Attic Early Iron Age chaînes opératoires

July 10th, 2017

Attic Early Iron Age chaînes opératoires: What do we actually know?
To begin the discussion on Early Iron Age chaînes opératoires it is necessary to define an appropriate mode of production for the Protogeometric, Geometric and Orientalising periods. Unfortunately, this is difficult due to lack of textual sources and other information related to Early Iron Age potters. Stissi (2002) discusses various problems related to Late Archaic and Classical fineware production and suggests that Peacock’s (1982) ethnoarchaeological model for Roman ceramic workshops (1) could describe some Greek ceramic production modes between the 6th and 4th centuries BC. Still, the situation during the Early Iron Age remains unknown.

A major problem in understanding the scale and mode of ancient Greek pottery production relates to the different views of scholars on ancient economic models. Some, who support the Primitivist Approach, see ancient craft production as a secondary activity. It functioned inside agricultural economies and targeted specific elite consumer groups who demanded craft products either to express social status or to maintain diplomatic contacts through gift exchange (e.g. Finley 1973; Austin & Vidal-Naquet 1977; Garnsey et al. 1983; Von Reden 1995; Möller 2000). Others, who support the Market Approach, see ancient craft production similar to modern. It was a primary source of income for artisans and operated in a market regulated by laws of demand and supply (e.g. Burke 1992; Cohen 1992; Sherratt & Sherratt 1993; Sherratt 1995, 152; Osborne 1996b; Loomis 1998). Finally, there is a group of scholars who have refined the Primitivist Approach by noting the complexity behind ancient economies and craft production (e.g. Morris 1994, 351, 354; Davies 1998, 230; Parkins 1998, 1-2).

Two studies that investigate fineware production models that are chronologically close to the Greek Early Iron Age are those by Arafat & Morgan (1989) and Osborne (1996b). By comparing the organisation patterns between Athenian and Corinthian fineware production during the Late Archaic and Classical periods, Arafat & Morgan (1989) see marked differences in their spatial organisation: Athenian production was centralised by contrast to Corinthian, which was more disperse. In both cases, however, Arafat & Morgan (1989) accept that Late Archaic pottery production functioned as a secondary economic activity in societies that were mainly agricultural. By contrast, Osborne (1996b) argues that potting was by no means a supplementary activity that aimed to meet the shortfalls of agricultural production. Instead, large trade networks across the Mediterranean indicate the existence of markets in which ceramic products were sold as a main source of profit.

Even though networks are considered highly complex (see Knappett 2005; 2011; 2013), Van der Leeuw (2013) simply sees them as lines that are drawn to connect different points on a map. The length of such lines varies: Greek trade networks could have either related to organised exports towards long-distanced markets or small scale transactions between local producers and consumers. Osborne (1996b, 43) argues that the rise of trade networks of independent markets “should be assigned a place with the other transformations that mark the revolution of the 8th century BC”. Pottery markets from the 8th century BC onwards have recently been investigated by Tsingarida & Viviers (2013); however, the majority of these markets relate to large-scale long-distance trade.

If we accept that a small-scale market existed in 8th century BC Athens, then the internal relationship between Early Iron Age ceramic workshops and fineware consumers is rather unclear. Even though there is epigraphic evidence of commissioned potters producing vases for wealthy patrons during Classical times (Webster 1972), the same assumption is adopted for the Geometric period based on archaeological evidence of social competition expressed through burial rites, connected to the consumption of fine pottery (Coldstream 1977; Morris 1987; Whitley 1991; Duplouy 2006). Regardless these assumptions for commissioned pots, the possibility of an open market cannot to be excluded. Even though the production of funerary finewares might have been based on commissioning, the production of domestic finewares might not have been. In fact, domestic pottery was produced with similar shapes and decorative characteristics as funerary pottery. For this reason, it is important to understand the level of specialisation in ceramic production.

