1Kilns are characterised by the separation of the combustion and the firing chambers, and enable the transformation of clay material into ceramic mainly by convection, i.e., the transport of heat in a fluid by the movement of molecules (the draught). The oldest known vertical draught kiln with superposed chambers is dated from the second half of the 7th millennium BCE and comes from the Hassuna level of Yarim Tepe I (Merpert and Munchaev 1993b). There are few studies on the use and significance of pyrotechnology in prehistoric societies (Delcroix and Huot 1972; Majidzadeh 1975; Hansen Streily 2000). The evolution of firing techniques lacks documentation (Rehder 2000), and efficiency—considered to be a cost-effective functioning with homogeneous atmosphere and rise in temperature—has often been regarded as the main driver of firing evolution and pottery production specialisation (Van der Leeuw et al. 2009; Costin 2020). Reviewing the historiography of ceramic pyrotechnology studies is necessary to understand the biases at the origin of the documentation gap. Indeed, the absence of a standardised methodology has limited our understanding of firing structures, regarding both their technical functioning and their significance as socio-technical markers for studying the dynamics of cultural transmission at the origin of societal changes (Roux 2013; Frangipane 2015). Our aim is to overcome the widespread idea that kilns are designed only in response to specific types of clay material to be fired, climatic factors, geographical settings, or economic constraints (Hansen Strailey 2000).
- 1 Observations conducted by King Houndekpinkou in Aiwedji, Benin, show that while the professional po (...)
2Depending on how the potter manages the firing process, similar firing structure designs can produce varying ceramic outcomes characterised by different thermal profiles, and conversely, different firing structure designs can achieve similar ceramic outcomes (Gosselain 1992; Livingstone Smith 2001). The use of a firing structure design (e.g., superposed-chamber kilns, juxtaposed-chamber kilns, single-chamber firing structures, firing pits, open hearths) necessitates extensive learning of a specific set of gestures associated with that design, which is not easily transferable to other designs.1 This combination of gestures and design constitutes a firing know-how, transmitted by guided imitation through cultural dynamics (Roux 2013; Roux et al. 2017; Roux and Gabbriellini 2018). Considering that multiple firing structures can achieve almost any type of firing process, the choice of a specific firing structure design from among various alternatives is socially constructed and expresses the potter’s firing tradition (Killick 2004; Baldi 2023).
3In the millennia following the adoption of the superposed-chamber kiln in Southwest Asia, kilns exhibited a high morphological variability (Padovani 2023b). The form of a firing structure corresponds to an association of volumes, in and through which the heat circulates. Therefore, the form impacts the operation of the firing structure, by enhancing or moderating specific aspects of the firing process related to the technical design chosen by the potter, such as draught force, heating rate or thermal homogeneity in the firing chamber. This is particularly true for kilns, which are more architecturally standardised than hearths and single-chamber structures, in that the distance between the fire and the pottery to be fired cannot be modified once the structure has been built (Thér 2014). Therefore, we assume the kiln morphology expresses the architectural concretization of a particular way of firing, with specificities of the firing process related to a potter or a community of potters, directly integrated into the architecture of the firing structure. A detailed analysis of the design and morphological features of firing structures in context can help to reconstruct some aspects of the firing process and outline different communities of potters.
4This paper introduces a new methodology for studying firing structures, involving the distinction of technical design from morphology and a detailed architectural analysis. The approach is illustrated through a case study, including bibliographical data and very recent excavation results, on two Early Chalcolithic kilns: Kiln 345 from Yarim Tepe II (Upper-Tigris region), related to Halaf occupation, and Kiln 14 from Tell el’Uwaili (Southern Mesopotamia), related to Ubaid 1 occupation. Halaf and Ubaid 1 communities are recognised as two different cultural horizons, distinguished by their construction techniques, pottery styles, and socio-economic organisations (Stein and Özbal 2007; Mühl and Nieuwenhuyse 2016; Vallet et al. 2019; Frangipane 2020). Although both Halaf and Ubaid 1 communities used superposed-chamber kilns for pottery firing, kilns were probably more common in Ubaid 1 sites, whereas they were not widespread in Halaf sites, in which single-chamber firing structures were mostly employed (Padovani 2023b). We are expecting differences in the morphology of the kilns from Halaf and Ubaid 1 cultural horizons. However, what are they? And what insights can these morphological differences reveal about Halaf and Ubaid 1 respective firing processes? The kilns chosen for the study probably do not represent the full range of kiln models used respectively by Halaf and Ubaid 1 groups. However, they exemplify one archetypal form used by some communities of potters in the heartland of each cultural horizon. A quantitative analysis is not necessary for the comparison, as the aim is not to identify patterns and regularities, but to reconstruct firing processes in cultural context, based on detailed analysis of traces and clues found on firing structures (Ribeiro 2024).
