ÇATALHÖYÜK 2004 ARCHIVE REPORT

RESEARCH PROJECTS

Fire and Burning at Çatalhöyük: Integrating Forensic Practice

Karl Harrison

Department of Archaeology, School of Human and Environmental Sciences, The University of Reading

Abstract

Whilst the traces of structural burning at Çatalhöyük arguably form one of the most vivid and dramatic images of the site, there would appear to have been comparatively little close analysis of the preserved material, from the point of view of the burning event itself.

This season's work has attempted to begin the process of integrating techniques from the fields of forensic fire investigation and fire engineering. It is hoped that a methodology integrating this practice, together with experimentation and observation in the field might be developed.   This would assist in analysing traces of structural burning and consequently with the formulation of hypotheses regarding the character and origin of these burning events.

Özet

Çatalhöyük'teki yangın geçiren yapıların izleri, yerleşimdeki en canlı ve dramatik tablolardan birini çizse de, korunmuş kalıntıların, üzerinde yangının kendisi ile ilgili çok az inceleme yapılmıştır.

Bu kazı sezonunda, bu süreci anlamaya yönelik olarak, yangın incelemeleri ve yangınla ilgili mühendisliklerin araştırma yöntemlerini uygulamak üzerine, bir çalışma geliştirilmiştir. Bu pratiğe bağlı olarak, deney ve gözlem yoluyla bir metodoloji geliştirilmesi umulmuştur. Bu, yangın geçiren yapıların izlerini incelemeye ve dolayısıyla yangının çıkış sebebine ve karakterine yönelik bir hipotez geliştirmeye yardımcı olacaktır.

Introduction

As a site, Çatalhöyük exhibits what would seem to be a wide range of uses of combustion, ranging from domestic fire installations, to dramatic evidence for apparently intense and extensive structural fires. It has previously been noted that the presence of traces of burning was one of the strongest initial impressions at the discovery of the site in 1958 (Cessford & Near, 2004).

Certainly it would seem that the ubiquitous traces of structural fires; discoloured burnt mud brick and lengths of charred timbers continued to interest Mellaart, as he devoted a significant degree of attention to the description of patterning and extent of temperature-related alterations. Furthermore, he attempted to use these observations to try to make sense of the origin and development of these fires, as well as the motivation behind them.

More recent research at the site of Çatalhöyük has continued to document and recover this extensive range of traces of burning, such as charcoal scatter and charred timber, mud brick variously discoloured by reduced or oxidised combustion, and baked plaster from floors, ceilings and walls (Matthews & Farid, 1996). These traces have been used in past research to describe and interpret events and activities associated with the Neolithic occupation of the site, such as the conflagrations that appear to plague Çatalhöyük from occupation Level V onwards (R. Matthews, comment in Gérard, 2002).

Evidence for these conflagrations in particular has fuelled debate on the perceived intentionality of the house and shrine fires, coupled with parallels made regarding the “use-life” of buildings, drawn from research conducted in southeast Europe (Cessford & Near, 2004; Stevanović, 1996). The stratigraphy of these burned remains has been further complicated by the reuse in the continuing occupation of the site (as exemplified in their use as floor packing; Stevanović and Tringham, 2001).

Close observation and description of traces of burning at Çatalhöyük have been recorded over a long period, relative to most archaeological projects. These observations have prompted discussion as to what physical or cultural processes in the Neolithic may have brought about this burning (Mellaart, 1966). It would seem that between observation of present traces and discussion of past behaviours, there is scope for the development of a means by which the initial descriptions of destruction by fire might be better understood and more reliably and rigorously modelled; a means which is necessarily scientific in character, being both repeatable and testable in character.  

Such analytical techniques have previously been referred to as ‘middle range' theories, due to the means by which they attempt to bridge a gap between archaeological evidence and interpretation (Binford, 1977); or else attempts to understand the formation processes of the archaeological record (Schiffer, 1987). More recently, work at Çatalhöyük has sought to contextualize such scientific techniques within the wider excavation process, in order to promote a multivocal response (Hodder, 1999).

The potential offered by traces of structural burning have generally been little considered for the information they may have to offer regarding the origin, development and nature of the fire, as well as the geometry, contents and fabric of the structure itself. Experiments with archaeological constructs have, in the past, provided an opportunity to observe the sequence of destruction within buildings (Reynolds, 1994; Hansen, 1962). In the context of archaeological sites, some attempts have been made to assess the intensity of structural fires from their preserved traces, such as with the fused glass of Fishbourne Roman Villa (Cunliffe, 1971), or in one notable case, the express attempt to develop a method of excavation in order to ascertain the origin and development of fires within structures of the Balkan Neolithic (Stevanović, 1996; Tringham & Bruckner, 1985). Special mention must be made of the work of Paul Murley, who has made a study of the effects of fire damage on the properties of limestone building materials (Murley, 2003a; 2003b).

