Athough the faunal analysis for the project is not yet complete, we must say something about the extraordinary number and kinds of dog bones at the KS settlement. All numbers cited here are preliminary and could change.

Dog bones account for 40% (NISP) of the bones in the excavated structure at the KS settlement. No other Srubnaya settlement anywhere has yielded more than 2% dog bones. At least 18 dogs were butchered (MNI), and probably many more. They occur more or less throughout the structure and in all levels, but seem to have been concentrated near the floor of the structure (levels 6 and 7) in an arc south and east of the probable Srubnaya-period well, Pit 14. (This observation requires confirmation from ongoing intra-site distribution analysis.) The photographs below show how these bone scatters appeared in Units L3 and L4, north and east of Pit 14. L4, on the left, is shown with the next level half-excavated on the right. A dark posthole can be seen in the floor of L3 on the right.

The dog bones recovered from the pits in the floor of the structure show skinning and dismemberment marks, and were broken (for marrow?) like those of cattle and sheep. The dog heads, which are over-represented compared to other body parts, were roasted or burned and then carefully chopped into 10-12 small, neat, almost identical segments with axe blows. One standard cut was down the middle of forehead at the intersection of the frontal bones. The palate was chopped longitudinally into thirds. A standard transverse cut was from the orbit to the middle of the tooth row. The back of the skull is also chopped into neat segments, and the mandible chopped in half. This complicated series of axe cuts was repeated in the same way for each dog head. It required practice and skill. Since it had no functional purpose, and was used only on the dog heads, not cattle or sheep, it almost certainly was a ritual act. The pattern of burning, chiefly on the teeth and alveoli, shows that the chopping happened after roasting/burning, since the chopped surfaces do not exhibit burning.

Some of the cattle heads also were chopped, but only to open the muzzle, a functional part of normal butchering (particularly to obtain the tongue). Cattle heads were not chopped into small pieces of equal size. The careful segmenting of skulls into pieces of 3-4cm size is unique to dogs. There is very little meat on most of these pieces. This suggests that it was necessary to segment not so much the meat as the substance of the dogs, especially their heads. The postcranial remains were not chopped into such tiny pieces. There was something special about the dogs' heads. We suggest below that the ritualized chopping of the heads might have been conducted to de-sacralize dog skins and carcasses that had been used in a ceremony that was itself conducted elsewhere, perhaps in the house now on the bottom of the lake.

In the photograph below, the standardized nature of these cuts can be seen. On the left specimen, the upper teeth can be seen still in the bone, and the lower curve of the right eye socket can be seen just above the upper teeth. The other two pieces are cut in exactly the same places, but are from the left sides of two other dog skulls.

Anne Pike-Tay and her students at Vassar College analyzed a sample of the animal teeth from KS and from PD1 to detect indications of the season of death from incremental banding in the cementum of the tooth roots. Her comparison of the seasons of dog deaths with the cattle and sheep/goat deaths revealed that while the cattle and sheep/goat were slaughtered year-round, winter and summer (see Seasonality) , 93.3 % of the dogs were killed in the winter. The dog deaths cluster in two seasonal groups, one probably around the winter solstice (winter 1) and another in mid-February to mid-March (winter 3). At least one cow was killed during each winter-season dog killing event (both winter 1 and 3). If the cattle and dogs were part of the same event, it suggests both a large crowd (many portions) and a complex ritual marked by dog and cattle sacrifices.

Several comparative Indo-European mythologists (Kershaw 2000; Falk1986:37-56; White 1991:95-100; Veesterman 1962) have reconstructed an ancient Indo-European mid-winter ritual associated with the beginning of the new year at the winter solstice, when boys were initiated into manhood and became warriors. The symbols of the ritual were dogs and/or wolves, accompanied by cow sacrifices. No archaeological evidence of such a ceremony has been reported before now. The Krasnosamarskoe settlement seems to contain the first reported archaeological evidence for this ritual.

In Kershaw’s (2000) reconstruction the principal ceremony occurred at the winter solstice, but there was also a second phase of the ceremony later in the year; at the KS settlement one cluster of dog deaths can be correlated with the solstice and another two or three months later. According to Kershaw the ceremony took place near a cemetery; the KS settlement was unusually close to a Srubnaya cemetery. One of the functions of the reconstructed Proto-Indo-European mid-winter ritual was to initiate young men into the warrior category (Männerbünde, Korios), and the principal symbol of this initiation was the dog. The dog was the symbol of death—multiple dogs or a multi-headed dog (Cerberus, Saranyu) guarded the entrance to the Afterworld. At the mid-winter initiation ceremony young men witnessed the death of both the old year and their old identities, while as warriors they would become men who fed the dogs of death by killing enemies. In the Rig Veda the priests who officiated over the midwinter rite were called dog-priests, the Vrâtyas. Their sacrifice was a cow. The winter-season ceremony at KS seems to have included both dogs and cattle.

In Germanic traditions, during the Twelve Days of Christmas Odin roared through the forest in the guise of a hunter with his pack of dogs; Odin also was the chief deity of the Mannerbünde, who were initiated at Yuletide and wore the skins of dogs or wolves. In the Roman Lupercalia, an obscure and archaic Roman ceremony conducted in mid-February, young male celebrants chosen from the best families ran naked around the walls of the old Palatine city of Rome with thongs made from the skins of sacrificed dogs and goats, drawing a magical boundary around the oldest part of the city. The KS settlement was bounded by a shallow ditch that stood between it and the cemetery, perhaps a parallel for the magical boundary-marking of the Lupercalia. All of these ceremonies were closely linked to other ceremonies that propitiated the spirits of the dead and purified the lives of the living (Harrison 1903:50-55). The Anthesteria, a related ceremony of ancient Greece also conducted in mid-February, also propitiated the spirits of the dead. In this season Hermes conducted the souls of the newly-dead out of their graves with a simple wooden stick, a magical wand called the rhabdos. The waterlogged well deposit inside the structure at KS (Pit 10) contained at least two, perhaps four long wooden poles or wands with notched ends, the only finished wooden artifacts in the well. The image of the rhabdos on the Greek vase on the left below was taken from a discussion of the Anthesteria in Harrison's classic 1903 book on Greek rituals; the image on the right is one of the wands we recovered from the bottom of the well, Pit 10. The third photo, on the bottom, shows the full length of the wand; a fragment of another carved end piece just to its left, from another wand; and other fragments of wood from the well bottom.

Archaeological evidence for an Indo-European midwinter dog-sacrifice has not been found before. Unique cases are notoriously difficult to interpret. But at least three separate lines of evidence converge on a ritual interpretation of the dogs: the ceremonial butchering of the dogs’ heads, contrasted with that of the other animals; the winter season of death for the dogs, contrasted with the year-round slaughter of cattle; and multiple correlations with a midwinter dog-centered initiation/purification ritual reconstructed by Indo-Europeanists.

The excavated structure seems to have been a normal outbuilding for a normal Pokrovka and Srubnaya settlement during most of the time it was occupied. But at midwinter it also served as the sacred compound for a mid-winter initiation ceremony, or, more probably, was the place where the refuse from this ceremony was dumped and buried after the ceremony was finished. The burning and careful segmenting of the dog heads could have been a way of ‘killing’ or de-sacralizing ritually-worn dog skins before they were discarded. Studies of the intra-site patterning of the dog remains inside the KS structure are ongoing, and will be described in detail in a future report.