5In addition, during the Ubaid 3–4 period (5300-4700 BCE), with the expansion of the Ubaid tradition in northern Mesopotamia, the morphological variability of the superposed-chamber kiln did not decrease, as it could have been expected, but rather increased (Padovani 2023b). It is currently admitted that the outward homogeneity of the Ubaid 3–4 pottery tradition resulted from several micro-regional borrowings between different pottery-making communities integrated into inter-village networks (Adams and Wright 1989; Stein and Özbal 2007; Stein 2010; Baldi 2016, 2022a; Baldi and Zingarello 2021). Therefore, we suppose the technical design and form of Ubaid 1 kilns were not transmitted as a package from southern to northern Mesopotamia communities. Rather, the range of kiln morphologies likely increased through the selective adoption, and integration into local kiln morphology, of specific architectural features, exchanged between potters of different communities using kilns.
6Early archaeological excavations in Southwest Asia focused on temples and palaces and the search for objects to fill European museums (Chevalier 2002), somewhat ignoring everyday crafting activities areas. Firing structures attracted little interest, and the lack of knowledge about their shape and material characterisation has prevented their discovery. Some archaeologists, such as Ghirshman (1938) in Iran, were personally interested in firing structures and documented the architecture of some selected examples. However, based on limited ethnographic observations and biblical accounts, the idea spread that pottery was manufactured on the edges of sites, as the kilns were considered to be polluting. From then on, the social status of potters was affected, and they were considered to be spatially and socially marginalised (Oppenheim 1948: 41).
7From the 1960s onwards, with the growing interest in the archaeology of techniques, firing structures attracted the attention of archaeologists but were treated as isolated objects, disconnected from their archaeological context and cultural frameworks. Two concurrent research directions emerged, each influenced by distinct intellectual movements affecting archaeological research. One trend focused on the general function of firing structures, whereas the other on their morphology.
- 2 Oxford English Dictionary
8Between the 1960s and the early 1990s, archaeometric analyses were performed to identify the maximum temperature reached during firing (Tite 1969; Goodyear 1971; Matson 1971; Morariu et al. 1977; Maniatis and Tite 1981; Perinet and Courtois 1983; Edwards and Segnit 1984; Klenk 1987; Kaiser and Lucius 1989; Rigby et al. 1989). Temperature data were primarily used to evaluate performance metrics and, consequently, to determine the level of technological advancement of societies employing specific types of firing structures. This trend was probably influenced by the popularity of the Anglo-American notion of “human capital” in economics, i.e., “the skills the labour force possesses regarded as a resource or asset”2. Within the context of this notion, the improvement of human skill is expected to amplify both productive capacity and general economic growth of the society, perceived as the ultimate goal of production. Humans are considered as rational economic agents focused on maximizing gains through their choices, while economic returns are the main driver of production and technical evolution (Kirchgässner 2008). This concept appeared first at the end of the 18th century in the work of Smith (2003), However, the concept gained popularity in social sciences from the late 1950s (Mincer 1958; Becker 1962, 1964). In this context, the detailed morphology of firing structures was often overlooked. Researchers focused on how protohistoric societies could achieve functional ceramic products without considering the firing structures’ specific characteristics or the potters’ agency. For instance, the kiln, which can reach high temperatures, was considered the most advanced stage in the evolution of firing structures. Only in the 1990s did experiments challenge the prevailing belief that temperature was a determining criterion for differentiating firing structures (Gosselain 1992; Livingstone Smith 2001).
9In France, in the 1960s-70s, an epistemological revolution took place in the field of techniques studies with the working group of the journal Techniques & Culture, formed around Cresswell (1976). The group explored the relationships between technical systems (comprising techniques, industries, and trades) and the evolution of socio-cultural organisations. It was influenced by the work of Leroi-Gourhan (1964) on the chaîne opératoire; the historical materialism of Marx and Engels (Marx et al. 2014), by focusing on objects and technical processes; as well as Structuralism (Lévi-Strauss 1949), which gained prominence in French institutions after May 68, by considering society as a general framework composed of multiple interacting systems. The working group Techniques & Culture advocated for the integration of all sciences to understand society as a whole and promoted the study of technology. In the 1980s, this anthropological approach of technology was tested by North American researchers (Smith 1982; Lechtman 1984).
10However, from the 1970s onwards, only a few studies unrelated to archaeometric analysis focused on firing technology (Delcroix and Huot 1972; Majidzadeh 1975; Hansen Streily 2000). These are still the most quoted references on Southwest Asia kilns, and should be examined in detail. Also, it is important to mention Crawford’s paper (1981) in which she discusses the challenges of identifying pottery firing structures at Abu Salabikh.