Regarding specialisation, Rice (1987, 189) suggests that a definition based on the intensiveness of production as full-time or part-time is already problematic even in the ethnological record. For example, Arnold (1985, 18) sees that specialisation is connected to full-time production modes, in which potters produce pots all over the year and their economic gain is based fully on potting. However, Rice (1987, 189) argues that specialised pottery production can be sometimes seasonal because of particular weather conditions, while in some other cases, workshops could hire part-time specialists for a certain period of time. By contrast to both, Roux (1990, 142) diversifies between technical specialisation and techno-economic specialisation. According to her definitions, technical specialisation relates to the production of an object that is meant to be consumed at a village or regional level. Such production is not the source of economic gain. However, techno-economic specialisation relates to the production of an object that is exchanged at a village or regional level for economic profit. This specialisation has two forms: “‘simple’ when the distribution of specialisation is on the basis of the raw material employed (ceramic or lithic), and ‘complex’ when production is also distributed in function of the type of object produced (i.e. ceramic containers of differing dimensions)” (Roux, 1990, 142-3).

With regard to Attic Early Iron Age fineware production, it is highly likely that Athens followed a model of techno-economic specialisation in order to meet the needs of a local market, related either to commissioned or non-commissioned products. Still, the form of this techno-economic specialisation in the production of specific shapes varied across time: Protogeometric neck-handled and belly-handled amphorae that have been recovered in Athenian burials (see Kraiker et al. 1939; Lemos 2002) were normally produced at standard sizes that could also relate to functional purposes; therefore, their production reflected simple techno-economic specialisation. However, when such vessels began to be produced in monumental sizes after LGIa (see Coldstream 1968; Whitley 1991), production probably shifted to a model of complex techno-economic specialisation.

In relation to spatial distribution, the largest quantities of ceramic test pieces, kiln waters and production debris have been excavated in the Athenian Agora (Papadopoulos 2003). The area that this material comes from extends between the Kolonos Agoraios, the south bank of river Eridanos, and the Areiopagos North-Slope cemeteries. Furthermore, this area, which was later built over by the Middle Stoa and the Odeion, was almost free of tombs during the Early Iron Age (Papadopoulos 2003, 275). Excavated pits and wells revealed large concentrations of potters’ debris and “it is likely that these wells, including those largely filled with domestic debris, served pottery establishments rather than private dwellings” (Papadopoulos 2003, 274).

Monaco (2000, 17) also notes that the quantities of perforated sherds used for kiln control, recovered under the Odeion and inside the wells extending 130m to its South-East, suggest that ceramic workshops at the Athenian Agora operated as early as the Protogeometric period. Furthermore, she notes the absence of production waste at the area of Kerameikos, which was exclusively used for funerary purposes from the Protogeometric period until the 6th century BC (Monaco 2000, 70). By contrast to Papadopoulos (2003, 274), Monaco (2000, 17, 20, 22) argues that the presence of Early Iron Age production debris and domestic pottery in the Agora wells suggest the simultaneous co-existence of pottery production units and houses. Although this possibility is likely, her characterisation of the well material as production waste or domestic pottery is problematic. In fact, the Agora wells contain ceramics that must be defined as either production debris or ready-made products. Furthermore, it is unclear if the latter were used in domestic contexts in the same area. If a pottery market existed next to the production site of the Athenian Agora, such ready-made products would have related to the workshops that sold them and not to the houses that purchased them; therefore, it is also likely that such ceramics relate to fragmented or unsold commercial waste that was dumped in the wells without having been sold or used at all.

Even though very few Athenian Geometric and Orientalising structures survive because of later building activity, there are remains of a 7th century BC kiln recovered at the foot of the hill south-west to the later Tholos. This kiln comprised of a round combustion chamber circa 1.33m in diameter, a column at its centre to support the upper floor, and a firing room of irregular shape. Even though no pottery debris was discovered, the clay floor of the chamber, the collapsed potsherds from the roof, the rich remains of charcoal and ash, and the presence of remains of a clay-lined basin in close distance, all undoubtedly suggested the presence of a ceramic workshop (Thompson 1940, 3-7).

In relation to Attic kilns, the one from the Athenian Agora resembles two potter’s kilns with circular combustion chambers of 1m in diameter from Skala Oropou, dating in the late 7th century BC (Mazarakis Ainian 1996, 21-124, pl.15b-16d; 1997; 1998). The Agora kiln is also no different to a smaller 8th century BC kiln from Torone, the shaft of which is 0.80m in diameter (Papadopoulos 1989; 2005, fig.38b; 2013, 39-42; Whitbread et al. 1997) and to an Early Orientalising round kiln from Knossos, which is 0.65m in diameter (Pariente 1994, 819-21; Tomlinson & Kilikoglou 1998, Coldstream et al. 1997). By contrast, the Agora kiln is significantly smaller and different to an Argive Protogeometric kiln of the 10th century BC, the combustion chamber of which exceeds 2.20m in diameter (Courbin 1963, 59-102), and also to a pear-shaped Geometric (or perhaps Archaic) kiln from Eretria, the length of which is 2m (Krause 1981, 86; Ducrey et al. 1993, 21-2, figs.13-14).

