Abstract

For at least the last half-billion years, fire has had a life of its own as a combustive interaction with fuel, air, and humidity. In its much more recent engagement with humans, fire has become a transformative technology, the control and spread of which are intertwined with human intentionality and innovation. Our earliest human ancestors approached fire deliberately, using it at first to improve comestibles such as meat and tubers and to heat-treat stone for flaking and later mastering the control of fire for pyrotechnical processes of ceramics, glass making, and metallurgy. Humans’ interactions with fire paralleled their creation of numerous flammables such as textiles and architecture, resulting in increased niches for the agency of fire through both accidental combustion and the use of incendiary devices in warfare. Although humans’ sponsorship of many intentional niches for fire suggests that it is a controllable entity, its inherent capriciousness still makes it a perpetual source of risk.

Fire is a natural phenomenon whose existence long predates the emergence of the human species. Started by processes such as lightning strikes or at the edges of lava flows when molten rock meets flammable organic material, natural fire is propagated only under accommodating conditions. Fire’s interplay with constantly shifting elements of fuel, aridity, and wind give it the capacity to change local pockets of land or to engulf entire landscapes, a scalability that is unleashed at every instance of ignition (figure 5.1). In cultural contexts, fire is both a welcome addition to the human technological repertoire and a sobering phenomenon when it escapes uncontrolled, as illustrated by the massive wildfires that are increasingly prevalent today and that result in billions of dollars in damage and the loss of human and animal life.

The natural history of fire indicates that, like earthquakes and hurricanes, it is a phenomenon that has been part of the Earth’s system for millions of years. The earliest primitive plants are discerned in the Ordovician period 450 million years ago (Rubenstein and Vajda 2019), constituting the first organic matter whose dried-out mass would have been susceptible to combustion. The advent of the first trees—and the first forests—in the Devonian period starting 393 million years ago represents what William E. Stein and colleagues (2020:1) have called “a turning point in Earth history, marking permanent changes to terrestrial ecology, geochemical cycles, atmospheric CO₂ levels, and climate.” The close tracking between the availability of vegetation and climate fluctuations culminated in numerous periods of heightened temperature in which there was increased burning of biomass recorded in ice cores, including during the Paleocene-Eocene thermal maximum 55 mya, in which the Arctic terrestrial mean annual temperature was 21˚C (Denis et al. 2017).

In the past as in the present, there are multivariate factors that make it difficult to accurately predict the directionality and ferocity of any particular fire or the rapidity of healing any particular burnt landscape. It also is difficult to directly correlate long-term effects because specific episodes of climate change do not always lead to predictable conditions for vegetation: although cycles of warm climate can increase the dryness of some species, rendering them vulnerable, cycles of increased precipitation also can result in more biomass to burn. Different species of plants recover from fires at different rates, which along with factors of moisture and temperature results in differential vegetation renewal after fire events. Vegetation itself responds to cycles of climate on both the short and long term, resulting in shifts in the types of plants and their abundance. The overall conditions that result in vegetation being fire-prone are additionally overlain by atmospheric loads of carbon, in which higher levels of CO₂ are correlated with an increase in lightning strikes that cause fires (Denis et al. 2017).

The complex interactions of fuel, climate, wind, vegetation growth, moisture, and species types provided the conditions for fires’ dynamic engagements throughout time. These interactions include obligate co-dependencies, such as plants that require fire to propagate or the cyclical opportunities afforded by the clearance of larger plants that allow other forms of vegetation to establish themselves in fire-cleared land, as seen, for example, in the establishment of savanna grasslands in sub-Saharan Africa 5–6 million years ago (Clark and Harris 1985). All of these developments long preceded the first human use of fire starting as early as 1.9 million years ago, which is indirectly suggested by the evidence for our ancestors’ occupation of cold climates and for the reduced dentition resulting from the use of intermediate mechanisms of food processing such as cooking (Wrangham 2009; Wrangham et al. 1999:567). The essential function of cooking is further supported by work by Karen Hardy and her colleagues (2015), who propose that the cooking of starches was the only mechanism that could have produced the glucose needed for rapid infant brain development, high-energy hunting, and overall hominid health in a genetic trajectory starting 2 million years ago and accelerating after 800,000 years ago.

