Khirokitia, Tholos III: Burials Under the Floors of a Neolithic House

Dikaios called the round buildings of Khirokitia tholoi and numbered them. Tholos III stands in his Area I, squares E–H 13–15, on a steep rock slope, built on top of earlier floors where people had lit fires in the open. Its back is against the long stone structure that Dikaios called the “main road” and that Alain Le Brun later reinterpreted as the village’s enclosure wall (Dikaios 1953, pp. 43–47; Le Brun 1997).

The wall of the house has two skins. Outside is a ring of stone about 0.30 m thick, capped with boulders against damp; inside it, a ring of pisé, massed earth laid by hand, 0.50 to 0.70 m thick and finished with a thin plaster of light yellow silty earth (Dikaios 1953, pp. 45, 47). Scaled from his plan, the room inside is about 2.5 m across and the house about 4.5 m across the outside. About two-thirds of its footprint is wall.

The door is on the south-east, and it opens 1.30 m above the foot of the wall. An outside stair of about five steps climbed to the top of the village wall. From there a walk of a metre along the wall top, behind a parapet of boulders 0.46 m high, led to the door, and four steps, partly of boulders, went down inside to a broad pisé step on the first floor (Dikaios 1953, p. 47). On the north-east a niche 0.45 m high, 0.41 m wide and 0.29 m deep let light into the room through a crescent-shaped slit between the stone and the earth of the wall. Dikaios noted that the sun shone straight in from early morning to early afternoon in June and July. Two shallow plastered niches were cut in the north wall.

Dikaios gave his tholoi domes. Le Brun’s team found a house at Khirokitia whose roof had burnt and fallen in one piece, and it had been flat, made of branches and reeds under coats of earth; the model gives Tholos III a flat roof of the same build, and the dome is offered only as a rejected alternative. The argument is set out in full in the household reconstruction. The only Khirokitia doorway that survives to its full height, in Tholos V, is 1.63 m high, and the model borrows that height.

A Khirokitia floor is a coat of earth plaster, “yellowish or darker”, 2 to 8 cm thick and as a rule smoothed (Dikaios 1953). When it wore out or was dug through, another went on top, over whatever had collected on the old one: ash, bone, broken tools, the trodden dirt of living. In Tholos III Dikaios counted ten house floors, which he numbered from the top, I, down to the first, X. In his section through the house, drawn north to south through the door and the centre of the room, they climb about 2.2 m from floor X to floor I (Dikaios 1953, fig. 22; the level of each floor is our reading of that drawing, to about 10 cm).

The first floors held no graves. Floor X is burnt pisé with ash and animal bone; floor IX a levelling of pebbles; floor VIII has an oval pit lined with pisé and a semicircular partition of pisé enclosing the south-west of the room, burnt along its edge, which Dikaios thought might have been a hearth (Dikaios 1953, p. 48). Two picrolite beads, a pig’s tooth and a scrap of a bone pin were lost in these floors.

Dikaios divided the life of the house into four stages (p. 52). His tally of its dead reads: “floor VII one adult, floor VI one adult and one infant, floors V–II none, floor I two adults and one infant” (p. 214). Six people under ten floors, three graves in the middle of the house’s life and three at its end, with four floors and most of a metre of fill between.

That spacing bears on an old argument. Dikaios took the number of burials in a house to follow from the number of its floors, that is, from how long it was lived in. Karin Niklasson suggested the reverse, that the floors followed the burials, a new floor going down because someone had been buried (Niklasson 1991, p. 169, as reported by Jones 2008). Tholos III fits neither view fully. Floor VI went down directly over a fresh grave, but floors V to II were laid with no burial under them at all.

The first grave was cut through floor VII on the north-east side of the room. Scaled from Dikaios’s floor plan it was about 0.85 by 0.38 m; its depth is not published, and the model gives it 0.25 m. In it lay a woman Dikaios identified as young, on her back, her head to the north-west. Her left leg was doubled up at an acute angle to the spine and lay flat; her right leg was bent with the knee raised against the side of the pit (Dikaios 1953, p. 48).

Over her head and the top of her chest lay a quern, 1345, an elongated grinding slab of andesite 32 cm long. A quern that size weighs something like 8 to 15 kg (our estimate). It is a working tool of the ordinary kind, and Dikaios assigned it to his period II (p. 292). By her jaw lay a bone pin, 1346, 7.5 cm long with its point missing, and by her left side four flakes of flint, 1347 (pp. 48, 53). The catalogue first printed these three numbers under floor VI and then corrected them to floor VII (p. 53).

Dikaios describes the stone three ways: “on the head and upper part of the thorax” in the account of the house, “on the upper part of the chest” in his survey of the burials, and, comparing graves, “the floor covered the body but not the head… the head was under a quern” (pp. 48, 218, 221). Read together, the quern lay over the head and the upper chest, and that is where the model puts it.