Rich iconographic evidence of kilns (Nobble 1960, 198-200, fig.230-8) and images of potters and painters at work (originally in Beazley 1946; summarised by Stissi 2002, pl.28-48; also see Chatzidimitriou 2005) have been traditionally used to define Athenian pottery production of later periods. Still, it is questionable whether such evidence can be used as sources of information to investigate pottery workshops of the Early Iron Age. It is certain that fineware production in Athens continued all along the Archaic and Classical era (Cook 1961; Oakley 1992; Monaco, 1999), while the distribution of ceramic production sites expanded towards the areas north-west of the Kerameikos, in the modern region of Academia Platonos (Baziotopoulou-Valavani 1994, 45). Despite the large number of excavations in Athens, no pottery production has been identified to have taken place within the Classical city-walls, apart from the area near the Dipylon Gate that was known in Classical antiquity as Kerameikos (Papadopoulos 2003, 276). The discoveries from the Agora excavations point to the direction that long before the construction of the Classical walls, potters’ activity and cemetery grounds expanded between the north-west of the Acropolis and east of the Kolonos Agoraios, leaving essentially no room for concentrated habitation (Papadopoulos 2003, 276).

Four Protogeometric ceramic kilns have been recently excavated by the 26th Ephorate of Prehistoric and Classical Antiquities at Palaia Kokkinia, Piraeus (Mazarakos et al. 2008). The kilns were made from clay mud, and although their dome did not survive, it is highly likely that it was made from the same fabric. The kilns comprised of a rounded combustion chamber separated from the furnace’s floor, which stood at ground level. Furthermore, all kilns were found in close distance from each other (Mazarakos et al. 2008, 155, pl.15), suggesting a clustered production. The broader archaeological site at Palaia Kokkinia included building remains, a road, and what might have been part of a larger cemetery, comprised of 17 burials divided in three broader burial groups. The site was in use during the Protogeometric and Geometric periods, although the actual burial groups dated in Middle and Late Geometric times (Mazarakos et al. 2008, 253-4). Palaia Kokkinia is relatively similar to Geometric Agora, as both sites combine burials and ceramic production establishments.

By ethnographic analogy, Papadopoulos’ (2003, 274-5) cluster of workshops at the Athenian Agora and the Palaia Kokkinia site (Mazarakos et al. 2008), could match Peacock’s (1982, 8-10) models of individual workshops, nucleated workshops or manufactories. The absence of separate kilns in favour of limited communal kilns and the spatial nucleation of the manufacturing debris at the Agora perhaps exclude the possibility of individual workshops. Furthermore, Peacock (1989, 9) notes that manufactories match the ergasteria of Classical Greek and Late Roman times, which produced highly specialised and standardised products at a large scale. Such production is highly unlikely to have existed in Early Iron Age Athens. This perhaps leads us to the possibility of nucleated workshops, but whatever the case, archaeological information is limited and no distinct mode of production can be postulated with certainty.

For defining the number of Attic Early Iron Age fineware production units, one can only refer to the information on Attic Geometric workshops by Davison (1961) and Coldstream (1968) based on connoisseurship. As discussed in the previous chapter and according to Coldstream’s (1968, 29-82) groupings, during the period between c.760 BC (LGIa) and c.700 BC Athens had at least 9 large fineware workshops. These were supplemented by 8 individual painters and 4 groups of affiliated painters that could have been working independently. This total of 21 Late Geometric ‘workshops’ may not necessarily reflect 21 separate production units; however, both Coldstream (1968) and Davison (1961) agree that the nine larger ones (2) must be regarded as such due to the relatively large amounts of vases attributed to them. For the 7th century BC, workshop numbers are unclear. The Analatos painter, the Mesogeia painter, the Vulture painter (Cook 1935; Davison 1961) and the Nessos painter (Jeffery 1961; Robertson 1978) are the only few artists that have been identified on the principles of connoisseurship. Still, these artists may not necessarily represent workshops, as by that time the scale of pottery production had increased.

