Tell Atchana/Alalakh

Burn the House Down 2008

Experimental Archaeology

Replication of a LBA Kiln 

  Kiln Gallery

Making Mud Bricks

   Mud Bricks Gallery

Making Tandir

   Tandir Gallery

Burn the Brick

Burn the House Down

   Burn the House Gallery

   Burn the House Video

Ceramic Fabric Exp.

The Mud House






Another experimental archaeology project was carried out during the 2008 excavation season at Tell Atchana/Alalakh. In order to observe the affects of fire on mud brick (adobe) construction, several burning experiments simulated different levels of firing. To initiate the burning experiments, a domestic house model was reconstructed based on the excavation results from Ilipinar and other mud brick architecture sites. In future experiments, models of palatial structures, such as Alalakh, will be modeled with the extensive use of timber and flamable contents.
The motivation for these experiments was that often burnt buildings, burnt building materials, and burnt debris levels have been found during excavations. This debris can vary between light brown, orange, red and pink-red depending on the temperature and the availability of oxygen. Some of the important examples in Turkey are at Alalakh, Kültepe, Acemhöyük, Kinet Höyük, Ilipinar Höyük and many more other sites.
In the first three sites, Alalakh, Kültepe and Acemhöyük, large administrative structures were found with large storage facilities often referred to as palaces. During the fires that occurred in these sites, enormous heat was released that colored the bricks not only to a reddish tint but sometimes even liquefied (vitrified) them. Traces still can be found in their walls indicating large timber beams used as the framework in the construction of these buildings, including possible heavy wooden shelving for storage. Based on the recovered melted and vitrified basalt orthostats (blocks) from Alalakh architecture, the temperature of the fire there must have risen to 1100 degrees celcius.

In the fourth example, at Kinet Höyük, the temperature was not that high and many bricks turned into orange-brown while remaining rather soft. In this example spaces indicating burnt out timbers were also found inside the wall construction.

The last example, Ilipinar Höyük, had a row of 15 Neolithic houses burnt down to the ground, which had turned into a sintered black, pink-red, red orange, and brownish-black burnt debris.

The first stage of our experiment was to build a small mud brick room, the second stage was to fill it with different fuel, and the final stage was to burn it down. The idea was to get an impression of the effects of a ‘domestic’ fire. No large quantities of oil were used to get additional heat as might have been the case in the palaces. Materials that could be expected to be present in an ancient house were collected and placed inside. Due to the size of the room, it became very full soon, but the total contents were not exaggerated. No extra fuel material was added during the burning process.After the burning we examined the colors of the bricks and the way building elements had collapsed. The first hour of the fire was filmed and many photos were taken. The experiment was possible thanks to Tell Atchana research projects which targets traditional building methods. One element of this program was the production of mud bricks. Large numbers of mud brick were produced to make consolidation and restoration of the monumental palace walls possible on the ancient site.

The experiment was possible thanks to the Tell Atchana research on traditional building methods. One element of this program is the production of mud-bricks. Large numbers of mud-brick are produced for a possible consolidation of the monument’s walls on the ancient site. Of this pile of ready bricks, we used 168 in total, while 22 poplar posts were used, with a length of circa 1.5 meter each and a total of circa 33 meter. The walls of the building were made with mud-brick of 40 x 40 x 12 or 20 x 40 x 12 cm. No mud-mortar was used between the brick masonry (dry walling). The reason here for is that it would take too long to dry the whole structure.

The room had a small suspended floor level (shelf) on the north side and was placed 53 cm above the surrounding ground level. The raised surface was made with six posts in an East-West direction and each post had a thickness of about 11 cm. The top of the raised surface was plastered with a layer of mud plaster made of ‘mud’ with heavy straw tempering. At the time of the fire this plaster was not yet entirely dry. Two vessels were placed on this suspended surface; in the Northwest corner, a small, empty jar and in the front middle an open mouthed pot was placed, filled with olive oil.

The ends of the posts were left sticking out of the wall on both sides; on the West side nothing was done to them and some open spaces occurred between the round wood. On the East side the wood was also visible in the mud brick wall, but the open spaces were filled with mud plaster.
Next to the suspended surface a vertical post was dug into the ground and was attached to the surface by a mud plaster joint. The East wall above the suspended surface was plastered with the mud plaster and this plaster stopped in a straight vertical line in the middle of this wall.