11Delcroix and Huot’s paper presents a description of the general firing process and a morphological analysis of 50 selected examples of firing structures from the Late Neolithic to Antiquity, excavated in Southwest Asia. The authors use two differentiating criteria: the depth of burial of the firing chamber in the soil and the relationship between fuel, gases and clay products to be fired (Delcroix and Huot 1972: 79). Based on this corpus and the criteria mentioned above, they propose an outline of the technological development of firing structures. They assume that a single technical principle was behind the transformation of pottery firing structures: the ability to increase the temperature by heat concentration. In other words, they presume all firing structures were part of the same “tree” and evolved from a simple technological stage, such as the open hearth, to a more complex one, such as the kiln with a buried combustion chamber (Delcroix and Huot 1972: 95, fig. 9). Nevertheless, the dating of firing structure types does not align with this evolutionary path. After the initial adoption of kilns, open hearth and pit firing continued to be used to a large extent (Padovani 2023c).
12Published three years later, Majidzadeh’s paper documents and compares 14 well-preserved examples of archaeological firing structures located in current Iran. He critiques the use of the depth of burial as a reliable typological criterion and employs ethnographic comparisons to propose reconstructions of firing structures. Due to limited data on kiln operation, he refrains from discussing technological evolution but emphasizes its non-linear nature.
13Hansen Streily discusses 89 kilns from Iran, Iraq and Syria. She suggests differentiating between technical and morphological features. However, she does not specify the elements included in these two categories. She validates Delcroix and Huot’s criteria of classification and assesses, on the one hand, that the primary purpose of a kiln is to reach high temperatures and, on the other hand, that the separation of fuel and pottery is characteristic of more advanced pyrotechnology, enabling better control of the firing process. She gives importance to the morphology and introduces the notion that the proportions of the kiln’s features are crucial to understanding how it operates. However, she does not provide an analysis of her dataset, based on proportions, technical characteristics or morphological features. She concludes with general considerations on the functional difference between single and double-chamber firing structures (Hansen Streily 2000: 79).
14As part of an exhibition at the Bochum Mining Museum, Boroffka and Becker (2004) proposed a new typo-chronological classification for Ancient Persia firing structures, building upon Majidzadeh’s morphological approach and incorporating the size criterion. They differentiate between single-chamber structures and five types of kilns: square-plan kilns, circular-plan kilns, oval-plan kilns with wall extensions, central wall kilns, and central pillar kilns (Boroffka and Becker 2004: 219, fig. 1). Their study demonstrates that no linear development can be traced over the long term, as the same kilns shapes were used in very different periods and regions. For instance, square-plan kilns were used spanning from the end of the 6th to the end of the 2nd millennium BCE, from the Susiana plain to the Kerman ranges. Classifying kilns by shape, without contextualizing them within cultural and socio-economic frameworks does not generate new data, but rather complicates the understanding of firing structure types distribution over time and space.
15In sum, this initial morphological approach was necessary to document the typologies of firing structures discovered at archaeological sites, overcoming the challenge of studying the remains of the humble practice of pottery firing, generally in poor condition. However, the interpretation of pottery firing structures has been limited by positivist approaches, and no clear method has been developed or tested to understand the anthropological aspect of these structures or why kiln technology merits further study.
16The Halaf cultural horizon is named after the site of Tell Halaf in northern Syria (Oppenheim and Schmidt 1943). This cultural horizon was composed of different communities inhabiting a territory spanning from the Mediterranean coast to the Zagros mountains (Nieuwenhuyse 2016) and covered a long period of time from 6000 to 5300 BCE (fig. 1). Halaf settlements were mainly agro-pastoral villages. In some of them, it is presumed that a part of the population was subject to frequent mobility within the surrounding areas (Bernbeck 2008; Pollock 2013; Nieuwenhuyse and Akkermans 2019). Single-chamber firing structures were widely used to fire pottery and only six Halaf superposed-chamber kilns have been excavated. In the Hamrin region, two kilns at Tell Hassan level 1 present a triangular shape (Fiorina 2014). In the Upper-Tigris region, we count one kiln at Tell Arpachiyah TT8 (Mallowan and Rose 1935), and two kilns at Yarim Tepe II levels 5 and 9 (Merpert and Munchaev 1993a). In the Jezirah, one kiln of oval shape with a large hole in the centre of the grate and an angled mouth, and two badly preserved and indeterminate firing structures, were excavated at Umm Qseir phase 1 (Tsuneki and Miyake 1998).
Fig. 1 – Map of Early Chalcolithic cultural horizons in Southwest Asia.
Vallet et al. 2020.
17Kiln 345 from Yarim Tepe II level 9 is the best preserved and documented structure. Despite the limited size of the Halaf kiln corpus, we assume that Kiln 345 is not an isolated model, since a very similar structure was found at Umm Qseir phase 1, less than 200 km west of Yarim Tepe II.
- 3 The kiln structure was given to the Mosul museum and can still be seen there (https://archeologie.c (...)