Webster (1972, 2) provides estimated numbers of painters and potters for Attic Black-Figure and Red-Figure style workshops between c.600 and c.400 BC. His work is based on painted signatures and previous attribution by Beazley (1951; 1956; 1963). For the 6th century Webster (1972, 2) suggests:

According to the numbers for each quarter of the 6th century BC, Attic pottery production showed a gradual increase in painters and painter groups practising their work across time. Even though evidence for potters is unclear at the beginning of the century, until c.525 BC there is an increase in the number of potters and pottery classes that can be attributed to a single artisan (43 potters for 18 pottery classes); however, after c.525 BC potter numbers decline (23) despite the increase in pottery classes (39).

Webster’s (1972) estimates suggest that during the first quarter of the 6th century BC there were a total of at least 13 painters and painter groups in Attic pottery production. If we accept that Attic Late Geometric ‘workshops’ (c.760-700 BC) were at least 21 according to Coldstream (1968, 29-82), then it seems that there was a sharp decline in the numbers of painters after c.700 BC. Numbers started to recover only after c.575 BC. Again, it could be likely that the number of Late Geometric ‘workshops’ suggested by Coldstream (1968) is unrealistic and that production probably depended on a smaller number of artisans.

A general problem with approaches that employ connoisseurship as a mean of defining workshop practice is that they presuppose painters and potters are the same people; therefore, a workshop is defined based on its painters. Unfortunately, evidence for the presence or absence of such labour divisions is not enough with regard to Greek Early Iron Age production. Even for Athenian production of later times, there has been a large debate whether the ΕΓΡΑΦΣΕΝ and ΕΠΟΙEΣΕΝ signatures on decorated finewares show clear distinctions among such artisans within pottery workshops (e.g. Robertson 1972; for a full discussion see Stissi 2002, 104-21). In general, any approach based on stylistic or epigraphic evidence needs to be treated with an amount of caution.

The average output of Attic Early Iron Age ceramic production units is another area with no information. Postulations can only be made with regard to workshops of later times, and more specifically those of the early Black Figure style of the 6th century BC. The earliest recorded workshop, that of Sophilos, has been attributed at least 45 vessels in a span of 25 years of work (Bakir 1981, 78-80). However, Beazley (1956, 216-37) suggests that the largest Black-Figure workshop was that of Nikosthenes. Based on maker signatures, Nikosthenes’ name is found on at least 186 vessels, produced in a span of 35 years and decorated by a number of painters including himself (Tosto 1999, 173-82); therefore, Nikosthenic vessels need to be treated as a brand that was produced by a significantly large production unit. Sophilos and Nikosthenes provide an output limit for any Geometric and Early Orientalising workshop, as it is highly unlikely that their production could have been larger than workshops of the Back-Figure style.

Another issue connected to the internal organisation of ceramic workshops is that of gender distinctions. Pacey (1993, 104) notes that in craft production, technology is a term that is conventionally used to define the activities of men. Women’s work also falls under the definition of technology; however, it is excluded from recognition not only based on the simplicity of the equipment they use, but also because it implies a different concept of what technology is about (Pacey 1993, 104). Dobres (1999, 130-2) argues that in communal modes of production the social organisation involves material, political and economic division of labour, in which every gender participates (Dobres 1999, 132). A good example is the case of hunting in primitive societies: hunting does not stop in killing the animal and distributing the meat by the men, but it also includes cooking by women, who are also part of the total hunting chaîne opératoire (Dobres 1999, 133-4).

Based on our current evidence, it is impossible to identify gender distinctions in Attic Early Iron Age ceramic workshops. It is likely that in the total chaîne opératoire labour was divided in primary and secondary tasks, which could have involved different genders. Judging from the signatures of potters and painter in Athenian vessels of later times, the dominance of male names is profane. Beazley (1956) records long lists of names of Attic potters of the Black-Figure style, who are all males. Additionally, the first name of an Attic potter signed on a vase is again male, that of Sophilos (c.610- 550 BC) (Cook 1960, 70-2). Following the thoughts of Dobres (1999), females must be treated as part of the ceramic chaîne opératoire in Early Iron Age Attica: the consumption of ceramic finewares in funerary rites was definitely guided by notions related gender (Whitley 1991; 2000); however, any gender notions behind pottery production are still unclear.

