See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271742957

Experimental approaches to archaeological ceramics: unifying disparate

methodologies with the chaîne opératoire

Article in Archaeological and Anthropological Sciences · March 2014

DOI: 10.1007/s12520-014-0177-4

CITATIONS READS

13 782

1 author:

Caroline Jeffra
University of Amsterdam
20 PUBLICATIONS 101 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Tracing the Potter's Wheel View project

Early Wheel Use in Crete and Cyprus: PhD Research View project

All content following this page was uploaded by Caroline Jeffra on 12 June 2017.

The user has requested enhancement of the downloaded file.

Archaeol Anthropol Sci (2015) 7:141–149
DOI 10.1007/s12520-014-0177-4

ORIGINAL PAPER

Experimental approaches to archaeological ceramics: unifying

disparate methodologies with the chaîne opératoire
Caroline D. Jeffra

Received: 30 November 2012 / Accepted: 26 January 2014 / Published online: 12 March 2014
# Springer-Verlag Berlin Heidelberg 2014

Abstract It has been stated that experimental archaeology is technological questions are experimental archaeology’s forte,
inherently better equipped to address technological factors while social questions remain less accessible, as highlighted
than social ones. In this respect, the chaîne opératoire ap- by Harry (2010). What is perhaps necessary to take experi-
proach has much to recommend it for experimental archaeol- mental archaeology research from its core strength into an area
ogy. In this contribution, a case will be made for the increased of weakness is the incorporation of technologically based
interpretive power of this combination by examining a num- approaches firmly anchored to elucidating social phenomena.
ber of examples where authors have incorporated chaîne In so doing, practitioners of experimental archaeology may be
opératoire explicitly or implicitly within their investigations better equipped to apply their technologically explicit results
of a single-material type: ceramics. It is argued that, by using and make solid assessments of social phenomena. It is sug-
the chaîne opératoire to frame and contextualise experimental gested here that the chaîne opératoire approach is a powerful
archaeology methodologies and interpretations, individual re- methodological tool, which may help experimental archaeol-
search programmes can produce more anthropologically rele- ogists to build bridges between the technical and the social. To
vant results. this end, a selection of case studies is explored below in order
to illustrate ‘best practise’ in combining experimental archae-
Keywords Ceramics . Experimental archaeology . Chaîne ology and chaîne opératoire approaches. By adding this ap-
opératoire approach proach to experimental archaeology’s methodological and
analytical toolkit, it is hoped that future researchers will be
able to craft more robust experimental methodologies which
Introduction can be readily directed toward anthropologically relevant
conclusions. In order to clarify how a chaîne opératoire ap-
Experimental archaeology has been applied to ceramic mate- proach might contribute to experimental archaeology, howev-
rials through a wide array of hypotheses, and one of the most er, it is necessary to define each.
striking observations when studying these contributions as a
group is the variety in methodological approaches. This is in
part thanks to the flexibility of the potting process itself, where Experimental archaeology and the chaîne opératoire
countless viable variations in technological behaviour are
theoretically possible within the chaîne opératoire of pottery Defining experimental archaeology
production. As such, experimental approaches can, and indeed
have, focused on ‘how’ and ‘what’ questions and have been In a well-cited definition, Skibo calls experimental archaeol-
less adept at addressing ‘who’ and ‘why’ questions; ogy ‘the fabrication of materials, behaviours, or both in order
to observe one or more processes involved in the production,
C. D. Jeffra (*) use, discard, deterioration, or recovery of material culture’
Fondation Fyssen, Université Paris Ouest Nanterre la Défense, La (Skibo 1992, 18). Beyond this, the field of experimental
Maison René-Ginouvès, Archéologie et Ethnologie, UMR 7055
archaeology has been delineated in the literature by differen-
Préhistoire et Technologie, 21 Allée de l’Université, France
92023 Nanterre CEDEX tiating it from related sub-disciplines. In his discussion of
e-mail: [email protected] experimental archaeology, Outram (2008) takes what may
142 Archaeol Anthropol Sci (2015) 7:141–149