Then the household laid floor VI over the grave and left an oblong hole in it above the quern, so the stone stayed visible from the room. Its size is not given in the text, so the model makes it 0.40 by 0.25 m, the stone’s outline with a margin, which is our reconstruction. When a stone wall was later built in the room, it was placed so as not to cover the hole (p. 48).

Dikaios offered a reason: the hole “may be connected with a cult of the dead, in which libations were poured through the hole” (p. 48). No vessel, residue or staining was reported at it, and the model shows nothing there. Other houses on the site kept a grave open in other ways. In Tholos XV the floor covered a woman’s body but left her head under loose earth (p. 221); Karageorghis thought libations were poured on heads left exposed (Karageorghis 1982, p. 28); Paula Jones notes channels left open into some graves, which would have allowed the dead to be reached after burial (Jones 2008, p. 117).

Floor VI is the busiest floor in the house. At the centre of the room was a paved hearth 0.30 m across; by the door, a sunken area 0.53 by 0.43 m paved with gypsum slabs; west and south-west of the hearth lay two stone pounders, 1321 and 1322. A diabase dish 24.5 cm across, 1320, and a fragment of a stone bowl, 1319, lay on the same floor (Dikaios 1953, pp. 49, 53). On Dikaios’s plan the hole over the young woman’s quern was less than a metre from the fire.

The second grave was cut through this floor against the south-west wall, inside the old pisé partition, which was now walled with a curved line of pebbles and pisé 0.85 m long. The pit, 0.90 by 0.35 m and widening as it went down, bit into the pisé of the house wall. In it a middle-aged man lay “exceptionally contracted”, face down, his head turned slightly up and wedged with lumps of pisé. The pit was filled with loose earth and then with pebbles up to the floor, and among the pebbles were pieces of two stone bowls: 1323, an andesite fragment 12 cm long, and 1324, part of a coarse limestone bowl 14.5 cm long (pp. 50, 53).

The third grave was a small pit dug through the same floor by the east wall, for an infant Dikaios judged to be under a month old. A pig’s tooth, 1341, and a small unfinished diabase bowl, 1427, rectangular with an open spout and a broken handle, lay next to it. Dikaios was unsure whether they were offerings (p. 50). Most infants buried at Khirokitia died very young: in one group of 51 from the French excavations, Françoise Le Mort found that 50 had died before six months (Le Mort 2000).

Dikaios noticed that hearths tend to sit directly over burials, though he saw no meaning in it (Dikaios 1953, p. 204, as cited by Jones 2008), and that in a house the earlier graves lie at the edge of the room and the later adult graves at the centre (p. 216). Tholos III follows the second rule. Its three early graves lie by the walls, to the north-east, the south-west and the east, around a hearth that people went on using with the graves under their feet.

What happened next in the room happened on top of the graves. Four platforms of boulders were built on floor VI, one of them partly over the infant’s grave. Against the south-west wall a stone wall 1.30 m long, 0.45 m thick and 1 m high was built over the grave of the man, stopping short of the hole above the young woman (Dikaios 1953, pp. 48, 50).

Then the platforms and the wall were joined into a continuous ring of stone, faced with pisé and raised to about a metre. The ring blocked the old door, the window and the inside stair, and a new door was opened on the east; it was not found, and Dikaios inferred it from a ramp paved with boulders on that side (pp. 47, 50–51). The room inside the ring was filled with earth, lumps of pisé, flints and bone to 0.30 m below the top of the ring, about 0.7 m in all, and floor V was laid on the fill (p. 51). The three graves of floors VII and VI now lay 0.7 to 0.8 m below the people living in the room.

Dikaios does not say when the hole over the quern was closed. The fill and floor V cover the spot, so it was closed by then at the latest; the model seals it with the ring, which is our inference. In our estimate the wall over the man took about 0.6 m³ of stone and the fill about 2.1 m³ of earth, far more work than all the graves together.

Floors V and IV carried no burials. Ash built up on them, seven small bone tools, catalogued together as 1310, were lost between them (an awl, a pointed tool, two pins, one of them doubtful, a needle and two borers), and a broken diabase tray 23.5 cm long, 1307, lay at the north edge of floor IV. Before floor III was laid, a layer of dark earth and ash and then a layer of hard light pisé went down over the room. Dikaios read them as the fallen superstructure or a deliberate fill. The roof was set again on the outer stone ring, the inner pisé was taken into the floor, and the room widened to about 2.9 m (pp. 45, 51; the width scaled from fig. 23). The model rebuilds the roof a metre higher at this point, because the floor had risen by more than that; the height is ours.