At present, the most robust direct evidence for deliberate fire use comes from the site of Gesher Benot Ya’aqov, dated to 790,000 years ago (Alperson-Afil and Goren-Inbar 2010; for discussion of this and other early sites of human fire use, see Roebroeks and Villa 2011; Twomey 2013; Marshall 2020). Fire enabled our ancestors to achieve mastery over other predators, to live in climates colder than the human body will otherwise tolerate, to modify environments at a large scale through the systematic burning of vegetation, to alter the physical qualities of materials such as food and stone, and to enhance a sense of “home” long before the development of permanent architecture (e.g., Hardy et al. 2016). The first fires used by our human ancestors were most likely scavenged from natural fires ignited by lightning or spontaneous combustion and curated as glowing embers that could be magically revived through the skilled addition of fuel. Eventually, people also learned to create fire at will, a process of invention that still did not come with automatic control over the process or prevent the dangerous escape of flames when handled carelessly.

Humans Approaching Fire

Contrary to popular thought, humans are not unique in their ability to approach fire without fear. Some years ago, J. D. Clark and J. W. K. Harris (1985:19) made a case for a very long association of hominids with fire by noting that other animals ranging from birds to mammals will gather near smoky fires to dislodge insect pests; this factor, along with the propensity of grasslands to burn periodically, led them to propose that it “would not be unreasonable to assume that early hominids had learned to live with natural brush fires and so were not particularly afraid of them.” The subsequent accommodation of humans to fire can be parsed into specific types of actions that reflect increasing deliberation and control:

Following the logic of Clark and Harris, habituation would have enabled early hominids to view fire as a predictable component of the landscape, making fire no different from other resources. Fire might sometimes be a point-specific resource similar to an outcrop of stone suitable for making tools; at other times, fire might be prey-like in its mobility and unpredictability of location. The deliberate use of fire would have been the next stage of human interaction. Because fires occur in nature through lightning strikes and lava flows, it is likely that our hominid ancestors’ first encounters with fire would have occurred in the course of scavenging fire from already burning materials available in nature (figure 5.2).

The next step, curation, would have involved significantly more cognitive interaction and capacity for planning, as well as an understanding of the specific physical interactions of fuel, air, and humidity that enable fire to be prolonged (see Twomey 2013). The most comprehensive type of control, encompassing the manufacture of fire, would require the careful manipulation of stones or friction in combination with kindling. Because of the paucity of data, there is a robust argument about the timing in which fire was deliberately manufactured, with some researchers proposing a date of 700,000 years ago, others favoring a date of 300,000–400,000 years ago, and still others advocating a much more recent date of 250,000 years ago (summarized in Shimelmitz et al. 2014).

Throughout the stages of use, curation, and manufacture, fire was implicated in the domestication of human social life long before the development of architecture. As Dragos Gheorghiu (2007:3) has noted, a hearth is a sum of artifacts, operations, and activities (see also Hardy et al. 2016). Engaging with fire requires a working knowledge of the component parts of the process in which compensation might be required for gusts of wind, for the placement of materials to be cooked on a fire without snuffing out the flames, and for the effective management of sub-par sources of fuel such as flammables that were too green or too wet. The multiple and ongoing calculations of input can be evaluated through an analogy with modern technological processes to understand the importance of recursive human behavioral patterns. Multiple inputs result in a “complex technological system,” as characterized by Michael Brian Schiffer (2005a:486), consisting of “a set of interacting artifacts (in which) interactions among these artifacts—and people and sometimes environmental phenomena—enable that system to function.”

In Schiffer’s view of complex technological systems, two key aspects are encompassed in the development of new technologies. One of these is the concept of the “life history processes” of technology, characterized by “interrelated activities, which in turn incorporate one or more technological objects” (Schiffer 2005a:488). Another is the concept of the “invention cascade,” which is defined as the way people address the shortcomings of technological systems through innovation and invention (Schiffer 2005a:486). Both of these concepts can be applied to the human use of fire. Although fire’s basic components are exceedingly simple (air, organic material, and a combustion source mingled in an atmosphere of appropriate humidity), each component can be modified by human action and has a potential for unwanted results that must be mitigated. For example, the need to protect people and objects from accidental burning prompts the development of pits, stone banking, and other containment features. The need to avoid (or capture) smoke leads to consideration for the siting of fire within dwelling areas. The curation of fire requires transport devices; the use of fire requires ancillary technology such as tongs, sticks, props, and hearths; and the creation of fire requires the selection of appropriate materials for the generation of sparks or heat accompanied by the effective placement of kindling.