be considered as either a broad view of archaeological science, incorporate sometimes rarely occurring variables relating to
or a narrow view of experimental archaeology. While he holds natural processes, as could be ascertained in a study of land-
that hypothesis testing is common to both, he suggests that scape alteration by a potting industry. A final category pro-
they are separated by what he refers to as ‘actualism’ (Outram posed by Reynolds is termed ‘technical innovation’, where
2008, 2), the incorporation of experimental conditions which archaeological techniques or equipment serve as the subject of
more closely approximate ‘real-life’ scenarios than laboratory experimentation such as the application of new imaging tech-
conditions. In his view, archaeological sciences include the nologies to ceramic vessels.
laboratory-based testing and replication to clarify those as- With no reference to Schiffer et al. (1994), it is not imme-
pects of a material’s physical properties which dictate the diately clear how these categories described by Reynolds
creation, use or decay of materials found archaeologically (1999) may relate to the suggested differentiation between
using methods which maintain tight control over variables. archaeological experiment and experimental archaeology. It
Experimental archaeology, on the other hand, builds on this is conceivable, however, that the two are not mutually exclu-
knowledge of physical principles in order to ‘investigate ac- sive; ensuring that each category of experiment suggested by
tivities that might have happened in the past using the methods Reynolds is composed of multiple archaeological experiments
and material that would have actually been available’ (Outram can perhaps fulfil the requirement that experiments be repeat-
2008, 2), although he does acknowledge that those methods ed, as identified by Reynolds (1999). What is less explicitly
and materials which are not pertinent to the hypothesis need established is the relationship between the testing of experi-
not be so rigidly actualistic. This strict use of ‘actualism’ to mental hypotheses and the derivation of anthropologically
demarcate experimental archaeology from archaeological sci- relevant interpretations through an understanding of the ce-
ence is not universally agreed, however. Indeed, Schiffer et al. ramic chaîne opératoire. If, as stated by Outram (2008, 2),
(1994) focus their definition of experimental archaeology on experimental archaeology allows for the impact of unpredict-
the extent of a research programme. Experimental archaeolo- able phenomena to ‘enabl[e] the refinement of hypotheses and
gy, in their view, is composed of many archaeological exper- archaeological interpretation’; then, perhaps, the next step for
iments; archaeological experiments are, in this perspective, ceramic experimental archaeology is to incorporate anthropo-
the individual instances where yes/no hypotheses are tested. logical interpretative frameworks by taking a chaîne
The context in which these yes/no hypotheses are addressed is opératoire approach.
left open by Schiffer et al., but a look at the methods employed
in their previous research indicates the inclusion of both
laboratory and actualistic methodologies. Furthermore, they Defining the chaîne opératoire approach
identify early-stage archaeological experiments as crucial to
the creation of appropriate experimental methodologies. The chaîne opératoire approach asserts that socially defined
While Schiffer et al. (1994) may have differentiated be- groups can be rendered archaeologically visible by their asso-
tween archaeological experiment and experimental archaeol- ciations with technical profiles (Roux 2011, 80). This asser-
ogy based on the extent of a research programme, Reynolds tion is rooted in the socially embedded nature of transmission,
(1999, 158–162) presents frameworks for subdividing exper- learning processes and motor habit acquisition (Manem 2012,
iments based on aims and complexity, which has implications 141). In short, because an individual learns their craft in a
for an understanding of the subfield of ceramic experimental social context, dictated by factors of identity such as family,
archaeology. The first of these is the construct, which relates to gender, culture, caste and/or ethno-linguistic group (see
hypotheses of the superstructure of foundation evidence iden- Kramer 1997; Degoy 2008; Gelbert 2003; Stark 1998;
tified archaeologically, as would be the case in experiments Lemmonier 1993; Herbich 1987; Roux 2010; Manem 2012;
involving kiln structures. The second type of experiment is Wallaert-Pêtre 1999; Gosselain 2002; Stark et al. 2008), the
termed ‘process and function’ and is designed to identify the technical profile of their products will resemble other mem-
way in which things were achieved; cooking-pot-use experi- bers of those social groups. This is true despite the fact that
ments fall within this category. This type relates most closely technical profiles are in fact ‘sociotechnical aggregates’ which
to the technological procedures used in the past. A third type represent the ongoing negotiation between a craftsperson’s
of experiment is termed ‘simulation’. Simulations, according learned techniques, borrowing, inventions and alterations
to Reynolds, are designed to address issues of site formation (see discussion in Gosselain 2000).
and taphonomic processes. The operation of a potter’s work- Components of these technical profiles have specific ways
shop may be the subject of a simulation experiment. Eventu- of relating to social groups within the potting craft, where
ality trials, the fourth category, require a combination of many aspects of the potting process are invisible to non-
construct, process and function and simulation experiments potters. Gosselain has been able to relate the visibility of
at a large scale and over a long period of time. In so doing, technical components (and how likely it is that non-potters
these experiments address the complexities of systems and may influence specific aspects of the potting process) to the
Archaeol Anthropol Sci (2015) 7:141–149 143

