Humans and fire: Changing relations in early agricultural and built environments in the Zagros,Iran,Iraq

Interdisciplinary theories and approaches in the study of human–fire relations

One of the most recent examinations of theories and approaches in the study of human–fire relations and the social and economic context of these was conducted by Clark and Yusoff (2014), building on earlier research by Pyne (1994, 1997) and Wertime (1973). Clark and Yusoff encourage consideration of energy-centred social theories (Bataille, 1991, 1993) that examine how communities produce and manage fire energy and how, in doing this, social existence is connected to the life of the planet more widely. They also explore Pyne’s (1994, 1997) environmental histories of fire use and management and Wertime’s (1973, 1983) analyses and histories of pyrotechnology to examine fire energy flow and how this is incorporated in social life and economic and political relations. They examine sensual and experimental aspects of human use of fire, drawing in part on Bachelard’s (1994) philosophical psycholanalysis of fire, as well as the transformative properties of fire and the social management and implications of this, drawing on Pyne (1994, 2001) and Wertime (1964). The potential implications of past histories of fire for more sustainable and equitable practices in the present and future are also considered, drawing on current political theories in the study of fuels as socially available energy and fire as an ‘entanglement of social and energetic “powers”’ (Clark and Yusoff, 2014: 211). In their analyses they develop and apply anthropogenic geologic approaches to study and trace the sources, properties, energy and impact of fire materials, properties and energy. A range of these interdisciplinary fire-centred theories and approaches are explored in this paper.

Interdisciplinary theory in fire ecology and fuel selection

Early theories from behavioural and evolutionary ecology of optimal foraging behaviour suggested that landscapes and resources such as fuel will be selected and managed to provide optimal returns in cost-benefit terms, and could thereby be examined by analysis and modelling of net energy expenditure and caloric value (Smith, 1983). However, it has been repeatedly demonstrated that many human foraging strategies are not optimal and involve consideration of a wide range of historically contingent ecological and socio-cultural circumstances and values (Zeder, 2012: 256). Current archaeological theories and approaches, therefore, are more contextual and examine in greater depth the complex co-relations between humans and environment, and the properties and networks of agencies, knowledge, energy and materials. Zeder (2012: 257) in particular, highlights fire as one of the ways in which humans modify the environment to enhance the abundance, accessibility and predictability of plants and animals in particular niches, and thereby create new types of environments and relations with and between species, drawing on niche construction theory (Odling-Smee et al., 2003). In the Zagros case studies below, this paper examines whether landscape fires in the Early Holocene were natural or anthropogenic in origin, and how fire may have impacted on environment and human, plant and animal species and relations and niche construction (section ‘Early Holocene environment and vegetation biomass: Fire incidence and fuel sources’).

The selection of materials for fuel and combustion is rarely directly correlated with environmental abundance or distance from source. Factors influencing the choice of firewood include: the production and abundance of readily collectable drywood which varies according to tree/shrub species, age and structure of stands, seasonality, climate and environment (Asouti and Austin, 2005: 8); microclimates, environments and responses of particular species to natural or anthropogenic disturbance by fire, grazers, browsers, clearance for cultivation, and management by coppicing and cutting (Asouti and Austin, 2005: 8). Other factors influencing fuel selection include: ecological and social strategies, which vary between mobile hunter-gatherers, nomadic pastoralists and settled agriculturalists (see summary in Asouti and Austin, 2005: tables 2–4); associations between particular tasks and goals in which fuel collection may be combined with hunting, gathering or foraging for other resources such as plants for food or crafts (Charles et al., 2014); as well as social and cultural perceptions of particular resources and their performance properties, value, use and materials/residues arising from these. Investigations of these factors can be informed by anthropological and anthropological theories on the inter-relations between environment, dwelling and taskscapes (Ingold, 2000), and the entanglement of flows of energy, matter and information (Hodder, 2012), which are used here to study the relations between fire and particular communities, places and tasks (Hodder, 2011: 157; Marchant and Lane, 2014). There has been some tendency to polarise male and female roles and tasks in the Neolithic, associating hunting and herding with males and the domestic sphere with females (Hodder, 1990). The interdependent roles of men, women and children, however, are more clearly understood if all stages in particular strategies are considered, such as here with regard to fire, fuel collecting, transport, processing and storage, and use of fire in hunting or land clearance and food processing and cooking, drawing on gender and life-cycle studies (see Bolger, 2010: 507). The interconnectedness of people with each other and environment, materials and energy have also recently been explored in studies of their entanglement (Hodder, 2011, 2012).