On the environmental level, the use of fire in a landscape was complementary to other human actions. Just as “landscape learning” is implicated in the collection of stone for tool making (see Rockman 2009), knowledge of the landscape was utilized by individuals for the identification of fuel sources, characteristics of the environment (such as seasonality) that affected the utility of fire, and the types of predators and prey that would be affected by humans’ fire use. In Australia, “fire-stick farming” by early human populations was a mechanism through which the landscape and humans co-adapted (Bliege Bird et al. 2008; Denham 2008; Jones 1969). Fire ecologists working in California have noted that the environmental record of fires shows a much higher frequency than would have occurred naturally, suggesting that human occupants substantially changed the ecology and environment of the region in the Indigenous era (Keeley 2002; see also Williams 2005). Although the agricultural period is usually associated with the physical clearance of vegetation using metal implements, fire would have been retained as a quick means of clearing brush and stubble (an observation raised long ago by A. Ghosh for India [1973]; see also Possehl [2002:43] for the use of fire as a pre-agricultural adaptation elsewhere in South Asia). Fire remains an essential component of slash-and-burn agriculture, in which successful burns of wild vegetation provide ash and nutrients for the support of cultivated crops (Fuller and Qin 2009:94; Padoch 1985).

The process of human habituation, use, curation, and manufacture of fire also entailed a number of social adjustments at the daily, lived level of human experience (figure 5.3). Indeed, most of the anthropological studies of fire have concentrated on its social components, such as the use of fire for symbolism and in art (e.g., Bentsen 2007; Jones 2007; Ronen 1998) and as an expression of cooperation and group interaction (Twomey 2013; Wrangham 2009). The habituation of our earliest primate relatives to natural fires may have facilitated a relatively rapid transition to use and curation once the mutually reinforcing feedback loops of starch consumption, group cooperation, migration, and cognitive capacity were established (see Hardy et al. 2015; Twomey 2013; Wrangham 2009). The emergence of regular cooking activities at a scale and predictability that would have had a selective genetic effect suggests that fire use became a routine part of human adaptations starting as early as 1.9 million years ago. The idea of a combination of predictable and unpredictable phenomena guided in part by a human hand also may have been the earliest transition to spiritual phenomena as people interacted with the unknowable or unseen to some visible effect. Fire use may have been the first physical manifestation of human agency upon spirit-world interaction, manifested in much later periods through durable evidence such as ornamentation (first appearing only 120,000 years ago) and burials (starting around 50,000 years ago; McBrearty and Brooks 2000).

Fire and Lithics Compared

As C. Karlin and M. Julien (1994) have noted, archaeologists have emphasized lithics as an entrée to early human cognition because the archaeological record preserves evidence of the entire process of manufacture, from core to waste flakes to a final product and subsequent retouching and resharpening. The manufacture of a stone tool requires a number of steps, including the identification and acquisition of suitable raw material, the use of energy in striking a stone to produce a sharp flake, and the use of the resulting tool(s) to cut meat, skin animals, or extract roots. Once struck from the parent rock, a stone flake cannot be reattached or resorbed back into the original mass. The production of lithics is often described through the rubric of the chaîne opératoire, in which stages of production can be identified because lithic manufacture is a glyptic process of reduction in which stone flakes are successively removed in the course of manufacture and use (Sellet 1993; see also Andrefsky 2008).

The archaeological record indicates that the blunt-force creation of lithics with a sharp edge went through several stages of development, resulting in increasingly elaborate tools. The current identification of the earliest stone tools at 3.3 million years ago consists of battered cobbles and flake fragments of what is called the Lomekwian tradition (Harmand et al. 2015). The next development consisted of Oldowan tools as early as 2.5 million years ago, which were simple cobbles with a sharp edge struck off as the result of free-hand knapping (Harmand et al. 2015:313–314; see also Karlin and Julien 1994; Toth and Schick 2009). The next substantial development was the Acheulian, a period marked most dramatically by the teardrop-shaped bifacial hand ax that first appears in the archaeological record as early as 1.65 million years ago (reported in Klein 2009:94). Lithics of a smaller size were subsequently developed for incorporation into composite and hafted tools (including projectile points and cutting implements) by 65,000 years ago, signaling the capacity of humans to understand a tool as something with both stable and disposable components (McBrearty and Brooks 2000:500).

