Along with an apparent selection of high ranking individuals (such as warriors and richly ornamented women) for burial in the communal cemetery, the presence of graves in silos beyond the cemetery lends weight to the idea of a mortuary hierarchy. Discriminatory post-mortem treatment awaited certain age groups such as the newborn, lower ranked individuals, criminals, certain warriors, and there were perhaps even human sacrifices.

These silo structures, long called "exclusion-graves", attest to the setting apart of certain members of the population who were buried in underground grain silos. They are evidence for marginalising practices and post-mortem distinctions that seem to have existed throughout most of Protohistory, but especially during the Early and Middle La Tène periods. Hypotheses concerning "cemetery banishment" have long emphasised the alleged status of the excluded, seen as criminals, sacrificial victims or slaves.
Over the past ten years, detailed and broader study of these structures has led to a reconsideration of this notion of social exclusion. Among the great diversity of cases, the only common denominator is the placing of one or several bodies in a silo. This practice seems to incarnate the relationship between the world of the living (the preservation of grain) and that of the dead, represented by the decomposing corpse.

Individual burials are the most frequent. The apparent disorder of skeletal positions (many lie on their stomachs) suggests a pragmatic reuse of pits that had been emptied of their contents and then rapidly refilled. Collective graves do exist however, such as that of three female subjects simultaneously buried at the site of "Le Grand Marais" in Varennes-sur-Seine, as well as collective deposits that generally contain two bodies interred consectutively and separated by a layer of soil.

The anthropological study of these skeletons has revealed a dominance of female individuals (mainly mature or old subjects), though there are also a few males and young children, and a very small number of infants.

Pushing bones aside
This treatment of the body is rare in the silos and completely absent in the cemeteries. At Varennes-sur-Seine, a jumble of dry bones was found piled against the curved silo side, while a few remaining bones in the centre indicated the original location of the deposit.

Desiccation
One of the bodies found at Varennes-sur-Seine was subject to numerous manipulations suggesting that it may have been passively dried. The conservation of certain flexions as well as forced deposit in a lateral flexed position indicate an intitial strongly bent posture. This was accentuated by a forcing of the joints and the inversion of the pre-existing flexions. This body could have undergone this treatment in a flexed or seated position. Similar preparation has also been observed at Acy-Romance where a natural drying associated the maintenance of easily displaced articulations and the dislocation of persistent connections through the drainage of fluids associated with decomposition.

Beheading and human sacrifice
Human sacrifice has also been proposed, in particular for certain subjects that seem to have been thrown down from the top of the silo or "decapitated". However, there is no tangible evidence (such as an impact on the cervical vertebrae) to support these hypotheses. The displacement of some skulls seems only to add to the complex inventory of post-mortem manipulations, and if "sacrifices" did occur, they did not have osteological consequences. However, we cannot exclude the possibility of "unreadable" causes of death, such as suffocation, drowning or strangulation.

These individuals were not stripped of their clothing and ornaments before being buried and cloth wrappings are often attested. Even certain body ornaments and clothing accessories remained in place, such as the bronze clasps, rings and bracelets of the woman from the site of "Le Grand Canton" in Marolles-sur-Seine. It thus becomes problematic to consider these clothed and sometimes adorned individuals exclusively as members of lower ranked groups. However, weapons are absent, when they are common in the neighbouring cemeteries. But before interpreting this absence in terms of social exclusion, we should recall the demographic structure of the interfluvial silos, in which two thirds of the individuals are female.

Many of the skeletons are incomplete and were manipulated after their initial burial. While some seem to have been rapidly dispersed, the excavation of many others suggests a methodical competence for this type of intrusive act. Though the skull was the element most often shuffled about, other bones, such as long bones, were also displaced.
While it is easy to see what is missing from an in situ skeleton, it is more difficult to evaluate the weight of dry bones arranged and displayed outside the cemetery. We find such deposits in rural buildings, cultural and collective sites, pits, shafts, trenches, sanctuaries, and even cemeteries! But most often, these prepared or retouched bones are found in secondary positions after the original structure has been dismantled. They are displayed as trophies or in shrines and "skull-boxes", for example, and the original arrangement can no longer be observed.
The presence of preparation traces is not systematic. When they do exist they are similar to those on the exposed skulls removed from fresh corpses described in antique texts. These skulls can probably be attributed to warriors because they are often found with weapon fragments. It is therefore possible that they represent the display of trophies associating the skull of a warrior with his weapons.

But before interpreting these as bloody or "barbaric" acts, or as evidence of human sacrifices, we must cautiously consider the simple association, in an underground space devoted to preservation, of a corpse and a harvest. A deceased human is the best mediator between the gods and the people (ancestor cults, "heroisation"). In a silo, the deceased perhaps becomes an intermediary and/or direct protector of harvests. This post-mortem association, approved by the group and whose criteria for selection are poorly known, would then indicate the existence of a true consecration ritual. In these pits, associated human and animal offerings, gradually emptied of their internal liquids, would then become ultimate libations offered to an underground god. They could be associated with the bending of objects such as swords and iron spear heads, followed by their massive corrosion, representing a symbolic decaying of weapons.
The sacrificial acts converge, establishing an intentional and controlled degradation of the offering, followed by differential manipulations: breakage of amphorae, displacement of animal carcasses, the damaging of weapons, etc. The human body, intentionally dried or naturally decomposed, can in turn be the object of codified manipulations.
Well before the establishment of collective cult spaces, animals, weapons and humans, associated with prestigious goods, were joined in a glorification of decomposition, thus creating a clash between the funerary and the domestic, as well as the religious and the profane. Graves are scattered rather than in a group, corpses are put on show and then their dry bones join the everyday world as surprising, macabre neighbours.