Amulet Atlas

storiesWho owns a charm

also: repatriation · returning amulets · whose object is it

storieEverywhereNorth AmericaWest AfricaAustraliaAotearoa New ZealandBritain and IrelandUnited States of AmericaUnited KingdomNigeriaAustraliaNew ZealandIrelandSwitzerlandIndia

A museum drawer of amulets is the hardest kind of collection to hand back and the easiest kind to have taken. The objects are small, they were bought by the handful, most were never accessioned with a name attached, and a proportion of them were never anybody's to sell. This record sets out the law where there is law, the returns and the refusals where they are on the record with dates and institutions attached, and the specific reason amulets sit at the difficult end of the problem.

cited

When

1976, the Iowa Burials Protection Act, to now still carried cited

Period1976 to the present
How the date is knowntextual — a dated text describes it
NoteDated from the first US statute specifically protecting Native American remains, not from the first claim; claims are older than any of this legislation.

Years are stored astronomically — 1 CE is 1, 1 BCE is 0, 500 BCE is −499 — and printed the way a reader says them. How these dates are made.

One country made it a duty

The Native American Graves Protection and Repatriation Act, Pub. L. 101–601, was enacted in the United States on 16 November 1990. It requires federal agencies and institutions receiving federal funding to return Native American cultural items — human remains, funerary objects, sacred objects, and objects of cultural patrimony — to lineal descendants and to culturally affiliated American Indian tribes, Alaska Native villages and Native Hawaiian organisations. It makes trafficking in those items a criminal offence. In 2022 Congress raised the penalty for a first conviction for trafficking human remains from twelve months to a year and a day, making it a felony, and raised a subsequent conviction from five years to ten.

The counts since 1990 are published: the remains of approximately 32,000 individuals returned, nearly 670,000 funerary objects, 120,000 unassociated funerary objects, and 3,500 sacred objects. For scale, in 1990 federal agencies reported holding the remains of 14,500 deceased Native people.

On 12 January 2024 revised Interior Department regulations took effect. Their operative sentence is short: "...museums and Federal agencies must defer to the Native American traditional knowledge of lineal descendants, Indian Tribes, and Native Hawaiian organizations." That is a rule about whose account of an object counts, which is the question underneath every dispute in this record.

NAGPRA was enacted primarily at the insistence and by the direction of members of Native American nations, and one person is usually named first. Maria Pearson has been called "the Founding Mother of modern Indian repatriation movement" and the "Rosa Parks of NAGPRA". In the early 1970s her husband, an engineer with the Iowa Department of Transportation, told her that road construction at Glenwood, Iowa, had uncovered both white and Native burials; the twenty-six white burials were quickly reburied and the remains of a Native American mother and child were sent to a laboratory instead. Pearson sat outside Governor Robert D. Ray's office in traditional attire until he saw her. Asked what he could do for her, she said: "You can give me back my people's bones and you can quit digging them up." The Iowa Burials Protection Act of 1976 followed — the first legislation in the United States specifically protecting Native American remains.

The law did not arrive without opposition and has not ended the argument. A number of archaeologists have objected that they are prevented from studying ancient remains that cannot be traced to a historic tribe, and investigations have stalled over it, the Spirit Cave mummy among them. Before 1990 the courts had already shown what no law looked like: in Wana the Bear v. Community Construction (1982) two hundred Miwok ancestral remains were bulldozed for a residential development in Stockton, California, and the California Courts of Appeal sided with the construction company.

The case everyone knows

The Benin Bronzes, taken from the royal court of Benin in what is now Nigeria, are the reference case because both sides have stated positions in public.

The British Museum has been claimed to be the largest receiver of "stolen goods" in the world and has consistently refused to return objects, citing the British Museum Act 1963 as preventing restitution. The directors of the British Museum and of the Victoria and Albert Museum, and the British Minister of Culture, have spoken against permanent restitutions; like several European institutions they prefer cooperation and a "circulation of objects" through temporary exhibitions in Africa. The British Museum and the Pitt Rivers Museum both joined the Benin Dialogue Group, set up to coordinate scientific exchange, assistance for a new museum in Benin City, and eventual restitutions.