However, lithic technology is limited as a proxy measurement of all early technological adoptions. One distinct characteristic of lithic manufacture is that stone is inert and does not behave independently of a human operator. The production of stone tools can only occur through direct and active human involvement when an individual chooses to apply energy to the raw material. This wholly human-initiated chaîne opératoire process means that the timing, speed, and amplitude of blows constituted a unidirectional series of activities; although stone has qualities such as hidden flaws that might prevent a desired result, it has no agentive capacity of its own. A slow process of stone tool creation also enabled the pace of learning to be spaced out according to the capacities of the learner, with the potential for the learner to carefully study the inert material between bouts of energy investment. Steps of the learning process appear to have included not only live demonstrations but also the existence of a demonstration toolkit that could be repeatedly viewed and handled by those who were learning the process (e.g., Karlin and Julien 1994:162).

Indirect evidence of fire use, as well as archaeological evidence, places the use and curation of fire within the same time frame as the development of stone tools, providing the opportunity to compare humans’ approaches to inert compared to agentive substances. Both stone and fire can be utilized in their scavenged state without further modification, a simple step that nonetheless includes aspects of scheduling and planning: how to carry the raw material, the path to take from raw material source to zone of use, timing the use to optimize the scavenged material, and how to adjust to the cessation of physical qualities such as sharpness (in the case of lithics) or heat (in the case of fire). Both lithics and fire are characterized by an understanding of unilineal cause and effect that is deeply engrained in our mammalian past, experienced in activities such as hunting as well as through irreversible bodily processes such as birth, death, and the daily realities of voiding bodily waste.

The thought processes encompassing the human use of both fire and lithics enable us to analyze the use of fire in the same way lithics have been studied, including the concept of the chaîne opératoire, the idea of raw material portability, the use of both basic and advanced techniques of working, the capacity to plan for desired outcomes, and the need to mitigate unpredictable outcomes. The material components of stone tool making and the use of fire would have been perceived as similarly governed by linear and irreversible processes: stone, once flaked, could not be reconstituted into an unbroken entity, and objects that were touched by fire could not be “unburnt” or food “uncooked.”

By contrast with the learning process of stone tool making, in which the natural element of stone is inert, the acquisition of knowledge about fire requires memory and predictive capacity because of the agency of the fire itself. Some rapidity of thought and adjustment to prevailing conditions are required on the part of the individual who uses and curates fire, which has behavior independent of human action depending on factors such as moisture and fuel flammability. Fire also interacts with other agentive natural forces, for example, when wind blows a fire out or out of control. Fire and stone tool making present different scales of danger to the operator as well. While a mishap in stone tool making can involve significant injury (for example, if a fragment enters the eye), most often the dangers to the maker are low. A stone tool production sequence that goes out of control usually results in nothing more serious than a cut or bruise to the hand and the waste of a piece of raw material. Even this aspect can be salvaged, with errors in stone tool making recycled into small or expedient tools or used “as is” for cutting and scraping tasks. By contrast, the consequences of fire being out of control are considerably more significant: injury or death of the human handler, burnt and uninhabitable landscapes, and destruction of animal and plant life.

The energy effects of fire can be vastly disproportionate to the energy expenditure of the human agent. Whereas the development of a stone tool is entirely dependent on a person, fire contains an inherent capacity for self-replication. In a controlled context, such as a hearth, some amount of human energy expenditure is required to supply the fire with fuel. However, in the outdoors, the simple application of fire to a suitable patch of vegetation means that all subsequent action is done by the fire itself, which makes fire an extremely efficient expression of human energy expenditure scaled up to the landscape level. Stopping a lithic production sequence is a simple matter of dropping the raw material, but stopping a fire almost always requires more expertise and energy expenditure than starting one.