To examine the environmental, economic and socio-cultural considerations influencing fuel selection and fire use, therefore, the case studies investigate the correlation between local vegetation, fuel sources and ecological and social strategies in the Zagros and how these changed during the transition to increasing sedentism and early agricultural lifeways (section ‘Fuel resources and combustion properties’).

Interdisciplinary theory in the socio-economic context and specific uses of fire

The social and cultural power of fire in drawing people together as well as excluding others have been widely studied, drawing on theories in cognitive evolution (Dunbar and Gowlett, 2014; Mithen, 1998), and is explored below in the case studies on the role of fire in the development of increasing associations with place and early sedentism (section ‘Socio-economic context of fire use: early sedentism and enclosure of fire’). With regard to how fire affects the senses, perception, decision-making and social relations, well-established research in archaeology on phenomenology can be drawn on (Pawłowska, 2014; Sorensen et al., 2014; Tilley, 1994), as here to examine the changing phenomena and management of fire within early sedentary settlements (section ‘Phenomenology of interior fires: the built environment and health’). Bloch (2010), in particular, has examined how constructed settings such as fireplaces may be used to mark continuity in social roles and relations by providing a place that represents these and that can be returned to. In light of this, the construction, placement and use of increasingly enclosed fireplaces in early built environments is critically evaluated to investigate social roles and relations in more sedentary early agricultural communities in the Zagros (section ‘Fire and social roles and relations’).

With regard to the role of fire in the transformation of materials, the significance of fire in enhancing the processing, preservation and nutritional values of newly domesticated foodstuffs in the Zagros is examined (section ‘Food processing and cooking facilities’), drawing on theories and approaches in anthropology, ethnoarchaeology, experimental and food sciences (Atalay and Hastorf, 2006; Hillman, 1984; Wrangham and Carmody, 2010). The use of fire in transformations of materials for construction and artefacts and in technological innovations can be informed by consideration of wider theories and approaches on technological choices and the biographies and life-cycles of materials, which are critically applied here to study innovation in early construction materials and artefacts in early sedentary communities (section ‘Transformation of materials: technology and ritual’). Approaches to technological choices (Dobres, 2000; Schiffer, 2003; Sillar and Tite, 2000; Skibo and Schiffer, 2008) examine how ecological, social, economic, political, cultural, technical and performance considerations are evaluated in the production, transformation, use and discard of materials and explicitly examine fire ecology, pyrotechnology and use (Wertime, 1983). Theories on the biography and life-history materials (Appadurai, 1986) examine the historical agency and significance of specific materials throughout their life cycle and can also inform studies of both the ecological and social context of fire and fuel from source to discard as well as fire as cultural agency. Fire, and the phenomena and materials generated by it, have long been attributed with ritual and symbolic properties and agency (Hastorf, 2001; Hodder, 1987). In particular, smoke has been seen as an intercessionary between beings and worlds as well as having properties of purification or danger. To explore these anthropological aspects of fire, the case studies below, therefore, examine where and how fire was located, managed and used to create and transform relations between people, place and materials in increasingly sedentary and agricultural communities.

Interpretations of destruction of buildings and settlements by fire in the Neolithic of Europe have drawn on theories and approaches in archaeological studies of conflict, warfare, catastrophe and ritual (Stevanovic, 1997). They also benefitted from forensic (Harrison, 2013) and IR investigations (Forget et al., 2015). The environmental and health aspects of fire in archaeology have been investigated by drawing on research in medical and pathological sciences, the built environment and experiment (Christensen and Rhyl-Svensden, 2015). Select examples of how these can be investigated in the Zagros are examined in the final section of the case studies (section ‘Fire as destructive force’).

This fire-centred study, therefore, like that of Clark and Yusoff (2014), aims increasingly to integrate ecological and social theories and approaches in order to examine more fully the interdependent environmental and socio-economic agencies, contexts and considerations influencing energy and carbon flows, and human access to, use and management of these.

Contextual and experimental approaches to fire: Linking theory, method and data

Central to all theories, and the approach developed in this paper here, are contextual approaches that examine the general, particular and specific context of agencies, phenomena and materials and their properties and life-histories, as developed in material culture studies more widely (Robb, 2010). Studies of the post-depositional context, taphonomy and preservation of fire traces and materials are of especial importance to archaeology, as understanding of these is fundamental to all subsequent analyses of the temporal and spatial associations and significance of fire. Major contributions to our understanding of archaeological traces of past fire are currently being made by experimental and ethnographic studies as well as developments in micro-contextual analytical techniques, including micromorphology, which is applied here. Mallol et al. (2007, 2013) in particular, have highlighted ways in which the best preserved traces will often be those that that were rapidly buried as they have often been less subject to post-depositional disturbance and alteration.