identity of potters, expressing the presence of specific three aspects: techniques (physical modalities by which raw
relationships: material is acquired and transformed), methods (the particular
sequence followed) and tools (Roux 1994; Roux 2003, 9–10).
in southern Cameroon, potters process clay in a way that By examining these aspects within assemblages, technical
makes them unmistakable members of a specific com- groups may be constructed which share common technical
munity; fashion vessels with a technique shared by a profiles. Through this, the relationships between varying tech-
larger, but nevertheless bounded, group of individuals; nological traditions, and the social groups practising them,
use the same ornamental tools and motifs as an even may be better defined (Roux 2011). Incorporating a diachron-
larger group of people; fire the pots in structures and ic component to this approach allows for the identification of
with fuels typically associated with communal or re- horizontal transmission of technical practise, which is, the
gional traditions; and treat the pots after firing with extent to which multiple categories of manufacturing tech-
techniques and materials distributed at still another spa- niques are incorporated by potters in post-learning contexts,
tial and social level (Gosselain 2000, 189). outside of the vertical relationship between teacher and pupil
during apprenticeship.
Gosselain (2000, 191) thus subdivides the potting process
into three categories of manufacturing stages based upon their An explicit combined framework
‘salience, technical malleability, and the social context in
which the techniques are learned and conducted’. The visibil- In practical terms, tailoring experimental methodologies to
ity of potting processes on a finished product therefore has an adhere to a chaîne opératoire approach requires but minor
impact on the salience of that process during a potter’s life- adjustments. Experiments frequently focus on specific seg-
time. Relating back to the concept of changing sociotechnical ments of the manufacturing process and are drawn from
aggregates, some elements of the potting process are more assemblage- or site-specific questions. What is currently lack-
readily influenced by what a potter observes around them and ing in many cases is the explicit consideration of how an
what demands a non-potting consumer may make; similarly, experiment’s foci relate to the wider considerations of the
they are relatively easy to transmit in post-learning environ- chaîne opératoire. In order to implement the chaîne
ments (Gosselain 2000, 191). opératoire approach, it is therefore suggested that those
Highly visible techniques, such as mixing clays to alter embarking on an experimental programme first address the
paste colour, decoration, some types of firing such as smudg- salience of the manufacturing process they hope to study. By
ing and post-firing techniques, are the most likely to change, first identifying the likelihood that the manufacturing process
and Gosselain highlights them as belonging to a first category observed was influenced by post-learning interactions, exper-
of manufacturing techniques. A second category is composed imental research will be better equipped to discuss the rela-
of clay selection, extraction and processing and vessel firing. tionship between the observed manufacturing processes and
The key differentiating factor between this and the first cate- those used by other potters. If one is to assume that potters
gory is the fact that those who may influence a potter are likely actively altered the type or frequency of temper, then it is
other potters or those who are otherwise involved in the important to be able to discuss the social conditions under
potting process; Gosselain (2000, 192) specifically cites the which such an alteration would have taken place, rather than
communal nature of the knowledge required for this category just functional costs or benefits. As such, any consideration of
of manufacturing stage, shared among potters and their assis- the salience of a given process should clearly enumerate, in as
tants. The final and most salient category of manufacturing much detail as is available:
stage relates to the fashioning of vessels, which draws on the
specialised gestures and motor habits learned during a potter’s – Steps taken within the process
apprenticeship. Furthermore, this leaves little trace on finished – Interchangeability of steps (e.g. under what conditions
vessels and is individual rather than communal in nature, may steps be incorporated by other potters? Would they
rendering it the least likely aspect of a potter’s preferred need direct contact with the potters using these tech-
technical profile to change (Gosselain 2000, 192). More recent niques, or could they be learned from word of mouth?
work by Gosselain (2008, 170) has modified this assertion, Study of finished objects?)
where he has stated that, while shaping techniques are viewed – Degree of specialist knowledge for communicating un-
by potters as a kind of cultural inheritance, they are not derlying concepts (e.g. can a non-potter describe steps to
immune to alteration and as such should not be considered a potter effectively?)
as ‘closed technical units’.
Putting chaîne opératoire into practise requires the detailed An explicit discussion of the salience of particular process-
study of technical practise within crafts, and as such, Roux es is particularly assisted by actualistic components of exper-
suggests that a description of a chaîne opératoire must address iments, as advocated by Outram (2008). By having the
144 Archaeol Anthropol Sci (2015) 7:141–149

experiential component integral to actualistic experiments, the Single segment experiments