Fire and Humans: An Intertwined Agency

Fire’s agency in natural contexts interdigitates with human agency of use. Compared to stone tool making, which serves as the standard measure of cognitive capacity and material engagement in ancestral time, the use of fire is a relatively easily learned skill. The period of apprenticeship in fire making is relatively short but allows individuals to gain experience and display expertise in commanding a socially useful technology. For most individuals, one or two demonstrations of the making and application of fire might be sufficient, meaning that fire is a technology that is simpler to make and use than even the most basic stone tool. Fire is scalable through the addition of flammable material, constituting a factor of management and control that could be easily grasped by users across the spectrum of age and dis/ability.

The diverse environmental, economic, and social uses of fire show how it was an all-purpose technology that could be controlled and manipulated by individuals of varying ages and levels of skill. As Clive Gamble and Martin Porr (2005) remind us, a focus on the individual in the hominid context constitutes a critical way of evaluating the development of human cultural diversity (see also Smith 2010). While many components of ancient technology are presumed to be the result of adult hands and adult thought processes (Karlin and Julien 1994:162), the utility and mechanics of fire would have been graspable even by young children. An increased interest in evaluating the role of children in the archaeological record (e.g., Baxter 2005; Crown 2007) means that we should anticipate all of the types of economic and social acts undertaken by young individuals in the past as part of their growth and maturity. From a young age, children could be instructed about fuel using the concepts of dead versus live vegetation, similar to the way children would be taught the relative utility and danger of dead versus live animals (cf. Barrett and Behne 2005). Children could be involved in the technology of fire both as fire tenders (responsible for adding fuel at appropriate intervals) and as procurers of fuel from the surrounding landscape.

Other distinct categories of users who may have gained agency and empowerment through the use of fire would include the elderly and the disabled. As Joanna E. P. Appleby (2010) has observed, there has been little explicit engagement with the concept of the elderly in the archaeological record. In part, this is because life expectancy in the premodern period was relatively low, resulting in a small proportion of individuals reaching old age; however, as the literature on “grandmothering” has illustrated, older individuals became increasingly represented in human communities over time and had a distinct sociobiological niche (Caspari and Lee 2004; O’Connell et al. 1999). The study of disability is similarly poised for greater consideration by archaeologists (see, e.g., Sneed 2020; Southwell-Wright 2013; Tilley 2012). The use of fire would have enabled individuals who were afflicted by physical limitations to display virtuosity in meaningful skills; at the same time, mishandling fire in ways that facilitated its escape and destruction of lives and property would have created or reinforced social stigmas against already marginalized members of the community.

Just as it provides equal opportunities for social actions within a group, fire is also a democratic means of destruction. A flame wielded by a child or a physically disabled person is as damaging and unstoppable as a fire propagated by the most able-bodied warrior, and the destructive capacity of fire rendered it an increasingly potent tool of violence over time. For foragers, the specter of a landscape damaged by accidental or badly managed fire was surmountable, as long as individuals escaped the flames alive. By contrast, many of the things in which agriculturalists make social and economic investments are inherently flammable, such as houses, corrals, textiles, baskets, standing crops, and stored food grains. In addition, many meaningful ritual and religious objects are made of perishable material such as wood, all of which can be destroyed in either accidental or purposeful fires. As Ruth E. Tringham (1991) has eloquently shown, the finality achieved by burning provides an irreversible phenomenological moment for actors and spectators.

The use of fire as a destructive mechanism gained momentum after the inception of agriculture (in many parts of the world, at the beginning of the Holocene 10,000–12,000 years ago). The archaeological record illustrates how fire was used in punitive raids and small-scale warfare, for example, in the American Southwest where researchers working at the site of Burnt Corn recovered large quantities of what would have been edible grain, signifying the catastrophic loss of provisions through intentionally set flames (Snead 2015). In additional to punitive effects, destruction through fire could also constitute a form of cleansing or purification on a large scale. In Neolithic Europe, for example, structures appear to have been destroyed by fire prior to being rebuilt (Bradley 2005; Dufraisse 2008). Historical accounts also illustrate large-scale purification through fire at ceremonies such as the “busk” or “New Fire” annual rites (e.g., James 2000; VanDerwarker et al. 2007). Many other forms of ritual purification through fire also existed, from the use of fire to cremate the dead (Kuijt et al. 2014; Oestigaard 2000) to the decommissioning of structures (e.g., Baltus and Wilson 2019; Tringham 1991) to the creation of aromatic smoke as an olfactory perimeter for ritual activities (Kolb and Murakami 1994).