Early Holocene environment and vegetation biomass: Fire incidence and fuel sources

With regard to the role of natural and anthropogenic fire in landscape modification and niche construction (section ‘Interdisciplinary theory in fire ecology and fuel selection’; Zeder, 2012), climate, environment and vegetation are fundamental parameters in the nature and incidence of wildfires and anthropogenic fire use and fuel availability, and are in turn impacted by fire, as investigated here.

Many Early Holocene sites in the Zagros were situated on ecotone boundaries in environs that provided access to diverse resources and watercourses or springs. Tree pollen, however, only represents <5–15% of taxa from Lake Zeribar cores in the Early Holocene, at 1000 m a.s.l. This was initially interpreted as indicating that the climate was colder and drier in the Zagros, and less hospitable to occupation than other regions (Hole, 1996; Roberts and Wright, 1993; Stevens et al., 2001). However, more recent palaeolimnological and stable isotope evidence from Lake Zeribar and Urmia suggests that the Central Zagros region was wetter than today in the Early Holocene (Stevens et al., 2012; Wasylikowa and Witkowski, 2008). It is likely that tree cover and wood resources were more extensive than these low pollen values imply. Some of the tree genera identified are likely to be under-represented in the pollen sequence, as pistachio is a poor pollen disperser and almond is insect pollinated. Oak pollen is also sparse (<5–10%) in this and the wider Irano-Anatolian phytogeographic region during the Early Holocene. Thus, the scarcity of tree species probably also reflects other factors, including greater distance from Mediterranean refugia of the Younger Dryas, and habitat competition from native browsers and humans as well as grasses (Djamali et al., 2010; Roberts, 2002). Browers such as wild goat were present in the upper Zagros in the Early Holocene and there is increasing evidence of human settlement in the region from at least 11,800 BP (R Matthews and Fazeli, 2013; R Matthews et al., 2013) and of wood-cutting for fuel and timber (W Matthews et al., 2013; Roberts, 2002; Willcox, 1990). Pollen data indicate that there was a rapid increase in grasses from <20% of the assemblage c. 12,000 BP to 50% by c. 10,500 BP, which is significantly higher than values observed today (<20%) and sufficiently abundant to have posed a risk to young oak growth and been less susceptible to seasonal drought (Asouti and Kabukcu, 2014: 168).

Fire is likely to have played a role both in the restriction of trees and in the spread of grasses as fire favours the rapid regeneration of grasses and shrubs over trees (Asouti and Kabukcu, 2014: 169). A rapid increase in landscape fire is attested from 11,800 cal. BP by a spike in microcharcoal from burnt grass awns in Lake Zeribar cores, as in other regions (Turner et al., 2008, 2010; Wasylikowa et al., 2006). This burning may have been due to wildfires from lightning strikes during rapid climate warming in this storm-prone region that is likely to have increased seasonal growth and then drying out of biomass (Lewis, 1972). It may also have been due to human fires, either accidental or deliberate to create open grassland for hunting and early collection and cultivation of large-seeded grasses (Roberts, 2002) in niche construction (Zeder, 2012), as human settlement is also attested from c. 11,800 BP in the region (R Matthews et al., 2013). A decrease in grasses from 10,500 BP is coincident with the appearance of Plantago lanceolata associated with disturbed ground and an increase in human settlement and evidence for goat herd management from at least 10,000 BP (R Matthews and Fazeli, 2013; R Matthews et al., 2013).

It is likely, therefore, that the Early Holocene environment was influenced by a range of climatic, topographic, biogenic and anthropogenic agencies that included anthropogenic fire and early settlement and agricultural and land management practices, as in other regions (Marlon et al., 2013; Roberts, 2002). Lake core data suggest available fuel sources in the lower Zagros mountain region at Zeribar included tree and shrub wood, leaves and nuts, grass, reed and sedge stems, leaves and roots, and other steppe and steppe-forest plants including legumes. Cores from Lakes Urmia, Mirabad, Lalabad and Nilofar suggest the vegetation in these regions was similar, although there were more Chenopodiaceae in lower more southerly regions of the Zagros at Mirabad (van Zeist, 2008).

Fuel resources and combustion properties

Selection and use of fuels, however, are not directly correlated with environmental abundance nor distance from source, and are also influenced by and can inform on ecological and social strategies and practices, as discussed more widely, above (section ‘Interdisciplinary theory in fire ecology and fuel selection’; Asouti and Austin, 2005; Asouti and Kabukcu, 2014; Charles et al., 2014).