18Kiln 345 was excavated by the Soviet Archaeological Expedition directed by Munchaev in the 1970s3. Although radiocarbon dates for level 9 at Yarim Tepe II are not available, levels 8 and 7 have been dated, using this method, from the early to mid-6th millennium BCE (Campbell 2007). Hence, level 9 can be dated to the beginning of the 6th millennium BCE. If there is no documentation on the exact sherds found in this particular kiln, three different types of pottery are documented at Yarim Tepe II level 9 (Merpert and Munchaev 1993a): 40% of painted pottery (uniform firing, sand and grit temper, slipped surface), 55–60% of simple smooth pottery (uneven firing, sand and grit temper, slipped surface) and less than 5% of coarse pottery (poor firing, organic temper).
- 4 The attribution of Abada level 3 to Ubaid 1 is debatable. According to Vallet et al. 2020 (figure 1 (...)
19The Ubaid 1 cultural horizon is named after the site of Tell Ubaid in Southern Mesopotamia (Woolley 1924) (fig. 1). Excavated Ubaid sites are rare in the region as most of them have been covered by the extensive alluvial deposits from the Tigris and Euphrates rivers (Geyer and Sanlaville 1996). As a result, Tell el’Uwaili represents this horizon’s most extensively excavated area, primarily confined to the Mesopotamian basin (Forest et al. 1996). Recent excavations at Tell el’Uwaili, by the Mission de Larsa-Oueili directed by Vallet, revealed a large pottery production area with 44 firing structures (Baldi et al. 2020, 2022; Vallet et al. 2020). Radiocarbon dates are not yet available for Tell el’Uwaili and the Ubaid 1 phase (level 4) is roughly dated from 6000 to 5600 BCE. The firing structures of level 4 were built above the granaries of level 5 and reused their walls (Vallet et al. 2020). They are eroded, but most of them seem to have similar characteristics, which leads us to suppose that a specific type of kiln was widespread on the site. One of these specimens is Kiln 14, excavated in 2019, and located at the centre of the settlement, in a cluster of three similar structures. At Tell el’Uwaili level 4, coarse pottery hasn’t been documented. The large majority of sherds are hard and homogeneously fired and present very small mineral inclusions. Slip is relatively uncommon compared to the very frequent occurrence of painted motifs (Vallet et al. 2020; Baldi 2022b). Apart from Tell el’Uwaili site, the corpus of Ubaid 1 kiln is very limited. In the Hamrin region, three firing structures have been identified from the Ubaid 1 occupation at Tell Abada level 34 (Jasim 1983, 2021). One of these structures is identified with certitude as a superposed-chamber kiln, featuring an elongated firing chamber and a central pillar that probably supported a grate. The two other structures are circular but only their foundations are preserved, making it difficult to conclusively categorise them as kilns.
20During the Ubaid 3–4 period, material elements related to the Ubaid tradition in Southern Mesopotamia were present on a large territory expanding towards the Mediterranean shores. This expansion of the Ubaid tradition is characterised by the hybridisation of technical processes in pottery manufacturing at several sites in Central and Northern Mesopotamia (Breniquet 1991; Akkermans and Wittman 1993; Stein and Özbal 2007; Campbell and Fletcher 2010; Özbal 2010; Stein 2010; Vallet et al. 2019), as well as the homogenisation of pottery shapes and decorations (Frangipane 1996, 2015), and the standardisation of the spatial organisation of work in pottery workshops (Padovani 2023b, 2024). Nevertheless, the outward consistency of the Ubaid pottery tradition is now understood to result from multiple micro-regional exchanges among diverse pottery-making communities interconnected within inter-village networks (Adams and Wright 1989; Stein and Özbal 2007; Carter and Philip 2010; Stein 2010; Baldi 2016, 2022a; Baldi and Zingarello 2021). During this period, we enumerate 45 kilns spread on 10 sites. Single-chamber firing structures were still used, particularly in the Upper-Euphrates region. Furthermore, superposed-chamber kiln morphology was not standardised and its diversity increased compared to the Early Chalcolithic period, possibly indicating exchanges and adaptations of the firing techniques (Padovani 2023b).
21We agree with Delcroix and Huot’s observation (1972) that the correlation between fuel, gases, and clay materials is essential for distinguishing firing structures. However, the authors approach firing structures as a unified and indivisible whole, from a functionalist perspective focused on the product to be fired, and so does Hansen Streily (2000). We can resume this functionalist approach by asking the question: What did this kiln fire? This approach objectifies the kiln and limits the possibilities for further study. In addition, as demonstrated through experimentation, ethnographic observations and archaeology (Gosselain 1992; Livingstone Smith 2001; Doherty 2021), almost any ceramic results can be obtained with any type of firing structure, depending on the potter’s skills and tradition of firing. The question to ask is rather: How did this kiln fire? This question facilitates an in-depth examination of the role of each component within the firing structure, enabling an analysis of how each feature, with its distinct function and form, contributes to the overall operation of the structure in firing the clay material. By exploring this concept further, the work of Simondon (1958) on technical objects and machines, is particularly relevant to understand firing structures from an intrinsic point of view. Our approach proposes to consider the kiln as a technical object, focusing on the way it operates to transmit heat to the clay material, in relation to the potter’s action.