Fire investigation and fire engineering are modern scientific disciplines with a range of techniques of potential use to archaeologists interested in optimising data relating to episodes of burning within structures. Whist the primary aim of forensic fire investigation is the detection of a seat of fire and its subsequent development within a building (Redsicker & O'Connor, 1997), it may also be able to shed further light on a range of variables which might affect the growth of a fire, such as available fuel load, room geometry, position and degree of ventilation, intensity and duration of burning, and the possibility that a fire may have been intentionally set, or extinguishing attempted (Cooke & Ide, 1985).

Given the limited time available, an initial concern was the identification of a range of case studies, which might successfully demonstrate the potential information that might be offered through an integrated approach of archaeology and fire science. It is hoped that these case studies may form the basis for further concerted study.

Case Study 1: Underground Burning and the Modelling of Fire Dynamics

The first study is initially drawn from an observation first made by James Mellaart. In commenting on the amazing intensity of the conflagration at the end of Level VIa, he reported that the fire was so fierce that the heat “…penetrated to a depth of about three feet or more below the floor level of the buildings, carbonising bodies and burial gifts alike and preventing all further bacterial decay. To this we owe the conservation of numerous perishable materials: flesh and tar on human bones, desiccated and carbonised brains inside human skulls…” (Mellaart, 1964, 85).

This conclusion, that a structural fire might be capable of charring combustible objects a meter below the ground, seems intuitively improbable, due to the degree of thermal energy that would be required. Despite this, however, more recent works that have examined fire and burning at Çatalhöyük, have not sought to challenge this observation (Cessford & Near, 2004; Matthews & Farid, 1996).  

Whilst the traces of burning observed by Mellaart himself cannot be reviewed, the techniques of fire science may be employed to quantify in broad terms the amount of heat required to produce the phenomena. From this, some conclusion may be drawn as to the likelihood as to whether a structural fire above might be the cause. One limitation of this approach is that it is a desk-based assessment only. It aims to produce a model that will test the viability of Mellaart's original hypothesis.

By utilising methods from fire engineering, it is relatively straightforward to calculate the amount of energy required to cause charring in buried fuel. From this, it is then possible to calculate the length of time a structural fire would have to burn in order to generate a critical amount of thermal energy to cause a reaction. Furthermore, it is then a relatively simple process to derive a figure for the mass of fuel required to support such a blaze.

In this instance, the most generous of assessments produces a range of statistics that suggest the burning of timber buried a metre beneath the ground is at very best improbable. In a 25m 2 structure, that might be considered characteristic of Çatalhöyük buildings (Mellaart, 1975), In the region of 5645 kilos of timber would be required to maintain a fire over a period of approximately 7.1 hours to begin to achieve the desired pattern of underground burning.

Such a quantity of fuel would be improbable enough, especially in a sealed compartment that would preclude constant fuelling from outside. However, the above model assumes total combustion of the fuel. In reality, a proportion of the fuel would survive as charcoal in protected regions of the structure. Likewise, no calculations have been considered for the level of ventilation required to supply such a fire with adequate oxygen.

Whilst this model is based on modern analogues for ancient materials, and as such cannot be regarded as anything more than a broad projection. It does, however, provide an order of intensity for an ancient structural fire, and the fuel required. This appears to be sufficient to call into question Mellaart's interpretation of the observation of buried charred timber, suggesting that another explanation might be looked for.

Case Study 2: Fire Investigation in Plan and Section

In addition to the ability of fire engineering to quantify the dynamic and potential forces involved in a structural blaze, fire investigation makes use of directional and intensity-related burning patterns to determine ‘seats' of fire (concentrations of burning that may constitute possible points of origin; Cooke & Ide, 1985), and its subsequent development from those seats (Redsicker & O'Connor, 1997). During my short visit to Çatalhöyük, I was able to briefly examine the burning affecting Building 45 in plan, and the section east of building VII.32, both of which exhibit traces of structural fire (Figs 110 and 111).

Figure 110: View of Building 45

The fire in Building 45 was largely indicated by the presence of several lengths of charred timber, which on cleaning appeared to have fallen from the corners of a raised platform, where the remains of timber uprights appear to stand in-situ.  

Three of the timber lengths were arranged parallel to one another and aligned along their grain, with the end of one piece terminating by the remains of one of the surviving upright sections, suggesting these may represent the remains of a single timber upright, subjected to a low-intensity smouldering combustion along its length. If this is so, then the apparent gaps in its fabric may have resulted from processes occurring following its collapse, rather than in-situ burning. This process would most likely be the continuation of smouldering combustion. This pattern of low-intensity burning continuing long after the structural fire proper is especially prevalent where the intersection of fuel masses increases the surface area of fuels within a close proximity.