No tool marks are reported from these pits. They were cut through rammed earth and plaster and into the soft deposit beneath, which digging sticks, stone picks and hands would manage, an inference from analogy. The spoil was small. In our estimate, grave VI gave about 0.06 m³ of earth, grave V 0.10 m³, the infant’s pit 0.04 m³ and the central pit of the last burials 0.3 m³. Each is under an hour’s work for one person, again our estimate. The labour lay in the plastering afterwards, and in what was later built over the graves.

The four grave pits of this house were sealed in four different ways. Grave VI went under a new floor with a hole kept open over the stone. Grave V was filled with loose earth and then pebbles to floor level, with pisé packed round the head. The infant of grave IV was later partly covered by a platform. The pit of the last three burials was capped with pisé and pebbles.

After four floors without a burial, a pit 1.0 m across and 0.40 m deep was cut at the centre of the top floor (Dikaios 1953, pp. 51–52). A young adult, probably a man, was laid in it face down, head to the north, the right leg drawn up at an acute angle so that the knee came close to the head and the left leg bent lower. A flat pebble lay on the back of his skull. Dikaios records a quern, 1308, of andesite and 35.5 cm long, “on the chest”; with the body face down, the model lays it across the upper back over the chest, which is our reading.

Between his skull and his right knee lay a small spouted bowl of andesite, 1306, 11.5 cm long, with a flat base and inturned sides. Dikaios judged that it had been broken on purpose, and one piece of it lay on his right thigh bone (pp. 51–53). How it was broken is not recorded. By the right knee lay a triangular red pebble of iron-stained andesite, 1309, 9.5 cm long. Broken stone vessels turn up in most Khirokitia graves; a spouted diabase bowl of the same family, lent by the Cyprus Museum to the Metropolitan Museum, shows the shape whole.

A middle-aged adult, probably a woman, was buried in the south half of the same pit, apparently later. Dikaios thought this burial disturbed the man’s and twisted his head into its “queer” position (p. 52). No objects are listed with her in the account of the house, but Dikaios’s general catalogue has a quern, 438, from “tholos III, grave II”, found 0.40 to 0.60 m down. It is not in the house’s own list of finds, so it may have lain in the fill of the pit rather than on her. The model includes it, placed in the south half of the pit and marked as placed by us. Above them both, in the upper fill, was an infant, “much decomposed”, with nothing beside it (p. 52).

The pit was sealed with a cap of pisé with pebbles pressed into it, and among the pebbles were two pounders: 1303, of andesite, 9 cm long with one flat face, and 1304, a near-sphere of grey chalcedony 8 cm across. An unfinished tray or quern, 1305, 25 cm long and never hollowed, lay a little to the north-east (p. 52). Then a last ring of boulders, three or four set one on another, was built against the wall, which Dikaios linked with yet another new superstructure. “Soon after this operation”, he writes, “the tholos was abandoned” (p. 52).

Dikaios printed every find from the house floors in one list, 23 catalogue numbers for floors X to I (Dikaios 1953, fig. 24, pp. 52–54). With the quern 438 from his general catalogue that makes 24 numbers. Two of them cover several things, the four flint flakes of 1347 and the seven bone tools of 1310, so the numbers stand for 33 objects, and the uncatalogued pebble on the man’s skull makes 34. Finds from the floors beneath the house, which are older than the building, are left out.

Of the 34, 15 lay in or on the graves or in the earth that sealed them, and 17 if the two pounders in the cap over the central pit are counted:

• Grave VI, the young woman: the quern 1345, the bone pin 1346, four flint flakes 1347. Six objects.
• Grave V, the man: fragments of two stone bowls, 1323 and 1324, among the pebbles over him. Two.
• Grave IV, the infant: the pig’s tooth 1341 and the unfinished bowl 1427, perhaps offerings. Two.
• Grave III, the young man: the quern 1308, the broken bowl 1306, the red pebble 1309, the flat pebble on his skull. Four.
• Grave II, the older woman: perhaps the quern 438. One.
• Grave I, the infant: nothing.
• The cap over graves III to I: the pounders 1303 and 1304. Two.

The other 17 are what a household drops over generations: pounders by the hearth, a dish, a bowl fragment, two trays, the bone tools, two beads, a pig’s tooth, a broken pin. The grave goods are the same kinds of things: grinding stones, stone bowls, a pin, flint flakes, pebbles. The two querns laid on bodies were working tools. The house belongs to the Aceramic Neolithic, and there is no pottery anywhere in it. Nor is there a necklace, though dentalium and carnelian necklaces, found only with women, occur elsewhere on the site (Jones 2008, p. 77; Le Brun 1997).

