page 272. It travelled to the Petit Palais in Paris for Sept mille ans d'art en Iran, the seven thousand years of art in Iran exhibition, from 19 October 1961 to 31 January 1962. It was number 7, page 76, in Dalu Jones and George Michell's catalogue of The Arts of Islam at the Hayward Gallery in London, 8 April to 4 July 1976, the exhibition that defined the field for a generation of British visitors. It went round the United States in The Heritage of Islam from 10 March 1982 to 5 September 1983, to Houston, Pittsburgh, San Francisco and Washington, and is number 32, page 50, in the notes for that tour.

Then it appears in the 1992 Ars Orientalis paper, reproduced at page 35 and discussed at page 15, and the object's public life changes tense. Louise Mackie's Symbols of Power, the museum's own 2015 survey of Islamic luxury textiles, illustrates it at page 154 as figure 4c and discusses it across pages 154 and 155.

The sequence is the argument. Four handbooks, three international exhibitions, four decades: this object was not obscure and its status was not marginal. It was a headline piece of Islamic textile art in the mid-twentieth century, and the discipline was wrong about it in public for as long as it was on display.

The chain, and why it is so short

The museum's provenance list has three lines. Originally reported as having been found in Bibi Shahr Banu, near Rayy, Iran. Adolph Loewi, 1888 to 1977, Los Angeles, sold to the Cleveland Museum of Art. The Cleveland Museum of Art, 1954 to the present.

A reader used to this archive will notice that only two of those are owners and only one of them is a private one. That is not an accident and it is not a gap in the research. An object manufactured to be sold has no history before the sale, and the shortness of the chain is the single most consistent feature of the whole group: a reported findspot with no excavation record, one dealer, one museum.

The pattern is visible across the accession numbers themselves. Cleveland's Buyid-style silks came in from Marguerite Mallon in Paris and New York, from Nassiri Amini in Tehran, from Heeramaneck Galleries and Rabenou Antiquities in New York, from Spink and Son in London, from Adolph Loewi in Los Angeles, from Galerie Heidi Vollmoeller in Zurich, and from Goumaronart in Paris, over more than forty years from 1941 to 1988. Six countries, a dozen firms, one findspot, and in almost every case no owner named before the dealer.

What it cost, and what the money did

The purchase price has not been published. There was no auction, so no hammer price and no buyer's premium exist for this object, and the only figure that could ever be produced is the museum's invoice from Adolph Loewi in 1954. The acquisition was made from the J. H. Wade Fund, the endowment that also paid for most of Cleveland's other purchases in this group.

The wider money question is answerable even though the specific price is not, and it is the more interesting one. The forgery of these silks was a supply response to a price signal. High prices in the 1930s for the genuine Rayy finds produced a market for more of them; the forgers' documented adaptation was to make pieces that were larger and complete, because larger and complete objects fetch more than fragments. A collector reading this record should take the general lesson rather than the particular one. In a category defined by fragments, completeness is not a bonus. It is a question.

What is unresolved, and what the museum did right

Nobody has published who made this. The forgers of the Rayy silks are not named in the literature consulted here, no workshop has been identified, and the question of whether the twentieth-century weavers were working from genuine excavated fragments or from photographs remains open. Nor is it established whether the pall was ever in the ground at Bibi Shahr Banu at all: a modern textile can be buried in an ancient necropolis, and that would be the efficient way to sell one.

What deserves saying plainly is that the Cleveland Museum of Art has handled this better than most institutions handle a bad attribution. It did not deaccession the object quietly, it did not leave the old date on the label, and it did not bury the finding in a scholarly journal. It printed a radiocarbon range on its own public record, in the date field, on roughly thirty objects, and released photographs of them under CC0 so that anyone can look. This record exists because that decision made it possible.