Why Did the Celts Keep Severed Heads? The Celtic Head Cult and the Evidence

Every surviving written account of Celtic heads being cut off and kept was written by Greeks and Romans looking in from outside. None of it comes from the Celts themselves.

The geographer Strabo, writing in the early Roman imperial period, passes on what he learned from the earlier Greek writer Poseidonius. According to Strabo, the northern peoples hung the heads of their enemies from their horses and nailed them up at the entrances of their homes. The heads of "enemies of high repute" were embalmed in cedar oil and shown to visitors, and their owners would not give them back even for a ransom of their weight in gold. Strabo says Poseidonius saw the display himself in many places. He also calls the custom "barbarous and exotic" and says the Romans put a stop to it (Strabo IV.4.5).

Diodorus Siculus, probably drawing on earlier material that included Poseidonius, tells a similar story. In his Library of History he describes Gauls cutting off the heads of fallen enemies and fastening them to their horses, displaying weapons at their houses, and preserving the heads of distinguished enemies in cedar oil. For Diodorus, the heads were trophies of a warrior's military reputation (Diodorus Siculus, Library of History V.29).

Historians read these passages with care. Strabo and Diodorus generalise across different Gaulish communities, periods and regions, and they wrote with a ready-made "barbarian" stereotype in mind. Their accounts show that the custom was reported. They do not show that every Celtic-language community did it, or that every head carried the same meaning. That is why the archaeology has to be read one site at a time.

The most direct evidence comes from Le Cailar, in the Gard region of southern France. Le Cailar was a fortified settlement, lived in from the 6th century BCE into the Roman period. Excavators uncovered fragments of human skulls together with weapons near the inner walls of the settlement and what may have been a gate area, the kind of place where heads could have been put on show (Ghezal and colleagues).

In 2018 Ghezal and colleagues published a chemical study of eleven of those skull fragments in the Journal of Archaeological Science. Using gas chromatography-mass spectrometry, a technique that separates and identifies the molecules left in a sample, they detected markers of conifer resin and aromatic compounds produced by heating in some of the fragments. The authors concluded that some of the heads had been deliberately embalmed. Their results support what the ancient writers described, though they do not show that every head at the site was treated this way. The study places the practice in Transalpine Gaul in the 3rd century BCE, and France's CNRS summarised the work when it was published (CNRS).

These fragments are the remains of real people from Iron Age Gaul, and they are what turned a line in Strabo into something scientists could test. If you are curious how so much can be learned from a few pieces of skull, our story on what bones tell archaeologists is a good companion.

Entremont stands on a plateau near Aix-en-Provence. It was a fortified settlement established around 175 BCE and lived in through the 2nd century BCE (Entremont, Dating the statuary). Its sculptures include seated heroic figures, carvings of human heads, and stone building pieces with recesses shaped to hold heads.

The French Ministry of Culture's Entremont site sets out two readings of these heads. In one, they are enemy trophies, shown off beside the heroes who won them (Entremont, Human trophies). In another, some earlier portico pieces were reliquary holders for honoured or heroic dead, a place to keep the heads of people the community wished to remember (Entremont, A sanctuary before the agglomeration). A proposed regrouping and display of the sculpture in the enlarged settlement dates to about 150 to 140 BCE.

Entremont also comes with a caution. Many of its sculptures were later broken up and reused as building material, and the site itself says a complete reconstruction is impossible. What survives is a set of pieces, not a finished scene.

Roquepertuse, near Velaux in southern France, was a Celtic or Celto-Ligurian sanctuary known for its sculpture, its carved building blocks and its imagery of human heads. A renewed study by Boissinot and colleagues dates its main phase to the 3rd century BCE (Boissinot and colleagues, Nouvelles recherches sur le "sanctuaire" de Roquepertuse).

One of its best-known pieces is a sculpture of two heads joined back to back, dated to the 3rd century BCE. The site's meaning has shifted as excavation and reassessment went on, so it is best seen as a sanctuary still being understood, rather than as a simple, fully explained "head cult temple".

Carved heads turn up in other Celtic and Celtic-influenced regions too, but a carved head is not automatically a severed head. The National Museum of Ireland describes Ireland's Iron Age stone heads as a varied group of carvings, often with simplified faces, long noses, close-set eyes and stern expressions. It says the proposed "Iron Age head cult" remains a theory, and that the meaning of the motif cannot be boiled down to one explanation (National Museum of Ireland, A Face From The Past).

The best known is the Corleck Head from County Cavan, a stone head with three faces. The museum identifies it as an Iron Age carving, while whether it shows a god, an ancestor or something used in ritual remains uncertain. People sometimes ask whether Ireland has good examples of headhunting; what the museum can show you is heads in stone, and it is careful not to read them as proof of a cult.

Britain has many objects called "Celtic heads", and dating them is often hard. A stone head from Calderdale, recorded by the Portable Antiquities Scheme, is probably Romano-British and probably from the 1st century CE. That record also notes the British Museum archaeologist J. D. Hill's caution that there is no secure evidence for a specifically Iron Age British cult centred on stone heads (Portable Antiquities Scheme record). The British Museum also holds a small copper-alloy fitting in the shape of a human head from a burial at Welwyn, dated about 50 to 20 BCE. It shows that head imagery was part of Late Iron Age Britain, not that it represents a real severed head (British Museum).

Did the Celts worship heads? "Worship" is too big a word for the evidence. What scholars offer instead are three overlapping readings, each tied to particular writers and sites.

The first is power. The Entremont interpretation says that possessing an enemy's head could mean seizing his power for yourself. Heads shown beside heroic figures would then announce military success, aristocratic authority and control over the defeated (Entremont, Human trophies).

The second is honour and reputation. Strabo and Diodorus both describe the heads of distinguished enemies as prized trophies, kept because they stood for the rank of the person defeated and the achievement of the person who defeated him. That reading gives the heads a social and political meaning without needing any doctrine of head worship.

The third is protection, ancestors and continuity. At Entremont, some earlier head-bearing pieces are linked with relic holders for honoured dead and with symbols of fertility or renewal (Entremont, A sanctuary before the agglomeration). The National Museum of Ireland notes that the head motif has been connected with supernatural power, while warning that a single "head cult" across the whole Celtic world cannot be assumed.

That warning is the heart of the disagreement. Some scholars look for one religious system behind all the heads; others see several practices whose meanings changed from community to community and century to century. The evidence clearly supports the cutting off and display of heads at particular sites in southern Gaul. It does not establish that all Celts worshipped severed heads. Celtiberian Spain, has its own evidence and its own scholarly discussion, separate from the Gaulish sites here.

Keeping and displaying heads is not unique to Iron Age Europe, and each culture gave it its own meaning. Compare the skull racks known as tzompantli, the shrunken heads called tsantsa, or the plastered skulls of Jericho, each with a meaning of its own.

Many of these objects are on public display, though collections move and galleries change. The museums and towns describe them this way:

For any single object, confirm the object and the gallery with the museum before travelling. The Musées de Marseille, the Ville de Velaux and the Entremont project all publish details for visitors.

These questions come up often. The short answers follow.

The heads of Iron Age Gaul sit where two kinds of evidence meet: Greek and Roman writers describing a custom they thought strange, and bones and stones that confirm parts of what they said while leaving the meaning open. Some heads were trophies, some were embalmed, some may have honoured the dead, and many carved heads may have meant something else entirely. If the Iron Age people of Europe interest you, the bog bodies of northern Europe are another way to meet the people of that era.

---