communicability and interchangeability of steps can be
demonstrated. Raw materials
A second step toward integrating a chaîne opératoire ap-
proach to experimental archaeology centres on defining the An excellent example of tackling the topic of raw material
relationship between the manufacturing processes studied and experimentally is found in the work of Stoner et al. (in re-
the wider picture of assemblages, periods or sites. Rather than view), which uses a four-faceted approach to define the point
focusing on a single type of manufacturing process (such as at which salts were introduced into ceramic fabrics from
shell tempering, when sand and grog tempering is also ob- Xaltocan, Mexico. High values of sodium and potassium were
served), researchers should explore the spectrum of related known to be present in the ceramics, but it was unclear
manufacturing processes within a site, assemblage or period. whether these were due to the nature of raw materials or the
This requirement relates directly to the distinction between post-depositional environment, particularly as the site is an
experimental archaeology and archaeological experiment put island at the centre of a saline lake. In order to experimentally
forth by Schiffer et al. (1994), by encouraging the creation of assess these two alternative hypotheses, they first compared
research programmes composed of many individual experi- Xaltocan ceramics against clays and ceramics from the Basin
ments. By looking at the spectrum of manufacturing process- of Mexico using 5,000 neutron activation analysis (NAA)
es, researchers can then quantify to what proportion of known assays. Second, they collected NAA data from samples of
material the investigated phenomena relates, while at the same clay which had been doped with water of differing salinities
time populating the understanding of a whole spectrum of and then fired to a range of temperatures. A third experiment
manufacturing processes within particular contexts. involved applying leaching techniques to a selection of ar-
By first identifying the interchangeability of steps within a chaeological sherds from the site, and later comparing the
particular manufacturing process, a more comprehensive pic- NAA results of pre- and post-leached sherds. Their final
ture of the salience of particular techniques may be created. experiment analysed both the clay and temper fraction of a
Furthermore, feeding this picture of differential salience into a sample of archaeological sherds using laser ablation-
framework of concurrent manufacturing techniques allows for inductively coupled plasma-mass spectrometry (LA-ICP-
experimental archaeology to contribute in a major way to MS) to try to clarify if either component is more responsible
demonstrating the give-and-take nature of technological iden- for the sodium and potassium levels found in the sherds
tities. These snapshots, when consolidated diachronically, can overall. Through these four separate methods, Stoner et al.
then be applied to interpretations of social interactions through were able to demonstrate that the post-depositional introduc-
time. What follows below is a discussion of individual exam- tion of salts could be removed through leaching processes and
ples of experimental archaeology which have, explicitly or that the raw materials did indeed account for the high levels of
implicitly, made use of some of the concepts integral to this sodium and potassium detected through NAA. These findings
chaîne opératoire-influenced experimental archaeology. have allowed them to argue for local production of Xaltocan
ceramics and to propose an association of raw material col-
lection with salt procurement activities on the outer shores of
the lake.
Categorising experiments Stoner et al’s (in review) work demonstrates a combination
of process and function and technical innovation experiments,
The application of experimental methodologies depends upon according to Reynolds’s (1999) definitions. But, while the
an understanding of the steps involved in the chaîne incorporation of several experiments within their programme
opératoire of production, regardless of whether or not this is supports the categorisation of their work as experimental
explicitly acknowledged. As such, experiments can be ex- archaeology according to Schiffer et al’s (1994) definition,
plored based on the aspect of the chaîne opératoire addressed the inclusion of a greater degree of actualism would have
methodologically. Most experiments which in some way ad- made a welcome further addition. This could take the form
dress the production of ceramics can be described as either of utilising clays and tempering materials found near to the
single segment or multiple segments. Single-segment experi- archaeological context and subjecting these to similar leaching
ments address one group of chaîne opératoire actions, such as procedures and chemical analyses.
those relating to raw materials (incorporating resource collec- By explicitly addressing the issue of raw material source,
tion, processing and paste preparation), production (including Stoner et al. cover the first stage of the chaîne opératoire.
formation, humid and leather-hard alterations and decoration) They focus on the procurement of clay, temper and water for
and firing. Combining these categories within an experimental potting and justifiably treat these three elements as compo-
methodology produces what is referred to here as combined nents which are not necessarily collected from the same
segment experiments. source area. The use of LA-ICP-MS to differentiate between
Archaeol Anthropol Sci (2015) 7:141–149 145

clay and temper directly relates to the potential for raw mate- Courty 1997), Roux and Courty hypothesised that during the
rial distribution throughout the landscape, which has the po- earliest use of the pottery wheel, potters did not start wheel
tential for identifying where potters may have been able to throwing vessels immediately, but instead employed pre-
outsource raw material collection. For, if the clay was found to existing coiling techniques in combination with rotation. To
contain sodium and potassium while temper did not, one explore this, they defined four potential combination methods
could argue that clay sourcing alone could be paired with salt which make use of differing amounts of rotative deformation
extraction activities along the lakeshore. If that were the case, in the potting process and had two professional potters man-
then it is possible that potters were making use of salt extrac- ufacture a type set of vessels using these techniques. The type
tors’ mobility to lessen their own need for making such a trip. set was then inspected for surface and microscopic features
As it stands, the evidence presented by Stoner et al. does not indicative of the forming methods used. By applying this
indicate that either material was devoid of sodium and potas- understanding of surface and microscopic features, they were
sium; so, a distribution of transportation labour is currently able to demonstrate that 4th to 3rd millennium BC potters in
inaccessible. As another relevant point, they argue that it is Asia (specifically Shar-i-Sokhta, Iran and Kalibangan, India)
necessary to construct a more explicit consideration of water were indeed using formation methods which combined coil
sources for potting, for the mineralogical composition of water building and rotative shaping, challenging the notion that the
sources can have an impact on the composition of fired introduction of the pottery wheel automatically signifies
ceramic bodies, as is indeed seen in Rye’s (1976) work wheel throwing archaeologically.
discussed below. Though not possible to establish within the Within Reynolds’s (1999) classification of experiments,
scope of this contribution, what is lacking from a chaîne this contribution should be considered a process and function
opératoire perspective is a discussion of the interchangeability experiment. This work is notable because it adheres to the
of the materials and indeed the processing of those materials. defining characteristics of experiment put forward by both
Although temper is considered from a chemical perspective, Schiffer et al. (1994) and Outram (2008). This publication
consideration of its procurement and sorting remains unclear. is, as noted above, one of several, which itself examines two
The degree of specialist knowledge for this aspect of the contexts and relies on multiple examples of type vessels
chaîne opératoire is debatable. While Gosselain has suggested manufactured by two different potters. The clay used for the
that raw material sourcing tends to remain stable during a manufacture of the type set of vessels was chosen because of
potter’s lifetime, provided that the source is accessible, the its compositional similarities to the archaeological material
paste preparation process itself would arguably require more against which it was to be compared, adding a further degree
specialist knowledge. So, a potter could perhaps have an of actualism to the experiment.
assistant or affiliate transport raw material, once a resource is Roux and Courty clearly enumerate the steps within the
identified within the landscape. Assuming that one in the role aspect of the chaîne opératoire which they investigate. Each
of occasional clay transporter is not a specialist in this task— formation method is described in terms of its constituent
that it is a task incidental to their own daily activities—it can techniques, particularly noting the aims of coiling or rotative
be argued that the identification of new sources of raw mate- manipulation during each stage of forming, joining, thinning
rials by a non-potter is highly likely. and shaping. By its nature, this contribution demonstrates the
Stoner et al. contextualise this type of pottery, demonstrat- interchangeability of potting steps. Each formation method is
ing its dispersal to other sites as they provide evidence of post- presented in terms of apprenticeship models, demonstrating
depositional environments which would not saturate the the inherent difficulty faced by potters who attempt formation
sherds in salts as at Xaltocan. Furthermore, they discuss the methods which differ from those taught during early potting
compositional difference between Xaltocan ceramics and experience. Through this, the specialised nature of this knowl-
those in neighbouring areas. Because of this, they are able to edge is made clear; because formation techniques are motor
describe the use of high-saline raw materials as a local behav- skill and gesture-based, potters would be unlikely to funda-
iour, based on the use of local materials, which presumably mentally alter their practises without significant incentives.
corresponds with the localised use of materials by other near- Roux and Courty are successful at providing technological
by neighbours as well. context to their experimental data, as it is at the heart of their
subject matter. The earliest use of the pottery wheel is, as
Production would be expected, set against a backdrop of handmade
potting. Therefore, the alteration of forming methods is pre-
An exemplary experimental approach to understanding ce- sented as discontinuous. Because of the discontinuous nature
ramic production which highlights the role of the chaîne of rotative potting methods and the firm basis Roux and
opératoire approach comes from Roux and Courty (1998). Courty present in the chaîne opératoire, they are able to
In this contribution (one of a number on the topic, including discuss the earliest use of the pottery wheel in terms of
Roux 1994; Courty and Roux 1995; Pierret 1995; Roux and symbols of urban identity, rather than simple manufacturing
146 Archaeol Anthropol Sci (2015) 7:141–149