The first British institution to return a statue after proof that it was looted directly from the royal court was Jesus College, Cambridge, following a campaign by the college's Legacy of Slavery Working Party; the handover to Nigerian delegates was announced for 27 October 2021. In April 2021 the National Museum of Ireland announced that twenty-one Benin Bronzes would be returned to Africa; its head of collections, Audrey Whitty, said the museum group would investigate holdings with a view to repatriating those viewed as "plunder". Fintan O'Toole had written in 2013 that much of that ethnographic collection "lies in the grey area between trade and coercive acquisition: an expansive terrain in imperial relations", while other objects were unambiguously loot. In June 2026 eighteen artefacts from the Kingdom of Benin went back to Nigeria from three Swiss museums — fourteen from the Ethnographic Museum at the University of Zurich, two from Museum Rietberg Zurich and two from the Musée d'Ethnographie de Genève — out of roughly a hundred items in Swiss collections assumed to have come from Benin.

The intellectual frame most of this argument now runs on is the 2018 Report on the Restitution of African Cultural Heritage by Felwine Sarr and Bénédicte Savoy. Since its publication the debates have produced restitutions on moral rather than only legal grounds, which is a shift in what counts as a reason.

Why an amulet is the hard case

Inference — a bronze plaque from a named royal court has three things an amulet usually lacks: it is singular, it is documented, and the taking is dated. An amulet is the opposite on all three counts, and the difficulties compound.

They were collected by volume. The Wellcome Collection holds 17,500 magic-medical amulets, talismans and charms picked up by Henry Wellcome in Islamic North Africa and elsewhere. A claim over one object in that number has to begin by identifying which object.

They arrived without paperwork. In 2021 the Wellcome Collection acknowledged that Wellcome's collecting was shaped by colonialism, that many objects bought at auction in the late nineteenth and early twentieth century carry no documentation of how they were acquired, and that they may have been "unjustly taken" from communities across Europe, Asia, Africa and the Middle East. The institution has committed to working with communities on provenance, and in 2025 returned a group of Jain manuscripts to India after researchers found they had been acquired without the consent of the communities they belonged to.

And a large share of them are restricted in their own tradition, which means the claim and the catalogue entry cannot both be public. Australia's International Repatriation Program is written for exactly that overlap: it "supports the repatriation of ancestral remains and secret sacred objects to their communities of origin to help promote healing and reconciliation". More than 100,000 Indigenous Australian objects are held in over 220 institutions worldwide, at least 32,000 of them in British institutions. Australia has no law directly governing repatriation. The museum anthropologist Jason Gibson notes that there is a lack of Aboriginal authority surrounding collections, and that protocols have instead been made by non-Indigenous professionals.

That combination — many, unlabelled, and not ours to describe — is why this atlas is thinner on some of the richest amulet traditions in the world than on Egypt, where the objects were dug up with field numbers.

Returns, with dates and institutions

The record is not only refusals.

In 2006 Tasmanian human remains — ash bundles and skeletal fragments — were returned from the British Museum to the Tasmanian Aboriginal community, following requests by the Tasmanian Aboriginal Centre with Australian government support, on the understanding they would be cremated in accordance with Aboriginal beliefs.

In 2008 several institutions including the British Museum repatriated Māori human remains to Te Papa Tongarewa in New Zealand, among them toi moko, tattooed heads, and bone fragments.

AIATSIS began its Return of Cultural Heritage programme in 2019. In late October that year the Illinois State Museum returned forty-two Aranda (Arrernte) and Bardi Jawi objects removed from central Australia in 1920 — the first group. In November 2019 Manchester Museum returned objects including sacred artefacts collected 125 years earlier from the Nyamal people of the Pilbara; it returned nineteen sacred objects to the Arrernte people during the pandemic, celebrated in May 2021, and 174 artefacts in September 2023. AIATSIS project leader Christopher Simpson said they hoped the project could become an ongoing programme. UNESCO's director for culture and emergencies said: "May this occasion be a source of inspiration, encouraging others to embark on similar journeys."

In 2022 the Pitt Rivers Museum and the Oxford University Museum of Natural History returned the remains of seventeen Aboriginal Australians to the Australian government. In 2026 the Pitt Rivers agreed to repatriate the skeletal remains of a number of Naga people, some of which had been displayed in a case titled Treatment of Dead Enemies.

The Pitt Rivers has also changed what it shows. During its 2020 closure it removed its tsantsa — shrunken heads displayed since the 1940s — and its human remains generally. Its director's statement: "Exhibiting Tsantsas (shrunken heads) reinforced racist and stereotypical thinking that goes against the museum's core values." It said it would rewrite labels to carry the voices of artists and Indigenous leaders, and began meeting originating communities to address errors and gaps and to discuss repatriation.