Set against a Bronze Age family tomb the scale is plain. The tomb at Vounous rebuilt on this site holds 142 objects for 9 burials, about 16 for each person and almost all of them pottery. Tholos III gave its dead fewer than three each, and the last infant nothing at all.

Tholos III is one house of many. Some 252 people have been excavated at Khirokitia: 152 by Dikaios, nine in later soundings and 91 by Le Brun’s French mission (Jones 2008, p. 77). About a fifth of them had a large stone set on the body, a rough boulder or a quern, mostly in adult graves, and broken stone vessels occur in the great majority of graves (Jones 2008, p. 77). Le Brun found the stone about twice as common with men as with women (Le Brun 1997, pp. 29–30). In Tholos III the young woman of grave VI and the young man of grave III both had one.

Vassos Karageorghis gave the classic explanation in 1982: the stone was put on the body “to prevent the dead from returning and harming the living” (Karageorghis 1982, p. 27). Paula Jones notes the same custom at Jericho, at Nahal Oren and as far back as the Natufian at el-Wad, and reads it as fixing the dead in their place (Jones 2008, p. 117). Neither reading can be tested, and the hole in floor VI sits awkwardly with both: the stone that held the young woman down, on either reading, was kept in view from the room. Neither reading appears on the model’s labels.

For the broken bowl there is less to go on. Dikaios judged that 1306 had been broken deliberately and saw one piece set on the body. Why it was broken is not recorded, and at Khirokitia breaking stone vessels for the grave was the usual practice.

Other arguments concern the whole village. Niklasson criticised inconsistencies in Dikaios’s readings of the burials (Niklasson 1991), and Jones asks whether parts of the site served to display the dead and keep them within reach, as well as to house the living (Jones 2008, pp. 89, 137). Le Brun, who excavated the site from 1977, put it plainly: the dead were buried inside houses that were not abandoned, so “the dead therefore remained with the living” (Le Brun 1997, p. 29). Dikaios himself read the late double and triple graves, like the pit of graves III to I, as a sign of rising deaths before the village was abandoned (Dikaios 1953, pp. 214–215). Later writers, Niklasson and Jones among them, read the burial pattern in social or ritual terms instead. Dikaios also spoke of human sacrifice for two other houses, never for this one.

The grave visitors know best from the guidebooks is one of Le Brun’s: an oval pit 85 by 70 cm and only 16 cm deep, with a quern on the chest of the body (Karageorghis 1982, p. 27, fig. 10). It is smaller and much shallower than the central pit of Tholos III, which was 1.0 m across and 0.40 m deep, and like grave VI it held one body under one quern.

When the house was left, the roof and the upper walls fell in and earth built up over them; the fall is our inference. Erosion, recent digging and cultivation then took away the south half of the last ring of boulders and damaged the top of the central pit (Dikaios 1953, p. 52).

Dikaios dug the house floor by floor for the Department of Antiquities and published it with a plan and section, nine floor plans with the graves and the finds marked by number, and the full list of finds (Dikaios 1953, figs 21–24, pls XIV–XV, XIX). Few houses on the site can be followed so closely. The finds went to the Cyprus Museum in Nicosia; two of them, the bowl fragment 1319 and the pounder 1321, were already missing when the report went to press.

The model divides the life of the house into twenty events, from its building to its excavation. The layers fix their order but give no date for any of them. Tholos III belongs to Dikaios’s periods II and III, the later part of the village’s life, and the Khirokitia culture came to an end in the later 6th millennium BC; how long the ten floors took to accumulate is unknown.

The house, its ten floors, the six graves, the postures of the bodies as Dikaios described them, the stone wall and platforms over the graves, the ring and the fill, and every object in the list are excavated. The model shows each of them, and its “Where positions come from” view colours every object by the source of its position. The positions of 17 come from Dikaios’s floor plans, where he wrote the catalogue number at the find spot; 16 from his text, which says on which floor or beside which body a thing lay; and 1, the quern 438, from our own rule. The floor plans do not register with one another to better than about 0.3 m, so positions carry an uncertainty of about 0.2 m.

The ages and sexes are Dikaios’s readings in the field, two of them marked “probably”; the model does not draw on any later study of these six skeletons. In the drawings, as in the 3D model, the dead are shown as skeletons laid in the postures he describes; the way the man in grave V and the woman in grave II face is our reading of small drawings.

Some things are ours outright: the depth of grave VI and the size of the hole over its quern; the thickness of the inner ring; three of the four platforms, whose positions the text does not give; the height of the walls and roof at each stage; and the moment the hole was closed. The roof is flat on Le Brun’s evidence, and the dome Dikaios drew is shown only as a rejected alternative. The orientation of the house rests on our reading of Dikaios’s north arrow, to within about ten degrees.