Plant remains identified on archaeological sites in the Zagros are attested not only as charred remains but also as pollen (Leroi-Gourhan, 1969); phytoliths (Rosen, 2003), impressions of plants, phytoliths, melted silica and calcitic plant ash (Figure 2; W Matthews, 2010; W Matthews et al., 2013, 2014). From this evidence, a range of fuel types have been identified at Early Holocene occupation sites in the Zagros, including trees, shrubs, reeds, grasses and ruminant and omnivore coprolites.

OPEN IN VIEWER

Figure 2.

Large interior oven with grass and reed fuel, burnt at temperatures up to >850°C (see Canti, 2003): articulated phytoliths, melted plant opal silica and calcitic plant ashes. (Plane polarised light (PPL)). Building 5, Bestansur, Iraq.

Ecological variation in combusted materials and fuel

There is some evidence that selection of fuels varied according to local environments, precipitation gradients and resource availability. Wood was more commonly burnt on archaeological sites in the highland areas than the piedmont and steppe where charred wood is almost absent and current precipitation is below that capable of sustaining oak, for example (Djamali et al., 2010). In highland sites, charred wood is present at Sheikh-e Abad from 11,800 BP and comprises 85–98% of the charred plant assemblages c. 10,000–9500 BP at Ganj Dareh (Van Zeist et al., 1986) and Abdul Hosein (Willcox, 1990), though trees are poorly represented (<15%) in pollen sequences from Zeribar. In lower piedmont sites charred wood is also present from at least c. 9000 BP at Jarmo (Helbaek, 1960) and Shimshara (see Figure 5). The presence of oak wood in considerable quantities at Jarmo suggests that oak may have recolonised these lower altitudes by this date (Helbaek, 1960; Watson, 1983) and/or was a preferred source of wood. By contrast, charred wood was sparse at sites in the piedmont-steppe region and absent at sites in the lowland piedmont-steppe and plains of Deh Luran and Khuzestan, at Ali Kosh and Chogha Bonut (Helbaek, 1960; Miller, 1996, 2003).

Reeds and grasses were used as fuel across all Zagros ecological zones in the Early Holocene. New phytolith and micromorphological analyses indicated that they were one of the principal sources of fuel in some lower Zagros, piedmont and lowland sites, including Jani (W Matthews et al., 2013; Shillito et al., 2013), Bestansur (Figure 2), and Chogha Bonut (Rosen, 2003). The abundance of reeds at these sites suggests wetlands may have been more extensive in the Early Holocene (Helbaek, 1969; Hubbard, 1990: 218; Rosen, 2003: 131).

Dung has also been identified as a common choice of fuel in all regions from at least: 10,000 BP in the upper and lower Zagros by new integrated micromorphological and GC/MS analyses (Figure 3; Shillito et al., 2013; W Matthews et al., 2013, 2014); 9660 BP in the piedmont zone at Bestansur (Elliott et al., 2015) and 9500 BP in the piedmont-steppe at Ali Kosh (Miller, 1996; Miller and Marston, 2012).

OPEN IN VIEWER

Figure 3.

Early ash layers in Trench 1: (a) field section profile (scale = 2 m); (b) resin-impregnated thin-section; (c) burnt aggregates, omnivore coprolites and bone (PPL). In situ burning and ashy deposits in Trench 2: (d) field section profile (vertical scale = 1.5 m, horizontal scale = 50 cm; (e) burnt deposits with nut shells, calcitic plant ash, charred plant remains and animal dung fuel (PPL); (f) fired-clay figurine.

Socio-economic context of fire use: Early sedentism and enclosure of fire

Fire-related activities were one of the foci of early visits to many Early Holocene Zagros sites, both before and after the development of agriculture c. 10,200 BP, and were thereby central in developing associations with place at many sites that later became longer-lived settlement mounds, and in creating and representing social relations (section ‘Interdisciplinary theory in the socio-economic context and specific uses of fire’; Bloch, 2010; Clark and Yusoff, 2014). Ethnographic studies by Watson (1979) in the high Zagros highlight how specific stone-constructed hearths are repeatedly returned to each year by individual family groups in mobile pastoralist communities, as the focus of their summer camp for food processing, cooking, light, heat, protection from predators such as wolves and socialising. In another ethnographic example from the Zagros, Arshi and Kasraian (1994/2001), observed that the fundamental household unit, a nuclear or extended family, is called the ojaq, ‘oven’, in semi–nomadic communities that have a village-base. During migration to highland summer pastures, these households coalesce to form larger social groupings, the obah, but when the group returns to their winter houses the larger grouping dissolves and most decisions are made within the household ‘oven/hearth’ group. Whilst these modern examples cannot be directly transferred to Early Holocene societies, they provide some insight into ways in which fire is a key element and phenomenon in social relations (see Atalay and Hastorf, 2006; Clark and Yusoff, 2014; Hastorf, 2001; Nelson, 2010), and how there may have been periodic fission and fusion of social groupings, as suggested for other Upper Palaeolithic/Neolithic communities (Roberts and Rosen, 2009; Wengrow and Graeber, 2015).