Such combustion of closely associated structural timbers has been observed in construct experiments, where wooden members were seen to smoulder for far longer at their joints than they did along their lengths (Harrison, 2001). This observation suggests that, had an intense fire burned in Building 45, these timbers would not likely be closely associated with the focus of destruction, unless they were somehow protected by the geometry of the structure, or by early collapse during the fire's development. Further excavation of this area might be expected to reveal further information on this sequence of destruction.

The section east of VII.32, originally described in detail by Matthews and Farid (1996), but since subject to further erosion, appears to preserve a more complex sequence of destruction. A layer of white ash appears sealed in the northern corner of the room, against the plaster of both the floor and the northern return wall. Sealing this ash is a layer of relatively lightly discoloured mud brick, the top of which appears to peak c. 1m south of the north wall, at c. 0.75cm above floor level. These mud bricks exhibit far less definition of shape than those above, more highly discoloured by fire.

Immediately overlying this heap of slightly discoloured mud brick are two charred timbers. Much like those of Building 45, they appear to be heavily charred, but otherwise well preserved. Their position beneath the more highly discoloured mud brick suggests that they may have collapsed early in the fire, falling away from a source of burning somewhere to the south.

Figure 111: Section East of VII.32

Fractured sections of mud brick, which exhibit a banding of discolouration, from pinkish sintering to black, reduced burning are most likely traces of spalling from the now-absent northern section of the structure's west wall. Such spalling is observed in modern plaster and brickwork, and is caused by the rapid heating and cooling of the wall of ceiling surface, which causes a rapid expansion of the moisture trapped within the fabric's substrate, and consequently causes laminar fracturing (Redsicker & O'Connor, 1997). The spalled sections visible here seem to be embedded in the largely unburned heap of material, again suggesting this heap's deposition occurring before the most destructive phases of the fire.

This section preserved numerous signatures relating to direction and intensity of combustion that may prove characteristic of the origin and development of fire; the most notable of these occurs c. 4m south of the internal plaster line of the north wall. Here the body of the western wall is relatively well-preserved, and an outward slope in the section allows the line of the internal plaster to be looked down at, between the brick of the west wall and the fill of the room's interior. The line of plaster preserves evidence of what appears to be two moulded outcrops, triangular in plan, jutting out from the line of the wall a short distance into the room.

Within the internal angles of these two mouldings, an area of plaster c. 1m in length appears to preserve evidence of extreme heating at a very low height within the structure. Pink discoloured plaster has been retained in-situ between the mouldings, but does not appear to occur outside of either of them. Fire investigators regard such ‘low burning' as a strong indicator of a potential seat of fire (DeHaan, 1997). The sharp distinction between pink and white-burnt plaster here may further suggest that the area between the mouldings was subjected to a low-intensity or smouldering fire over an extended period, before the development of an episode of burning sufficient to destroy the structure.

The evidence preserved in this section suggests a process of burning that goes beyond a simple explanation of a single, catastrophic fire. The ashes first mentioned appear to be sealed and separate from the rest of the burnt material, and do not appear to feature any charcoal inclusions. Further microscopic analysis of the sample taken from this level would hopefully be enlightening of this point.

Much of the most dramatically discoloured material occurs in the upper level of the fill, sloping from c. 1.5m up the north wall down into the structure's interior. This material is some distance from the possible seat of fire, and as such might originate from the ceiling of upper portion of the walls, which would likely have been subject to the majority of the thermal energy produced (Cooke & Ide, 1985); particularly given the apparently limited ventilation of structures at Çatalhöyük.

These suggestions are necessarily preliminary due to being based on observations of the cleaned section. Excavation of the structure's interior in future seasons will hopefully provide far greater information on which to base hypotheses regarding the cause, origin and development of this apparently complex fire on. Hopefully, this in turn may assist in postulating why such dramatic fires became a characteristic feature from Level V onwards.

Taken as a whole, the incidents of fire use and burning discussed here can be seen to move away from a simplistic concept of the intensity of fires and the effects of high temperature, and towards an understanding of complex relationships; both the physical relationships balanced in combustion processes (energy release rate, energy absorption, available fuel and ventilation), as well as the interaction of people with the processes of combustion. It is hoped that by evaluating the available data in the light of integrated techniques from fire science, it may be possible to further inform on a broad range of these issues, in a quantifiable and testable manner.

This initial survey of the traces of burning preserved at Çatalhöyük has only begun to assess the potential information that might be gained by such methods as these. The wide range of issues that have been briefly discussed here will almost certainly benefit greatly from the application of further research.

This article began by referring to the integration of fire science within an archaeological context as being a ‘middle range' theory; a rigorous and testable methodology by which greater insight might be gained into a range of past human behaviour. Whilst such the division of archaeological theory into levels has more recently been brought into question, the distinction still serves to sound a cautionary tone; such work as this cannot lose sight of the fact that it exists to inform about human subjects. A priority for future research must therefore be a wholly integrated work, which considers the science of structural burning in the context of the anthropology of response to fire.