22A firing structure, like any technical object, is composed of various functional elements, each with its specific function (Padovani 2023a, 2023c). These elements are interconnected to achieve the overall goal of the structure, which is the firing of clay material. The way in which these elements are connected is essential to understand how the structure operates. In addition, these elements present a certain degree of independence and can be transposed into other technical systems (Simondon 1958). Instead of focusing on elements extrinsic to the firing structure (fuel, gases, and clay materials), our study concentrates on the architectural components of the firing structure that support the fuel and clay material, and facilitate the circulation of gases. A firing structure is composed of three types of elements, each with its own function: (1) elements that enable heat to be produced (such as the combustion chamber where the fuel and the hot gases burn), (2) elements that facilitate the circulation of heat (such as the mouth, chimney or flues), and (3) elements that allow the transfer of heat to the clay material (the firing chamber). The way these elements are interconnected constitutes a technical design. Two firing structures exhibiting different technical designs can both transform clay material into ceramic, thereby achieving the same overall function. However, these structures differ in their heat transfer mechanisms and require different types of expertise to operate effectively.
23The first step of the approach is to identify the technical design of the structure, i.e., to determine its functional components and their relative positions to one another. For instance, in a superposed-chamber kiln, the combustion chamber is situated below the firing chamber, and the mouth, flue and chimney are positioned to facilitate a vertical draught. Heat transfer in this configuration occurs primarily through convection. On the other hand, in a single-chamber structure, the combustion and the firing chamber are integrated. The direction of the draught, either horizontal or vertical, depends on the placement of the chimney. However, heat transfer occurs mainly through conduction, facilitated by contact with the embers.
24The form of the firing structure can modify the direction and the strength of the draught, as well as the distance over which the gases, produced by the breaking of the fuel molecule chain due to heat, are consumed (Rehder 2000: 16). These parameters impact the temperature rise and the distribution of heat within the kiln, thereby affecting the amount of heat transferred to the clay material depending on its position inside the structure.
25The second step of the approach is to document the morphology of each technical design. This documentation includes the measure and shape of the functional components of the structure (mouth, combustion and firing chamber, grate’s hole, chimney, etc.). The upper part of the firing structures is rarely preserved, forcing us to work in two dimensions rather than volumes. For instance, in a superposed-chamber kiln, the external shape of the firing chamber (e.g., circular, oval, rectangular, triangular, square, elongated) often forms the footprint of the entire kiln at ground level.
26The kiln is set in a shallow pit and preserved until the grate. The floor level associated with the kiln is estimated to be at the top of the circular cubicle attached to the kiln’s wall (fig. 2a). Records indicate that all the structures of Yarim Tepe II earliest levels were built with clay slabs (Merpert and Munchaev 1993a: 144). It is likely that the grate was built using clay slabs (fig. 2a), but no precise information is given for the kiln’s walls that measure around 40 cm in width. The combustion chamber is almost circular and has a depth of 1 m. Its external diameter varies between 1.9 and 2 m, whereas the inside diameter measures 1.15 m (fig. 2). The grate is not supported by a central pillar but by the walls of the combustion chamber forming an internal shelf (fig. 2b). The grate is 20 cm thick; it appears to be made of two superimposed parts, possibly indicating a renovation. The grate has a central hole that measures 40 cm in diameter. There are six other holes surrounding it that measure 15 cm in diameter and are spaced every 50 cm (fig. 2b). The design of the grate, with a large central hole that balances the weight of the slabs on either side, is such that it does not require a central support. It is likely that the mouth of the kiln (0.4 m large) was situated to the southwest, where the holes of the grate were spaced furthest apart (fig. 2b). It opened directly into the combustion chamber.
Fig. 2 – A. Photograph of Kiln 345, Yarim Tepe II; B. Plan and vertical sections of Kiln 345, Yarim Tepe II.
A. Merpert, Munchaev 1993a; B. Merpert, Munchaev 1993a, re-elaborated by C. Padovani.
- 5 The entire covering of the firing chamber is not mandatory for the structure to work (Desbat and Sc (...)
27The original height and covering of the firing chamber are unknown5. Additionally, the function of the circular cubicle is unclear (fig. 2a). It could have been associated with clay preparation or pot shaping, fuel storage, ashes or embers disposal after the cleaning of the combustion chamber.
28The combustion chamber of kiln 14 has an elongated shape, with a length of 2.2 m and a width of at least 1.5 m. It is 0.55 m deep and dug into the former level of occupation. Its northern wall is damaged, revealing the filling of the combustion chamber (fig. 3a), whereas Kilns 23 and 24 are situated against its southern wall, measuring 2.5 × 2 m and 2 × 1.5 m respectively (fig. 3b). The recurrence of a difference equal to or greater than 0.5 m between the kilns’ length and width suggests the dimensions follow a specific pattern. Kiln 14’s upper chamber was not preserved. The grate is situated a few centimeters above the floor level. The grid structure of the granary of level 5 was reused for supporting the kiln’s walls (figs. 3–4). Kiln 14’s walls were built with two rows of elongated bricks, each 10 cm wide (fig. 4b). The inner row of bricks forming an internal shelf, as well as wall pillars and possibly a central pillar, were used to support the grate (fig. 4a). The grate itself was probably made with the same elongated bricks as the kiln walls. The grate’s holes have a diameter of approximately 10–15 cm and are irregularly spaced. The kiln’s mouth was situated to the west. Between the mouth and the combustion chamber, an elongated space in the form of a 0.5 m long tunnel is clearly visible on the cross-section (fig. 4a). This is an early example of a firebox, enabling to ignite the fire at the entrance of the combustion chamber before pushing it within.