The earliest visits to Early Holocene sites in the Zagros left evidence of: digging and occasional lining of pits, for storage and/or fires; accumulations of ash and fire-cracked stones probably used for grilling and roasting food and radiant heat; and debitage and chipped stone tools from a range of activities, including hunting and butchery (Figure 4).

OPEN IN VIEWER

Figure 4.

Ganj Dareh, Iran: (a) Neolithic pits with fire-cracked stones Level E (after Smith, 1975: pl. IIIa); (b)–(c) Level D burnt architecture and features (after Smith, 1972: pl Ic, 1990: figure 1). Jarmo, Iraq (d) Ovens and pits in communal areas (J-I 8-9, after Braidwood et al., 1983: figure 74). Courtesy of the Oriental Institute of the University of Chicago; (e) ovens within architectural units (after Braidwood et al., 1983: figure 51).

The scarcity of groundstone tools for food processing in many early levels suggests these sites represent shorter visits for management and collection of resources. Examples of sites with evidence of such periodic activities include in approximate chronological order from 11,800 to 8000 BP: Sheikh-e Abad, East Chia Sabz, Asiab, Jani, Ganj Dareh E, Tepe Abdul Hosein, Tepe Guran, Jarmo, Chogha Bonut and Sarab (SV5). Sites, such as Ganj Dareh Level E and Asiab were clearly used for hunting, collecting and perhaps management of seasonally abundant resources, including wild birds, land snails and clams, and social gathering and perhaps feasting. Ganj Dareh E is likely to have been occupied between the months of May and October, to collect Pistacia nuts and harvest and sow barley. These dry months are also a preferred time for collection of firewood in semi-arid environments as wooded areas are more accessible and drywood is more abundant (Asouti and Kabakcu, 2014: 12). At Jani and Tepe Guran, residues from the earliest fire-related activities are fragmented and mixed with natural sediment, suggesting less frequent deposition and greater exposure to erosion and thereby more intermittent visits rather than permanent occupation, as observed in ethnographic and micromorphological studies of temporary hunter-gatherer hearths (Jani Phase 1, Figure 5; Mallol et al., 2007). Traces of architecture have been identified in early levels of some sites, including burnt aggregates from possible fire installations and isolated fragments of construction materials at Sheikh-e Abad and Chogha Bonut, and short stretches of walls perhaps to provide a wind break for fires at Ganj Dareh E and Jarmo JII9-8.

OPEN IN VIEWER

Figure 5.

Sequence of traces of fire at Jani, Iraq: (a) Jani mound and environs; (b)–(c) reed and grass stem phytoliths and charred plant remains in stratified midden deposits (Phase 2, scale = 50 cm) and (d) dung fuel (cross-polarised light (XPL)); (e) charred dicotyledonous wood in sediment-rich deposits with fire-cracked stones (Phase 1).

The low mounds of ash from fires in these early sites (Figures 3, 5) may themselves have been viewed as representative of associations with place and of land rights, wealth, roles and status, including access to woodlands and animals for firewood and dung fuel, rivers and marshes for fish, clams, fresh water mussels and reeds, and land for nuts, large-seeded grasses, legumes and land snails, as observed in ethnographic and anthropological studies more widely (Bloch, 2010; Hodder, 1987; Ingold, 2000). Hayden (2009) argues that periodic feasting was one of the ways in which social relations, obligations and any inequalities were created and naturalised in many sites in the Neolithic, as also observed in Central Turkey (Twiss, 2008). Certainly, many of the Zagros sites were visited early in their history for collection and consumption of seasonally abundant resources that could have supported large social groups. In these early levels, fire-related activities were conducted in communal open area contexts, suggesting food processing and cooking facilities and roles may have been shared by several groups (Bernbeck, 2004: 144; Bloch, 2010). The repeated plastering and renewal of many of fireplaces, as at Jarmo JII9-8 and at Jani Phase 3, suggests specific fire installations may have been associated with and maintained by individuals or small groups within these communal areas such as the hearth-groups of seasonally mobile communities observed in ethnographic research in the Zagros that returned to the same fireplaces each year (Arshi and Kasraian, 1994/2001; Watson, 1979). This evidence suggests dual associations to both community and smaller social groups, as explored by Wengrow and Graeber (2015) with regard to Upper Palaeolithic hunter-gatherers.