Where this atlas stands

Inference — this project holds no objects and can return nothing, so its only decision is what to publish. Three consequences follow and they are stated as descriptions, not undertakings.

The `holdings` field on a record names a museum and cites a published catalogue page. An accession number with no catalogue URL is rejected by the validator, because a number a reader cannot check is a number a reader cannot use in a claim.

Where a museum has published a position on its own provenance — the Wellcome's 2021 statement, the Pitt Rivers' 2020 one, the British Museum's reliance on the 1963 Act — the record carries the position and not a summary of it.

And where a tradition says a class of object should not be shown or described, the description is absent and the restriction is printed instead. That rule has its own record.

Gweagal man Rodney Kelly and others have been working for the return of the Gweagal Shield and Spears from the British Museum and from the Museum of Archaeology and Anthropology at Cambridge. That claim is live as this is written, and no verdict belongs here.

What this is near

orgThe Native American Graves Protection and Repatriation Actthe statute this read opens with, and the only one in the set that makes return a legal dutyorgAustralian Institute of Aboriginal and Torres Strait Islander Studiesthe body running Australia's Return of Cultural Heritage programme since 2019orgThe British Museumthe institution that refuses most often, and publishes the Act it refuses underorgThe Pitt Rivers Museumthe museum that removed its human remains in 2020 and published the reasoningorgMuseum of New Zealand Te Papa Tongarewawhere the toi moko went back to in 2008, and where the terms are set by the originating peopleorgWellcome Collection17,500 amulets, an acknowledgement in 2021, and a return to India in 2025orgMusée du quai Branly — Jacques Chiracthe French collection at the centre of the restitution argument Sarr and Savoy opened in 2018parallelNkisia class collected in quantity out of Central Africa, and one where the object's own community defines what it isparallelHei-tikia taonga whose return is governed by the tradition's own terms, not by a museum'sartThe Lovett collectionthe counter-example — a collector who wrote down who was carrying each charmstoryWhat this atlas does not printthe companion read: what may be said, once you have settled what may be kept

What names this one

Each in its own words, from its own page.