economics. In later contributions, this discussion is taken which were used are described in sufficient detail to replicate
further and, based on the coexistence of multiple distinct the experiment, but the processes and rationalisation for those
technical traditions, used to identify groups as socially hetero- processes is little discussed. For example, the method of
geneous or homogeneous (Roux 2008). stacking fuel, and indeed, the decision-making process behind
the ratio and mixture of fuel types is absent from the article.
Firing Because of these omissions, from a chaîne opératoire per-
spective, it is difficult to address the interchangeability of the
Balkansky et al. (1997) conducted an experiment which inves- steps used except in a very general sense. Ideally, similar
tigated the firing aspect of the chaîne opératoire, applied to the experimental reports should communicate the stacking
context of Ejutla, Oaxaca, Mexico where it was previously methods used for fuel and vessels, the techniques employed
assumed that potters used formal kilns in the manner of Ejutla’s for stoking the fire and increasing temperature, and—espe-
northerly neighbours at Atzompa and Monte Albán. In this cially where temperatures are recorded—a firing curve
context, the excavators found shallow features overlain by showing temperature change over the course of the
‘dense midden deposits with an abundance of figurines and whole firing process.
figurine wasters, figurine moulds, and hundreds of wasters from Specialist knowledge is a prerequisite for successful firing,
a variety of ceramic vessels’ (Balkansky et al. 1997, 142). The but within this aspect of the chaîne opératoire, there are
lack of formal kiln structures led the excavators to hypothesise differences. Most interchangeable within the spectrum of req-
that four on-site pit features played a role in the firing of pottery, uisite knowledge is the pit itself; a non-specialist could easily
especially in light of the conditions of the fill and the charac- see the pit, learn of its function and communicate its dimen-
teristics of the surrounding soil and bedrock. To test this, an sions and purpose to specialist. Fuel type used could also be
experimental firing was conducted based on the archaeological transmitted with relative ease, although this becomes less
evidence. A pit was excavated and lined with sand to ease re- communicable via non-specialists when fuel is prepared in a
excavation following firing and then filled with a range of special way or is mixed to particular specifications. Non-
pottery forms separated by spacers to improve airflow between specialists would have a more difficult time communicating
vessels. A range of fuel types including cow dung, grasses and the ratio of fuel to pots used during pit firing, as this is
fallen tree branches were placed around the ceramics and then informed by the experience of the specialist firing the pottery.
covered with straw which was then coated with mud plaster This ratio has a direct effect on the duration and intensity of
insulation. Over an 8-h firing, temperature was recorded every the firing, and its underlying rationale can be difficult to
15 min by four installed thermocouples. The experimental pit understand unless the non-specialist has the opportunity to
firing created a feature which, while more ephemeral, was very ask questions and learn directly from a practising specialist.
similar to the archaeology. Balkansky et al. suggest that, if The process of stacking the pots and stoking the fire are the
sealed soon after use or reused for subsequent firings, it is likely most inaccessible aspect of the chaîne opératoire to the non-
that this feature would in fact closely parallel those found at specialist. Both stacking and stoking rely not only on concep-
Ejutla. Furthermore, the mud plaster which was applied as a tual knowledge of the process but also previous experience to
final layer to the filled pit hardened during firing, ultimately guide actions. Stacking vessels for firing in modern contexts is
resembling material recovered at Ejutla, Peñitas, and a number guided by experts, who place vessels within the firing area to
of other pottery production sites in Mesoamerica (Balkansky maximise air (and heat) flow, and allow for controlled settling
et al. 1997, 148). These experimental by-products may there- as fuel is consumed and vessels contract. Because of this, a
fore provide further support to the interpretation of the pits at non-specialist could certainly watch the process, but would
Ejutla as firing installations. likely be unable to communicate it back to specialists with
This experiment, combining aspects of a construct and sufficient detail to ensure its replication.
process and function types (as defined by Reynolds 1999), The authors are, however, successful at contextualising this
successfully incorporates actualistic components within its technical practise within the wider picture of firing in Meso-
design. It does not, however, constitute a wide-ranging exper- america. The experiment was pursued because of the differ-
imental programme composed of multiple individual experi- ence between Ejutla and its nearby neighbours, where formal
ments as advocated by Schiffer et al. (1994). To fulfil this, the kiln structures were employed. Furthermore, the identification
experimenters could have conducted multiple firings within of the types of archaeological evidence which may be associ-
the pit described (and perhaps others with differing lining ated with pit kilns—the pits and the hardened mud plaster
material), thereby testing the assumption that re-use of these fragments—highlights previously unknown technological
features could render them less ephemeral. connections between this site and others within Mesoamerica.
Balkansky et al. (1997) do describe the steps to this aspect The degree of specialist knowledge required for successful pit
of the chaîne opératoire, although there are some areas in firing demonstrates that these similarities may provide evi-
which they could have been more explicit. The materials dence for significant social connections between sites.
Archaeol Anthropol Sci (2015) 7:141–149 147