In the later levels of many Zagros sites, by contrast, hearths and ovens were less frequently placed in communal areas, and were instead incorporated and bounded within architectural units (Figure 4), suggesting food processing and cooking were less communal and more segregated within individual social units. This change corresponded with alterations to the design of buildings from very small-roomed structures to buildings with larger, more rectilinear rooms that could have housed greater numbers of people and a wider range of activities, as at Ganj Dareh, Level B (Smith, 1990), and Jarmo J–I, 6b–d and J–II,5 (Figure 4; Braidwood et al., 1983: figures 41, 51). The increasing segregation and compartmentalisation of fire and other activities has been observed at a wide range of Neolithic sites across the Middle East and is argued to be part of wider socio-economic changes that emphasised the household as a unit of production and consumption (Bogaard et al., 2009; Byrd, 2000; W Matthews, 2012a) and arena for social relations (Baird, 2005). This greater segmentation of social groups and activities around fire installations correlates with increased sedentism and the adoption of agricultural practices in many communities. That many of these discrete architectural units had fireplaces highlights the significance of fire-related activities and the roles associated with them for these social groups, as observed by Bloch (2010) with regard to the continuity of social concepts and roles afforded by more permanent fixtures and fittings within houses at Çatalhöyük (W Matthews, 2005a) and more widely.

Food processing and cooking facilities

Some of the earliest facilities for fire use at many sites in the Zagros comprise open or pit fires associated with fire-heated rocks, which themselves provided a versatile local or portable source of radiant heat, as at Ganj Dareh E (Smith, 1975, 1990), Sheikh-e Abad Trench 2 (R Matthews et al., 2013) and Bestansur, c. 10,000–9600 cal. BP. Hot rocks could have been used for heating as well as food processing or cooking through: drying/smoking of foodstuffs by suspending these over fireplaces with rocks in situ; parching/toasting/grilling/griddling/baking by placing foodstuffs onto heated rocks either in the fire or onto other surfaces; roasting by placing rocks within pits – often covered; boiling/stewing/grease extraction by placing rocks directly into foodstuffs in containers of skin, wood, basketry or pottery (after c. 10,000 BP) (Atalay and Hastorf, 2006). Within the fire-pits with heated rocks at Ganj Dareh E, the presence of Pistacia nut shells suggests that heat was used to open shells and roast nuts to enhance access, preservation and taste, and/or that shells were used as kindling after consumption as they are highly combustible; and of barley grains suggests that heat may have been used to toast/parch grains to enhance ‘shelf-life’, storage and taste (Hillman, 1984; van Zeist et al., 1984: 219, 221).

Technological innovation in construction of pisé/mudbrick ovens, significantly enhanced management of combustion temperature, oxygen flow, duration and safety by providing a controllable contained environment for fire use, from at least 9750 BP at Ganj Dareh B. The plastered interior and exterior surfaces of these ovens would have generated radiant heat and enabled many of the food processing and cooking stages provided by hot-rock technology, as well as enhanced and added to these. The smooth plastered surfaces of constructed ovens could be more readily used without contaminating foodstuffs and cleaned and renewed over time, concomitant with as well as facilitating more year-round sedentary occupation. The discovery of quantities of charred domesticated and wild barley in a domed oven within a building at Tepe Guran suggests that these structures enabled larger-scale toasting/parching of grain than that permissible using hot-rock technology (Meldgaard et al., 1963; Renfrew, 1969). At Bestansur, a range of fire technologies were used including use of heated rocks and open fires in external areas, as well as large constructed and plastered ovens, within rectilinear multiroomed building complexes (R Matthews et al., in prep).

These fire technologies would have favourably impacted the nutritional value of foods, making them more digestible (Wrangham and Carmody, 2010). They would have increased the range of substances that could be eaten and stored as well as dishes and tastes, particular from newly domesticated and more palatable varieties of plants such as cereals and legumes, and at some point milk products – known in Turkey from c. 7950 BP (Evershed et al., 2008). As food is often central to identity as well as sensual pleasures (Hastorf, 2001) and there is some evidence for local variation in food crops across the Zagros (Charles, 2008), it is likely that sites may have provided different meals with which to nurture and entertain each other and their guests, and that these would have varied seasonally. Suggested Neolithic menus are provided by Atalay and Hastorf (2006).