The archaeological record of massive fires in sites associated with complex societies shows the regularity with which fire was utilized as a cheap, effective way of waging war. Purposeful conflagration resulted in thick lenses of ash and charcoal far greater than the occasional out-of-control household fire (what Possehl [2002:49] calls “the unhappy grist of daily life . . . considerably erased by the process of cleaning up the mess and rebuilding”). Throughout the world, major archaeological sites were decommissioned by fire in antiquity: Hazor in Israel (reported in Lev-Tov and McGeough 2007:89), Kerkenes Dağ in Turkey, Pylos in Greece, Ebla in Syria, Fishborne in Britain, Kot Diji in Pakistan (Possehl 2002:49), and Aguateca in Guatemala (Inomata et al. 2001), to name a few. Fire continues to be a destructive force in the present, whether in the form of wartime actions (such as the fire bombing of many cities in World War II) or of anarchic individual expressions of arson.

Conclusion

Fire has characteristics that are shared by other contributions to this volume: like monsoon rains and animals, fire is a natural phenomenon that humans can attempt to predict, monitor, and manage for their benefit. Like vegetation and diseases, fire responds to anthropogenic niches as a source of new loci in which it can thrive. Like hurricanes and earthquakes, fire can clear a landscape and enable humans to envision new configurations and new economies. The existence of fire is, however, also intertwined with concepts of risk at every instance of use, from the risk of burns sustained in campfires to the risk of widespread landscape destruction from a single errant spark.

As a combination of both predictable and unpredictable characteristics, one might argue that fire was the key transitional technology in the shift from simple, linear activities such as lithic production to all subsequent technological systems. Fire and its management represented both an essential component of cognitive development and the basis for pyrotechnologically informed innovations such as metallurgy, lime slaking, ceramic production, and glass making in the early sedentary period starting 10,000–12,000 years ago (Miller 2007; Roberts and Radivojevič 2015) and the development of electrical devices, internal combustion engines, and steam-powered technologies in the modern era (Schiffer 2005a, 2005b). Indeed, the global effects of fire wielded by human hands and as a result of human land-use changes provide an impact that is so significant that S. J. Pyne (2020:1) has suggested that the entirety of the last 2.5 million years should be termed the “Pyrocene.”

In our interdigitated wild and cultivated landscapes, the boundary between “natural” fire and “cultural” fire has become blurred. The complex feedback loop of human and fire interaction is similar to the mutually constituted realms of agency discussed throughout this volume, including the realm of animals (Ammerman; Bishop; Quintus et al.; Tomášková), vegetation (Dine et al.), and diseases (Juengst et al.). While the mechanics of fire have not changed over the past half-billion years (there is still only the basic configuration of a source of ignition applied to fuel), natural and human activities have become accelerated and intertwined. Neither the agency of fire nor the agency of humans has been superceded; rather, each agent has recourse to more opportunities of engagement through recurring episodes that began the moment “a fire-wielding species met a fire-receptive world” (Pyne 2020:1).

Starting with the first experiments with natural fire and continuing into a closer and closer relationship with curated fire and created fire, humans developed increasingly diverse approaches to flammability. Fire also gained increasingly diverse targets because humans created more things to burn, including agricultural fields, organic artifacts, and architectural structures. In more recent times, combustion has gained new opportunities through humans’ distillation of hydrocarbons into kerosene, gasoline, and rocket fuel; the creation of more technologies that throw off sparks, such as internal combustion engines and power lines; and the use of chemicals incorporated into buildings and furnishings that serve as unintended accelerants and increase the potency and heat of unintentional fires. Most sobering of all, humans have contributed to global processes of climate change that have resulted in forest fires being increasingly destructive and intense, as seen in the American West, Indonesia, Australia, and increasingly throughout the world.

Acknowledgments. I would like to thank Jed Docherty, Heather M.-L. Miller, Michael Brian Schiffer, and James Snead for much-appreciated comments on an earlier version of this chapter. Thanks also to Dennis Sandgathe for his interest in the topic.

References