Combined segment experiments side by side in terms of function, the explanation of technical
choice is justified primarily on functional grounds. Consider-
Raw materials and production ing the steps from the perspective of the non-specialist, it may
be stated that different steps within the chaîne opératoire are
Skibo et al. (1989) provide an excellent case study into the differentially communicable. The substance of tempering ma-
relationship between raw materials and production which may terials, if not necessarily the performance attributes, is visible
be assessed from a chaîne opératoire perspective. This multi- to non-specialists and can be easily communicated to special-
component experiment investigates the functional advantages ists. Sand, crushed shell, grass or manure are each visible in
of organic tempering by comparing it against untempered and the vessels themselves and would be recognisable to non-
mineral tempered clay. A number of individual experiments specialists who witness the potting process. How the temper-
were conducted to test drying effectiveness and ease of man- ing materials are processed, where they were collected from,
ufacture, which included clay workability and expediency of and what beneficial characteristics are conferred are less vis-
manufacture. Drying effectiveness was assessed recording ible to the non-specialist. In this respect, certain aspects of
weight loss between briquette forming and the leather-hard temper choice as relating to vessel production are more likely
stage. It was found that tempered clays dry more quickly than to be transmitted between specialists.
those which are untempered and that mineral temper allows This work demonstrably fits within a wider context of
for slightly faster drying than organic. Furthermore, untem- temper choice. By examining organic temper from a techno-
pered and organic-tempered clays were more likely to warp functional perspective, Skibo et al. are able to argue that
and crack than was mineral-tempered clay. Ease of manufac- Eastern US Late Archaic potters were not necessarily choos-
ture was tested by measuring workability and expediency of ing temper type based upon aesthetics alone. Furthermore,
manufacture. Workability was established by mixing dry clay they connect the positive formation attributes conferred by
with either coarse sand or horse manure before adding water the inclusion of organic material temper to the nomadic
until clay reached plasticity. After wedging, half of each hunter-gatherer lifestyle which would have likely valued ex-
mixture was set aside for 14 days of ‘souring’ in plastic bags. pedient potting technology. They also relate the later shift to
The soured and unsoured samples were tested identically: five using mineral tempering material to the rise of sedentism in
coils were made by hand and wrapped around a beaker in the Early Woodland period.
order to observe cracking and bending properties. Both sand
and manure tempered mixes were prone to some cracking, Raw materials and firing
although it was more visible in the latter. Skibo et al. note that
the organic temper did have a positive impact on wet strength The relationship between raw materials selected and the firing
of clay, making a more rigid vessel. Expediency of manufac- process has been studied experimentally by Rye (1976). The
ture was tested by adding manure, fine sand or coarse sand to aspects of the chaîne opératoire specifically addressed include
excessively wet clay until the mixture reached workability. It raw material sourcing, temper sourcing, sorting and mixing
was found that both the fine sand and the manure were better and firing temperature. This excellent example addresses a
suited to this task than coarse sand. Furthermore, the organic well-illustrated problem in potting which relates to the chem-
temper was demonstrated to be suitable for creating slab ical reactions of calcite bodies (when used as tempering ma-
vessels immediately after it was mixed with excessively wet terial) during the firing process, which may cause spalling or
clay. Taken together, these findings allow Skibo et al. to complete vessel collapse. Rye (1976) addressed this problem
suggest that organic-tempered pottery was favoured in the in a Melanesian context, aiming to establish the conditions
Eastern USA during the Late Archaic period due to its expe- under which CaCO3 can be used in low-fired ceramics while
diency, which fits with the hunter-gatherer subsistence pattern. avoiding CaO hydration damage after firing (Rye 1976, 121).
This process and function (Reynolds 1999) experiment Clay and unsorted beach sand (with naturally occurring shell
relies on both actualistic and non-actualistic data to draw and coral fragments) were collected from three different areas
conclusions and successfully incorporates a number of indi- of the south Papuan coast, mixed in varying ratios of sand to
vidual experiments, thereby fulfilling the requirements pro- clay and wetted in two parallel samples with either seawater or
posed by both Outram (2008) and Schiffer et al. (1994). freshwater. Test bars were formed and shrinkage was mea-
Furthermore, the authors are particularly successful at enu- sured during drying. These were fired in five different sets,
merating the steps they have taken within the chaîne ranging in temperature between 650 and 1,050°C. To assess
opératoire of paste production, both in terms of the processes the degree of spalling, test samples were exposed and moni-
and the underlying rationale for those processes. tored on a beach for 2 months where absorption of atmospher-
The interchangeability of the steps enumerated is not di- ic water vapour could take place. Based on this experiment,
rectly addressed from a chaîne opératoire perspective. As may Rye found that CaO hydration damage among seawater-
be expected of experiments which compare multiple options wetted clay only occurred in samples formed with high sand
148 Archaeol Anthropol Sci (2015) 7:141–149