storyAnkh amuletthe long read this sign lands in — worn now by Kemetists, by goths and by people with no line to Egypt at allstoryCimarutathe long read the Stregheria argument belongs in: who a folk object belongs to once a revival claims itstoryCippus of Horusthe long read it lands in: a pasha's gift to a prince, a castle in Bohemia, and a New York accession in 1950storyCoptic cross tattoothe long read this mark belongs to — the tattooists call it resistance and solidarity, which is a statement about who may wear itstoryCorn dollythe long read the 1950s revival belongs in: new folklore attached to an old craft, and the source saying so outrightstoryCowrie strandthe long read this shell complicates: 30 billion of them reached the Bight of Benin as trade goods in the era of Atlantic slaverystoryCrystal pendantthe long read where a supply chain and a spiritual claim end up in the same paragraphstoryDol hareubangthe long read this belongs in: 47 or 48 originals, an island's worth of replicas, and three of them given away as state giftsstoryDreamcatcherthe long read this record is the test case for — who may make a thing, and what changes when somebody else doesstoryHei-tikithe long read this record is central tostoryHex signthe long read for the rest of it: neopagans making them, the Amish refusing them, and a state radiation office using one as a logostoryIllathe long read this belongs in — nearly every illa in a museum left a corral to get therestoryIncantation bowlthe long read this corpus forces: 2,000 bowls excavated, tens of thousands dug and soldstoryKeris pusakathe long read the Banyu Sumurup workshop belongs in: one bench, one blade for an enthusiast's shrine and one for a visitor's shelfstoryMagical gemthe long read; most of these gems reached museums through the trade, and their findspots went with the diggersstoryMjölnir pendantthe long read this pendant forces: a sign claimed at once by Heathen practitioners, by a national cemetery service and by white supremacistsstoryNkisithe long read this object is the centre of: purchased, confiscated or removed by colonial authorities, and argued over by name since 2018storyRabbit's footthe long read it belongs in: a Black folk-magic object sold as generic American luck, with the lineage left off the packetstoryRed stringthe long read; a custom with no textual basis, sold worldwide under the name of a restricted traditionstorySak yantthe long read the 2011 ministry move belongs in, with both sides namedstorySenninbarithe long read this belongs in: a belt made by a thousand named-nobody women, now bought and sold as militariastoryBakongo nkisi craftthe long read this tradition sits at the centre of: purchased, confiscated or removed, and argued over by name since 2018storyTaonga Māorithe long read this tradition is the sharpest case in — a resin copy made offshore, a statute with a $100,000 fine, and a collection whose sources are on the recordstoryThe swastikathe long read on who gets to say what a sign means, which this record is the hardest case forstoryBone and ivorythe long read for objects whose trade is now illegal and whose display is argued aboutstoryCowriethe long read for a material whose supply history is the Atlantic slave tradestoryGoldthe long read on value, ownership and what happens to a charm worth meltingstoryJadethe long read this stone's 1997 settlement belongs tostorySkinthe long read the 2011 Thai argument belongs tostoryFired claythe long read this record's last paragraph belongs to: bowls dug without a findspot, and who ends up holding themstoryTooth, tusk and clawthe long read for an object under export caps, permits and a 2007 repatriationstoryWoodthe long read behind the Fante dolls that ended up in European museums after the 1800sstoryFeeding the charmthe long read this practice complicates: a figure that is no longer fed is, in its own tradition's account, not the object on the shelfstoryGrave depositthe long read this practice leads straight intostoryInheritingthe long read that starts where an heirloom leaves the familystoryAkodessawa fetish marketthe long read this market sits inside: who gets to describe a practice, and in whose wordsstoryRoom 1: Enlightenment, the British Museumthe long read on ownership; Wall Press 96 is where it stops being abstractstoryThe amulet and charm cases, Pitt Rivers Museumthe long read; this museum answers it out loud, on its own labels, since 2020storyQuiapo Churchthe long read this address argues in public: two systems, one plaza, both naming the otherstoryMana Whenua, Te Papathe long read on ownership, argued from the claimant's side — 850 ancestors returned since 2003storyThe Lovett charms at Wellcome Collectionthe long read; a charm swapped face to face on a barrow is the awkward case in itstoryAugustus Pitt Riversthe long read; his museum is where the question of who a charm in a case belongs to is asked most directlystoryE. A. Wallis Budgethe long read on acquisition; his own account of getting tablets out of Baghdad is one of its documentsstoryWalter Leo Hildburghthe long read on how a charm changes hands; a Hildburgh object has usually changed them four timesstoryAustralian Institute of Aboriginal and Torres Strait Islander Studiesthe long read on ownership; the deposit clause here is the clearest model of a collection held under somebody else's conditionsstoryThe British Museumthe long read on ownership; the 1963 Act argument is its clearest statement of one sidestoryThe Metropolitan Museum of Artthe long read on ownership, where seizures and returns since 2006 are the evidencestoryThe Native American Graves Protection and Repatriation Actthe long read on ownership, of which this is the only piece of binding law in the setstoryThe Pitt Rivers Museumthe long read on ownership; this museum's own labels since 2020 are part of the argumentstoryMusée du quai Branly — Jacques Chiracthe long read on ownership; the Sarr–Savoy report of 2018 is its clearest recent documentstoryMuseum of New Zealand Te Papa Tongarewathe long read on ownership, argued here from the claimant's side rather than the holder'sstoryUNESCO Intangible Cultural Heritagethe long read on ownership, where inscription is the thing most often mistaken for a property claimstoryWikimedia Commonsthe long read on ownership; a free licence on a photograph settles nothing about the object in itstoryfetishthe long read this record feeds: who gets to name a thing, and what the naming doesstorymascotthe long read the Native American mascot argument belongs tostorymojothe long read on words and objects that leave the community that built themstoryVodou and Vodún — and the spelling this atlas does not usethe long read on naming: who gets to spell somebody else's religion, and what the spelling then carriesstoryBudge's amulet platesthe long read that starts from how objects like these reached LondonstoryThe Lovett collectionthe long read about how collections were formed, and what the Wellcome has since said about its ownstoryFree to usethe same question about objects instead of about text: what may be kept, and by whomstoryThe fake marketwhat happens after an object with no provenance reaches a museumstoryThe swastika problemthe same question one step over: who gets to decide what a sign or an object means nowstoryWhat this atlas does not printthe companion read — this one is about what may be said, that one about what may be kept

Where this came from

Elsewhere

Record updated 2026-09-16. This record as JSON. Every field carries its own provenance; the tier chips above say which.

Pass it on