The room’s entrance is a low open ‘doorway’. The lower half of the door was made with three baulks that had been laid over the entire south wall ending inside the West and East walls. On the West side this wood was completely covered with about one centimeter mud plaster and no wood could be seen. On the East side the wood was covered but a few centimeters of the door baulks stuck out of the mud.
The most southern lower door baulk was covered up to 50% in mud plaster that had been used for the roof construction. The roof was a mono-pitched roof; the West wall was one brick higher than the East wall. Six posts were placed between these two parallel walls. The open space between these posts was covered by thin, small wooden planks. A layer of dead plant stems were placed over these planks, substituting for reeds. This was all covered with a layer of mud plaster.

The room-fill

As mentioned above, a small pot filled with olive oil was placed on the suspended surface. Under the West side of this floor, dry, thin wood was piled up. A large number of tezek (dung) pieces were placed under the East side of the suspended surface. This dried cow dung has been in use over the centuries and has been used as the traditional fuel in areas with a dry, harsh climate. The room was filled from the standing post onwards with the dry stems of cotton plants. More tezek was placed near the doorway. The door was partly blocked with a few courses mud brick.

The fire

In the evening of a warm and almost windless day the fire was lighted at 18.29 PM. An initially slow start was soon raging out of control. After a few moments smoke came through the roof plaster. After about 3 minutes and 20 seconds the fire penetrated the roof and flames ‘ate’ a constantly larger hole in it. The 3 baulks in the doorway started to burn and the inner one soon collapsed while the second baulk fell out of the doorway at 18.51 PM. The top of the doorway collapsed completely at 19.01 PM when all the mud-brick that rested on the baulks came down. This collapse also toppled part of the lower door blocking.
The third beam, however, did not break or collapse, since the wall above the door fell and there was not enough carrying surface left for the bricks to rest on. This third beam smoldered a little more but survived the fire. It might be because this beam was partly covered with roof plaster and its position was on the far side of and high above the fire. The two standing posts burnt down soon after the start of the fire and collapsed in about half an hour.

On the West side the flames started to appear between the beam stubs of the suspended floor after about 13 minutes. Soon flames were licking the outside of the wall, showing that if this wall was a separation wall between two rooms, the other room would have caught fire by its floor beams in about 21 minutes. However it is more probable that it would burn by the flames that raged through the roof construction much earlier. On the east side this process of burning beam stubs started much later; about 45 minutes after the start of the fire.

In about one hour everything that was stored in the house had burnt away except for the suspended surface and the pile of tezek below it. This became the center of the burning for a long time. The beams of the suspended surface became thinner and thinner leaving the hollow beam impressions, similar to excavations like Ilipinar, Karatas and other sites. The suspended floor collapsed within two hours after the start of the fire. The floor collapsed by breaking in the middle and that section fell down. The sides that were attached to the walls stood up against it, their highest points were about 15 cm below their holes in the wall.

The next day and 23 hours after the start of the fire, the layers inside the building was still very warm and smoldering at some points. Although ash formed the bulk of the room fill almost no charcoal could be found, confirming the total absence of charcoal in the burnt houses of Phase VI in Ilipinar .

The collapse of the roof at the very first stage resulted with a decrease in the temperature of the fire. The durability of poplar is low and it does not include high amount of resin which will increase the temperature. Nothing remained of the beams on the ground, the beams of the suspended floor or the standing posts. The two vessels that collapsed with the suspended floor both survived the initial collapse. However, the vessel that had contained the oil had become very brittle and probably was cracked internally. When lifted from the debris it suddenly fell apart. The other vessel that was empty showed no negative effects. Both vessels had discolored from reddish buff to a grey-black tone. The empty jar was a ‘real’ vessel for daily use and is of a heavier fabric than the wider, open decorative type of pot, that contained the olive oil.

Most of the bricks that were located above the suspended surface or in that part of the room where there was no suspended floor were black with smoke and soot or were a grayish-pinkish tone. No heavy burning effects were visible in these wall segments. The walls that were located below the suspended surface with its seven wooden beams and the pile of tezek was pinkish-reddish in color. The burning effects did not penetrate very deep into the mud bricks which were a reddish color on the exposed outside but were reduced to black inside the brick followed by it natural colors. The bricks became very brittle and broke easily when lifted. The plaster layer on the East wall became very hard and was difficult to remove from the standing brick wall.



A simulated domestic fire in a heavily packed room with good oxidizing conditions with a duration of about 2 hours did not give the effects as found in the earlier mentioned burnt archaeological sites. Although some of these effects were present to a slight degree, they were not comparable to the heavily liquefied and vitrified state of these sites. Different wood types, storage materials, burning durations, and maybe different weather conditions, like windy storms may yield different results.


Ben Claasz Coockson & Murat Akar