Fig. 3 – A. Photograph of Kiln 14’s section, Tell el’Uwaili; B. Photograph of Kilns 14, 23 and 24 taken from the southwest, Tell el’Uwaili.
Mission Larsa-Oueili 2019.
Fig. 4 – A. Section of Kiln 14, Tell el’Uwaili; B. Plan of Kilns 14, 23 and 24, Tell el’Uwaili.
Mission Larsa-Oueili 2019, elaborated by C. Padovani.
29We can observe that the combustion chamber is entirely filled with multiple layers of black ash and sandy earth (fig. 3a). The same accumulation of layers has been found in the filling of the street to the north of the kiln. The regularity and horizontality of these layers and the thinness of the sediments suggest they have been formed by aeolian deposits after the abandonment of Kiln 14. The layers of black ash may indicate regular firing episodes carried out in another part of the site.
30Both Kiln 345 and Kiln 14 present the same technical design, featuring a combustion chamber beneath a firing chamber enabling a vertical draught. The firing chamber was not preserved, but its presence is inferred from the remaining grate intended to support the pots during firing (fig. 5). In addition, the potters that built the two kilns followed similar architectural principles. Both kilns have an average diameter of 2 m, which is relatively large and enables the firing of hundreds of medium-size bowls. Furthermore, the rather round shape of the two kilns provides structural strength due to the pressures exerted by the bricks against each other within the curved walls (Norsker 1987: 95). According to Rehder (2000: 16), the round shape minimises heat loss compared to a square or rectangular shape. Moreover, at both sites, the potters took numerous precautions to ensure the kilns were built robustly. Kiln 345 featured 40 cm thick walls, while Kiln 14 walls were supported by an underground grid structure and buttressed by the walls of two contemporaneous kilns. These elements demonstrate that the kilns were built to last.
Fig. 5 – Schematic technical design and morphologies of Kiln 345 from Yarim Tepe II and Kiln 14 from Tell el’Uwaili.
C. Padovani.
31Despite the structural similarities of the two kilns, their morphological differences suggest variation in internal draught circulation, affecting the rise in temperature and fuel consumption. The draught is generated by a pressure difference between the interior and the exterior of the kiln. During firing, cold air is drawn in through the kiln mouth and is heated. The thermal increase raises the internal pressure, inducing a vertical airflow. As the heated air exits through the top of the structure, it creates a vacuum that pulls in more fresh air through the kiln mouth. The heat is primarily produced by the combustion of gases—manifested as flames—released by the chemical degradation of solid fuel. Efficient fuel degradation requires a sufficient temperature and supply of oxygen in the combustion chamber (Friren and Girel 2022). However, the gases generated by the process also need enough oxygen and space to burn and increase the temperature before exiting the structure.
- 6 It is uncertain which specific fuel was used in each kiln. Reeds could have been the primary fuel s (...)
32In Kiln 345, the fire was ignited in the combustion chamber and the fuel was introduced by the mouth under the grate level (fig. 5). In this relatively confined space closed by a thick grate, most of the available oxygen was probably consumed by the chemical breaking of the fuel6, and insufficient space and oxygen were left for the complete combustion of the generated gases. Consequently, these gases exited the kiln without fully burning, resulting in a significant loss of thermal energy. Furthermore, without adequate draught and sufficient oxygen at the base of the combustion chamber, there was a risk of the fire being smothered. Yet, the risk was offset by the deep combustion chamber and the large central hole in the grate providing height and a clear passage for the hot gases, ensuring an efficient draught. Most of the hot gases escaped through the central hole of the grate, creating cold spots at the edges of the firing chamber. This uneven heat distribution in the firing chamber may have left visible marks on the pottery or induce tensions in the clay pots, with one side being fired more rapidly than the other, potentially leading to cracking. Increasing the temperature within this kiln would have been difficult due to the lack of space for the combustion of the gases and the thickness of the grate absorbing heat. With a large amount of heat being lost, the firing process was not necessarily extended in time, but additional fuel was required to achieve it.