Phenomenology of interior fires: The built environment and health

With regard to the phenomenology of fire and its impact on experience, decision-making and social relations discussed above (section ‘Interdisciplinary theory in the socio-economic context and specific uses of fire’; Sorensen et al., 2014; Tilley, 1994), the enclosure of fire within constructed buildings created sheltered environments and thereby itself facilitated early sedentism and year-round occupation, particularly during the extremes of cold and wet seasons in the Zagros. Problems of overheating of interiors in the summer would have been mitigated by use of open fires in exterior open areas or placing ovens in small rooms to insulate particular sections of buildings from heat generated, as at Jarmo J-I II 5, c. 8500 BP. Ovens may also have been constructed and used on flat roof tops, as attested by a rare example on a collapsed roof at Neolithic site at Çatalhöyük. These rooftop ovens were probably principally used in warmer months, as oven rake-out is periodically covered by accumulations of rain-wash/snow-melt sediment (W Matthews, 2012b).

Use of fires, hearths and ovens within buildings is also likely to have had an impact on human health as fire heat, smoke and particulates are a health hazard affecting respiratory tract and eyes and tissues (Christensen and Ryhl-Svendsen, 2015; Scott and Damblon, 2010). That smoke did affect interior spaces is attested by the accumulation of soot on multiple layers of whitewash on wall plaster fragments from Shimshara, 9th millennium BP (Figure 6; W Matthews, in prep). These layers of soot and whitewash closely resemble those from the Neolithic site of Çatalhöyük, where seasonal exterior and interior variations in soot accumulations have been identified corresponding with evidence for seasonal oven use (W Matthews, 2005a, 2005b, 2012b).

OPEN IN VIEWER

Figure 6.

Traces of soot accumulation on Neolithic wall surfaces. Shimshara, Iraq (a) field-section profile (scale = 50 cm; (b) resin-impregnated thin-section; (c) layer comprising aggregates of re-deposited wall plaster (PPL); (d) white wall plasters/washes and soot accumulations (PPL). Çatalhöyük, Turkey, Building 5 (e) white wall plasters/washes and soot accumulations (PPL).

Carbon residue on ribs of some individuals buried below house floors at Çatalhöyük suggest smoke impacted elderly more house-bound individuals (Molleson et al., 2005: figure 12.19). At Tepe Abdul Hossein in the high Zagros Iran, one fireplace had a flue to conduct interior smoke and thereby improve health and comfort for this Neolithic community (Pullar, 1990: plate 8). Air circulation within rooms, prevailing winds and landscape and settlement topography would have had an effect on fire performance and efficiency, including smokiness and draw, and may have influenced the location of fireplaces. The location of fire installations within buildings would also have been influenced by and affected interior illumination as well as articulation of activities and social relations. Fire is also a purifying agent capable of killing microbiological organisms and reducing the toxicity, harm and bulk of noisesome materials such as decaying organic materials and coprolites (Schiffer, 1987). Many components in refuse deposits examined from the Zagros are burnt (Figures 3, 5), highlighting both the value of sources of waste as combustible materials as well as fire in refuse management.

Fire and social roles and relations

The repeated plastering of hearths/ovens in the Zagros, as at Jani (R Matthews et al., 2013) and Jarmo (Braidwood et al., 1983), for example, attests to remarkable continuity in use. This practice supports Bloch’s (2010) suggestion that such fixtures, by their durable nature, provide enduring representations of key social concepts, roles and relations and enable these to be returned to and passed on, affording social stability and continuity (section ‘Interdisciplinary theory in the socio-economic context and specific uses of fire’). In the case of hearths and ovens, these roles are likely to include food processing and preparation, production of craft materials, social and cultural events and gatherings, and the social identities and relations that make these possible, as noted in ethnographic studies for more mobile communities (Watson, 1979). Other aspects of fire that contribute to a sense of belonging and to social memory are the smells and sounds of fire (Pawłowska, 2010; Sorensen et al., 2014), whether evocative of the home, a particular social gathering or catastrophe.

Transformation of materials: Technology and ritual

One of the principal effects of the thermodynamic energy of fire is transformation of material properties. In some societies these transformations may also be considered ‘magical’ and used to mark and enhance particular places or objects (Hastorf, 2001). Fire energy was increasingly harnessed in the creation and elaboration of human material worlds during the Neolithic, as well as in crafting more durable construction materials such as fired-lime, and artefacts such as ceramic vessels and figurines, in its wider potential as a cultural agent (Clark and Yusoff, 2014; Pyne, 1997; Wertime, 1973) and as a source of energy in technological choices (Sillar and Tite, 2000). In Southwest Asia, one of the first materials markedly transformed by fire was fired-lime, c. 12,000 BP, which once rehydrated hardens to provide a rigid waterproof and highly durable material (Goren et al., 2001; Kingery et al., 1988). Production of fired-lime requires large quantities of fuel and sustained temperatures >750–850°C, and thereby represents considerable investments of energy as well as pyrotechnological skill, which was shared across Neolithic communities along exchange networks in Southwest Asia (Özdögan, 2002). In the Zagros, fired-lime or gysum was used for containers at Ganj Dareh A, c. 10,000 BP. It was also used to surface two large elaborate buildings at Tepe Guran (Levels J and E) c. 9000–8600 bp with a distinctive ‘terrazzo’ flooring, and probably mark their social significance as at Göbekli Tepe, Nevali Çori, Çayönü, in SE Turkey and Jericho in the Levant.