proportions fired at high temperatures (Rye 1976, 128). Fur- experimental and archaeological publications, but Gosselain
thermore, nearly all of the freshwater samples underwent such himself acknowledges that despite the base levels of salience
significant Ca(OH)2 hydration damage that Rye dismissed the identified within his work, socio-cultural factors can and do
possibility that they could have been used for vessel manu- lead to alterations in potting practise. By investigating and
facture (Rye 1976, 128). Reportedly, ‘most clay-sand mixtures successfully defining the specific chaînes opératoires of ar-
wetted with seawater could be fired safely to most of the chaeological material using experimental archaeology and
temperatures used in the experiment, whereas almost none expanding the analytical scale to search for longer-term pat-
of the mixtures wetted with freshwater were usable over most terns of stability and change; a better picture of technical
of the temperature range’ (Rye 1976, 128–129). practise through time can emerge. That understanding of
This process and function experiment successfully technical practise, and the individual components which blend
incorporated both actualistic and multifaceted ap- together to constitute it, creates a window into the social
proaches supported by Outram (2008) and Schiffer interactions of potting communities. If one considers that
et al. (1994), respectively. Rye does address the steps potting practise can, and frequently is, a domestic activity,
within the chaîne opératoire of raw material processing then it should be clear that this approach is a window into the
which were followed within the experimental protocol social interactions of households as well.
and uses this description of steps to elucidate the It is obviously beyond the ability of every experi-
knock-on effect of raw materials on firing. Although mental archaeology programme to present an expansive
the potential areas of interchangeability of steps are review of the technological context in which one aspect
clear from a chaîne opératoire perspective, this is not of the chaîne opératoire was practised by a specific
addressed explicitly within the article. Rye’s realisations group. But, by incorporating the chaîne opératoire ap-
throughout the experiment highlight areas of specialist proach suggested above, the structured accumulation of
knowledge which affect the communicability of steps complementary information is within reach, and details
within the chaîne opératoire. Most visible to the non- of technological contexts can be more clearly expressed
specialist would be the inclusion of beach sand as a as interrelated. For example, Skibo et al. (1989) argued
tempering material, but the removal of larger calcite- that organic temper use gave way to mineral temper due
bearing particles may be less obvious. Similarly, non- to changing settlement patterns from the Late Archaic to
specialists may not recognise the importance of using Early Woodland. If this is the case, further experiments
saltwater, rather than freshwater. Furthermore, the inter- should demonstrate the presence of a technical negotia-
play between sand proportion, saltwater inclusion and tion as practises changed, and perhaps the alteration of
firing temperature would likely require direct communi- other aspects of the chaîne opératoire, facilitated by the
cation between specialists for successful transmission of social networks of potters. Change to the chaîne
this particular variety of technical practise. opératoire is in fact where social phenomena are most
As it is a general study of underlying principle, this exper- readily apparent; these changes act in the same way as
iment was not contextualised technologically in great detail. dye used for magnetic resonance imaging scans. The
Instead, Rye discusses various archaeological and ethno- route that change takes, its pace, and the changes which
graphic contexts in which salt or saltwater are explicitly become areas of stability within the chaîne opératoire
included as a raw material. through time demonstrate the presence and character of social
connections and boundaries between potting groups. Experi-
mental archaeology’s ability to, in great detail, define aspects
Discussion of the chaîne opératoire ensures that there are many more
opportunities to recognise alterations to potting practise.
The above case studies demonstrate the implicit role that the Through this, it is more likely that the interactions of crafts-
chaîne opératoire approach already has taken within experi- people will be visible in their social contexts: families
mental archaeology. The authors above have been able, with established in new communities, artisans traded by elites,
varying success, to propose social interpretations for the tech- wives moving to their husbands’ villages and potters rep-
nical phenomena highlighted. By shifting from implicit to licating imported styles using local materials. In this
explicit incorporation of this approach, these interpretations respect, the chaîne opératoire and experimental archae-
benefit. ology are natural partners.
A major area of progress for experimental archaeology
stemming from this chaîne opératoire incorporation is the
Acknowledgments I would like to thank the guest editors and two
more successful integration of ethnographic and
anonymous reviewers for their insightful comments on an earlier version
ethnoarchaeological models. The work by Gosselain (2000; of this paper. This research was supported by a Postdoctoral Research
2002; 2008) cited above has been touched upon in many Fellowship from the Fondation Fyssen.
Archaeol Anthropol Sci (2015) 7:141–149 149