33In Kiln 14, the combustion chamber is half as deep as in Kiln 345. The fire was ignited in the firebox, located between the mouth and the combustion chamber (fig. 5). This particular element as well as the elongated form of the combustion chamber provided a 2 m space for a long flame to develop horizontally before rising through the grate and into the firing chamber. In this configuration, enough space and oxygen were provided for gases to burn and release heat without building a higher structure. The firebox was an important innovation. Although not accelerating the firing process, it enhanced the draught and may have lowered the fuel consumption compared to Kiln 345. Additionally, the uniform size and distribution of the grate’s holes contributed to a more homogeneous heat distribution within the firing chamber.
34Overall, both kiln morphologies were carefully engineered to achieve firing but Kiln 14 from Tell el’Uwaili hypothetically demonstrates greater energy efficiency and the capacity to generate a stronger draught. Nevertheless, this kiln was not necessarily easier to operate. Halaf potters using Kiln 345 had to meticulously manage fuel loading and combustion to maintain a delicate balance between raising the temperature and preventing fire smothering. However, they could have exploited the differential heat distribution within the firing chamber by positioning coarser vessels in the hotter zones and finer vessels in the cooler zones. Meanwhile, Ubaid 1 potters firing with Kiln 14 needed to be cautious not to produce an excessive draught, which could have elevated the temperature faster than the pottery could endure and dangerously increase the risk of breakage.
35From the analysis of Kiln 345 from Yarim Tepe II and Kiln 14 from Tell el’Uwaili, we assume that the large central grate’s hole, the rather circular or oval and deep combustion chamber can be considered as Halaf morphological traits, whereas the elongated combustion chamber and the firebox can be considered as Ubaid 1 morphological traits. The diffusion of these traits can be traced in the 45 kilns of the Ubaid 3–4 corpus (fig. 6). In figure 6, each line of geometric patterns corresponds to a kiln, with each pattern representing a distinct morphological trait. The number in brackets at the beginning of each line indicates the number of kilns sharing the same traits.
36During the Ubaid 3–4 period, specific Halaf or Ubaid 1 morphological traits spread beyond their core territories and were integrated into local form, generating hybrid kiln models.
Fig. 6 – Distribution map of the morphological features of Ubaid 3–4 kilns with illustrations of two morphologies of kilns.
C. Padovani; background map M. Sauvage.
37For instance, the firebox has been documented on several kilns at Tell Abada level 1, Tell Songor B level 1 and Tell Hassan level 5, but also at Tell Feres level 10a–b and Telul eth-Thalathat levels XIII and XIIb. At Tell Feres (Baldi 2015), Kiln A exhibits a short firebox, achieved by thickening the kiln walls. This feature is uncommon among the kilns of the workshops. At Telul eth-Thalathat (Fukai et al. 1970), most kilns have an oval combustion chamber over 1 m deep, akin to Kiln 345. However, K101 and K119 also exhibit a firebox. In K101, the firebox opens at the floor level, below the grate. It measures around 0.4 m in length and 0.35 m in width. Although the grate was not preserved, the central pillar found in the middle of the combustion chamber implies the grate possessed multiple small holes rather than a large central hole. It is worth noting that the spread of selected kiln traits probably occurred in multiple directions rather than exclusively from south to north. K11 at Tell Abada level 1 (Jasim 1983, 2021) provides a clear example of north-south transmission: this kiln features an elongated, buried combustion chamber and a firebox, associated with a grate with a large central hole, a characteristic element of the Upper-Tigris region.
38However, during the 600-year period under consideration, instances of kiln hybridisation are relatively scarce. There is a noticeable continuity of regional kiln morphology from the Early Chalcolithic to the Ubaid 3–4 period. For instance, in the Upper-Tigris region at Tepe Gawra (Tobler 1950: 41), Kiln 71 of level XVI shares the same morphological traits as Kiln 345 from Yarim Tepe II (fig. 6). It presents a thick grate of 18 cm with a large central hole and eleven smaller holes on the edges. The combustion chamber is less deep (only 0.65 m) but the firing chamber has a comparable internal diameter of 1.10 m. A similar circular cubicle is attached to the kiln, with the kiln’s mouth located above it. In the Hamrin, at Tell Abada level 1 (Jasim 1983, 2021), K8 features a triangular combustion chamber and its small holes grate was supported by internal buttresses. This shape, uncommon for superposed-chamber kilns, had previously been identified only at level 1 of the nearby Tell Hassan site, related to Halaf occupation (Fiorina 2014). Any kiln from this period has been identified in the Mesopotamian basin. However, in the Hamrin, K3 at Tell Songor B level 1 (Matsumoto and Yokoyama 1995) and K7 and K13 at Tell Abada level 1 (Jasim 1983, 2021) exhibit a buried combustion chamber, a firebox and a grate with small holes, closely resembling those of Kiln 14 at Tell el’Uwaili. In addition, elongated combustion chambers were more common in the Hamrin, than in Upper Mesopotamia.