The production of fired-lime was initially argued to have been one factor in the environmental degradation that lead to the collapse of Neolithic sites in the Levant (Garfinkel, 1987), although climate and social factors are now considered more likely. Meldgaard et al. (1963) suggests that traces of a white powder in a fire-installation from Guran may have been from fired-lime production, but no analyses like those at Çatalhöyük have yet been conducted (see Anderson et al., 2014; W Matthews, 2005b; W Matthews et al., 2013). In the Zagros, fire was also used in modification of bitumen to affix flint in implements such as arrows or sickle blades and cowry shells, probably to skulls as eyes (R Matthews et al., in prep).

Fire was a key element in the innovative production of ceramics across Southwest Asia and the Zagros in the Early Holocene (Boivin, 2008). The lining of ovens was fired during use and manufactured from specially selected alluvial sources with abundant sand size minerals of quartz and feldspar that absorb thermal shock, attesting specific technological choices, as at Çatalhöyük, Turkey (W Matthews, 2005b). Fired and non-fired-clay figurines of human and animals are found at many Epi-Palaeolithic–Early Neolithic sites in the Zagros, including Karim Shahir, Asiab, Ganj Dareh, Ali Kosh, Sarab and Jarmo, where >5500 have been found. The hard-fired ceramic figurines would have facilitated repeated handling either in ritual practices or by children as symbolic objects, ‘toys’ (Figure 3). Some figurines were only fired at low temperatures, are more schematic or remoulded and often occur in secondary discarded contexts (Braidwood et al., 1983), as at Bestansur (R Matthews et al., in prep). In these cases, fire may have been perceived as potent energy and incorporated into acts such as wish magic, supporting suggestions by Morales (1990) and Daems (2008). ‘Clay’ tokens were also variously fired and repeatedly handled and may have been used symbolically and/or as accounting devices or gaming pieces in these communities.

Use of fire as a potent focal point and energy in itself is suggested by micro-contextual analysis of the occasional presence of red pigment in ashes, as at Jani (W Matthews et al., 2013). The colour red is generally considered cross-culturally to be symbolic of life (Tacon, 2004), and was used throughout Southwest Asia in the Neolithic, particularly in burials and wall paintings. At Jarmo red pigment was apparently randomly scattered (Braidwood et al., 1983). At Jani, micromorphological analyses reveal that flecks of red pigments are associated with floor sweepings in layers of charred reeds and grasses. Traces of charred reeds and flecks of red pigment were also identified scattered across a floor with four wild male goat skulls, one smeared in ochre, and a wild male sheep skull in what is probably a ritual building (Building 2) at Sheikh-e Abad c. 7600 bc (R Matthews et al., 2013).

Fire as destructive force

There is considerable scope for fire-forensic investigations in the Zagros (section ‘Interdisciplinary theory in the socio-economic context and specific uses of fire’). At Ganj Dareh, Level D, c. 9950 BP, an entire level of buildings was destroyed by fire (Smith, 1990). At Abdul Hosein at least one building in 12-H was catastrophically burnt, trapping a crouching male, female and infant (Pullar, 1990: 61, plate 12). The bodies themselves were not burnt and cranial injuries suggest they may have been crushed by falling debris. The site lies close to one of the major Zagros fault lines so it is possible that the fire occurred after an earthquake, as suggested by Berberian et al. (2012). At Shimshara, at least three building levels (16, 13 and Figure 7) were destroyed by fires with temperatures in some areas high enough exceptionally to partially melt and crack stone artefacts (Mortensen, 1970: figure 44), suggesting attainment of temperatures rarely encountered on other Neolithic sites. The cause of these fires is currently unknown, but they may have been sparked and/or fanned by the high winds that currently periodically funnel through an adjacent gap in the mountains cut by the Lesser Zab river, <4.7 km, distant to the northeast.

OPEN IN VIEWER

Figure 7.

Burnt Neolithic floor surface with overlying phytoliths, Shimshara, Iraq: (a) field section profile (scale = 50 cm); (b) resin-impregnated thin-section; (c) lenses of burnt plant remains on burnt floor surface (PPL); (d) articulated phytoliths and charred plant remains (PPL); (e) articulated phytoliths and charred plant remains autofluorescence (incident light fluorescence (IFL)).