It can be generally postulated that from circa 615 BC onwards, the time of production of the Nessos amphora (Cook 1960, 69), Attic pottery workshops were monopolised by men, despite which gender was destined to consume the ceramic products. Langdon (2015) argues that some miniature drinking cups recovered in children’s graves from Late Geometric Kerameikos were decorated in a ‘clumsy’ way, which implies that children were perhaps involved in Athenian Geometric ceramic production. At the moment, it is still unclear if children were drastically involved in workshops as full-time workers or apprentices, or if the decoration of a specific class of vessels was depended on children due to symbolic reasons.

Gender or age divisions in Attic Early Iron Age workshops may need to be examined under broader terms, such as skill variation and apprenticeship duration. Both terms are interconnected. According to Roux (1990, 143), the duration of apprenticeship of a technique is linked to three aspect: firstly, the nature of know-how involved, which depends on the technique and the method of production used; secondly, on the physical properties of the worked materials; and thirdly on the perceptual-motor capacities put in action by the potter.

Ethnographic work by Roux & Corbetta (1989) has shown that wheel throwing is a complex technique that takes long time to learn. The duration of apprenticeship necessary for mastering wheel throwing depends on four aspects: 1) the process of apprenticeship that is locally followed by young potters; 2) the fashioning phases of producing a pot and the organisation of two-handed gestures that are required in each apprenticeship stage, and for each type of pot; 3) the motor abilities developed by potters; and 4) the potters’ performances according to their stage of apprenticeship (Roux & Corbetta 1989) (3). In general, new potters need to integrate these factors in a progressive way. The acquisition and successive mastery of motor abilities depends on strategies employed at each stage of apprenticeship (Roux 1990, 144) (4). Given that Attic Geometric and Orientalising decorated finewares are all wheel-made, apprenticeship duration should be regarded as a long-lasting process too. In that sense, it is highly likely that the work of children in pottery production could have related to a learning process that started at an early stage of their lives.

Finally, there is no information in relation to whether an individual potter produced indiscriminately every and any ceramic object, or if there was further specialisation and focus on specific shapes attributed to specific potters. Roux (1990, 147-8) provides three criteria to facilitate this distinction: 1) the diversification of forms and dimensions of vessels; 2) the quantities of pots of each type; and 3) the standardisation of the products. In pottery production of the Harappan culture in the valley of Indus (Roux 1990, 148) the throwing of small vessels does not require the competences noticed for the production of larger vessels. Furthermore, a substantial demand for larger vessels would produce a division of tasks among potters according to their dimensions (Roux 1990). Based to the above, a key concept in understanding specialisation in pottery production is the analysis of standardisation of the final products.

Previous stylistic approaches on Attic Early Iron Age decorated finewares by Desborough (1952), R.M. Cook (1960) and Coldstream (1968; 1977) have described the evolution of typologies and decorative styles as a linear process, in which specific shapes, painted motifs and figurative themes belonged to specific chronological periods. This approach was convenient in order to establish relative chronologies based on style; however, it also produced the notion that Attic Early Iron Age fineware production was standardised across specific periods of time. Even though changes occurred gradually, it was thought that these were adopted and followed by almost every workshop simultaneously. This understanding of general and linear evolution of ceramic typologies and styles is to a great extent responsible for the notion that Athenian Early Iron Age workshops functioned as a cluster, where diversification among workshops was highly unlikely to have occurred. It remains a question if this was actually true. The author's doctoral thesis approaches standardisation from another angle, that of the chaîne opératoire theory, and offers some different answers to this problem.


  1. Peacock (1982, 8-11) identifies eight major modes of production that probably existed during Roman times: 1) household production; 2) household industry; 3) individual workshops; 4) nucleated workshops; 5) manufactories; 6) factories; 7) estate production; and 8) military (or other official) production.
  2. Dipylon Master, Hirschfeld painter, Lambros painter, Athens 706, Birdseed painter, Soldier-bird workshop, workshop of the Hooked-Swastikas, Athens 894 and Athens 897.
  3. For more on wheel fashioning techniques and their identification on the final products see Roux & Courty (1998). Roux (2003) also suggests the Dynamic System Framework approach in understanding technological change in wheel throwing practices.
  4. Such procedures involving the gradual acquisition of motor abilities across time have already been discussed for the production of hand-made pottery in prehistoric societies by Kamp (2001) and Loney (2007).