39Furthermore, we suppose that potters did not consistently use borrowed morphological traits for their original function, but instead adapted them to fit their specific needs or way of firing. For instance, some fireboxes used at Tell Abada and Telul eth-Thalathat feature excessively steep inclines, not conducive to holding a large fire, associated with deep combustion chambers (fig. 7). It is likely that during firing, fuel was placed vertically against the steep incline of the firebox, with the fire being sustained at the bottom of the combustion chamber. Therefore, these fireboxes were probably not used to increase the distance over which the flame could develop. Instead, they provided a large opening that facilitated the influx of oxygen into the combustion chamber, thereby preventing the fire from being smothered. Proper draught enhancement requires a firebox to be horizontal enough to accommodate a large fire or to be associated with an elongated combustion chamber, exemplified by the kilns of Tell Feres (Kiln A), Tell Abada (K7, K13) Tell Hassan (J13) and Tell Songor B (K3).
Fig. 7 – Profiles of three Ubaid 3-4 kilns combustion chambers with steep fireboxes indicated by an arrow.
Fukai, Horiuchi, Matsutani 1970 and Jasim 2021, re-elaborated by C. Padovani.
40The absence of the Ubaid kiln package (firebox + elongated combustion chamber + grate with small holes) and the adaptation of the firebox function in Upper Mesopotamia suggest that northern communities did not adopt an Ubaid way of firing during the Ubaid 3–4 period and that efficiency was not a driver of kiln transformation. Despite limited data, the continuity of kiln morphologies is significant at the regional scale since the Early Chalcolithic, with deep and closed combustion chambers in Upper Mesopotamia against elongated ones and fireboxes in the Hamrin. While all the morphological traits identified during the Ubaid 3–4 period were present in the Hamrin region, the large majority of them also existed in the Upper-Tigris and the Jezirah, but were not spread in the Upper-Euphrates, suggesting that interactions and exchange of knowledge did occur but probably resulted from favored meso-scale contacts.
41Kilns are complex systems and cultural markers that have been overlooked by archaeology. Early typological studies are valuable for documenting kiln morphologies, but their contribution to general knowledge on ancient societies is limited by decontextualisation and positivist preconceptions. The proposed method aims to overcome these challenges by looking at the kiln’s operating system from the inside. The method differentiates between the technical design and the morphology of firing structures, thus avoiding the comparison of different technical objects sharing a similar shape. The key tenet of the method is to deconstruct each firing structure into its component parts. By analysing how each component of the structure contributes to the firing process, this approach enables a deeper understanding of the kiln operation, the expertise required for its use, and its transformation through time. While deconstructing the kiln into individual components offers valuable insights, we acknowledge that a successful firing is contingent upon the synchronous operation of all its parts and that “kiln design has been more an art than a science. There are mysteries and uncertainties about it, but existing successful kilns are of so many varied shapes and proportions that it would seem almost anything works” (Rhodes 1981: 168).
42In this paper, only objects sharing the same technical design, i.e., superposed-chamber kilns, have been examined. Our goal was to test the scope of this method by demonstrating how kiln typologies in context could help to outline different firing procedures, and also to track the dynamics of cultural transmission. We chose to focus on the two millennia following the adoption of the superposed-chamber kiln because of the significant morphological variability of kilns and potential to address questions related to the expansion of the Ubaid tradition in northern Mesopotamia.
43The Ubaid 1 kiln morphology at Tell el’Uwaili was established to be more energy-efficient than the Halaf model of Yarim Tepe II. However, this observation does not negate the fact that pots could be perfectly fired in both models by potters who knew how to operate each of them. The absence of diffusion of the Ubaid 1 kiln model in northern Mesopotamia during the Ubaid 3–4 period corroborates that efficiency and cost-effectiveness are relative criteria and should not be considered as the primary drivers of kiln transformation. Our results suggest that kiln morphology evolved through the borrowing of selected kiln elements, rather than entire kiln models, among different groups of potters socially connected. We suppose that borrowings were bidirectional and the selected kilns traits were integrated into local forms, increasing kiln morphology diversity. These interactions and exchange of knowledge participated into creating a wide range of morphological traits, shared through the Ubaid social network. Indeed, the choice of a trait from this pool of possibilities expressed the potter’s firing tradition, shaped by his own beliefs, and the firing habitus and social structure of his group, as well as external contacts with other communities. The absence of morphological standardisation across the Great Mesopotamia and instances of morphological traits borrowed but not fulfilling their primary function, suggest that no distinct Ubaid firing tradition emerged, and that local potters maintained their traditional firing practices.
44I warmly thank R. Vallet, Project Director of the Mission Larsa-Oueili since 2019, for allowing me to use unpublished data from Tell el’Uwaili. I would like to express my gratitude to J. S. Baldi, Field Director of the Tell el’Uwaili excavation, who shared with me the suffering of excavating kilns 14, 23 and 24 (and more) in the scorching heat of South Irak in June 2019. Furthermore, I would like to thank the anonymous reviewers whose valuable comments have allowed me to enhance this paper further. I extend special thanks to Victoria de Castéja and Michèle Chapouton for their invaluable assistance with the English text and support.
Flux RSS