References Reynolds PJ (1999) The nature of experiment in archaeology. In: Harding

AF (ed) Experiment and design: archaeological studies in honour of
John Coles. Oxbow, Oxford, pp 156–162
Balkansky AK, Feinman GM, Nicholas LM (1997) Pottery kilns of Roux, V. (1994) La technique du tournage : définition et reconnaissance
ancient Ejutla, Oaxaca, Mexico. Journal of Field Archaeology 24: par les macrotraces. IN Binder, D. & Courtin, J. (Eds.) Terre cuite et
139–160 société. La céramique, document technique, économique, culturel.
Courty MA, Roux V (1995) Identification of wheel throwing on the basis Juan-les-Pins, Editions APDCA, 45–58.
of ceramic surface features and microfabrics. J Archaeol Sci 22: Roux V (2003) A dynamic systems framework for studying technological
17–50 change: application to the emergence of the potter's wheel in
Degoy L (2008) Technical traditions and cultural identity: an the Southern Levant. Journal of Archaeological Method and
ethnoarchaeological study of Andhra Pradesh potters. In: Stark Theory 10:1–30
MT, Bowser BJ, Horne L (eds) Cultural transmission and material Roux V (2008) Evolutionary trajectories of technological traits and
culture: breaking down boundaries. University of Arizona Press, cultural transmission. In: Stark MT, Bowser BJ, Horne L (eds)
Tucson, pp 199–222 Cultural transmission and material culture: breaking down bound-
Gelbert, A. (2003) Traditions céramiques et emprunts techniques dans la aries. University of Arizona Press, Tucson, pp 82–104
vallée du fleuve Sénégal, Paris, Éditions de la Maison des Sciences Roux V (2010) Classification des assemblages céramiques selon le con-
de l'Homme et Épistèmes cept de 'chaîne opératoire': une approche anthropologique de la
Gosselain OP (2000) Materializing identities: an African perspective. variabilité synchronique et diachronique. Les Nouvelles de
Journal of Archaeological Method and Theory 7:187–217 l'Archéologie 119:4–9
Gosselain, O. P. (2002) Poteries du Cameroun méridional: styles tech- Roux V (2011) Anthropological interpretation of ceramic assemblages:
niques et rapports à l'identité, Paris, CNRS éditions. foundations and implementations of technological analysis. In:
Gosselain OP (2008) Mother Bella was not a Bella: inherited and trans- Scarcella S (ed) Archaeological ceramics: a review of current re-
formed traditions in Southwestern Niger. In: Stark MT, Bowser BJ, search. Archaeopress, Oxford, pp 80–88
Horne L (eds) Cultural transmission and material culture: breaking Roux V, Courty MA (1997) Les bols élaborés au tour d'Abu Hamid:
down boundaries. University of Arizona Press, Tucson, pp 150–177 rupture technique au 4e mill. avt J.–C. dans le Levant sud. Paléorient
Harry KG (2010) Understanding ceramic manufacturing technology: the 23:25–43
role of experimental archaeology. In: Ferguson J (ed) Designing Roux V, Courty MA (1998) Identification of Wheel-fashioning methods:
experimental research in archaeology. University Press of technological analysis of 4th-3rd millennium BC oriental ceramics.
Colorado, Boulder J Archaeol Sci 25:747–763
Herbich I (1987) Learning patterns, pottery interaction and ceramic style Rye OS (1976) Keeping your temper under control: materials and the
among the Luo of Kenya. Afr Archaeol Rev 5:193–204 manufacture of papuan pottery. Archaeology & Physical
Kramer C (1997) Pottery in Rajasthan : ethnoarchaeology in two Indian Anthropology in Oceania 11:106–137
cities. Smithsonian Institution Press, Washington Schiffer MB, Skibo JM, Boelke TC, Neupert MA, Aronson MA (1994)
Lemmonier P (ed) (1993) Technological Choices. Transformations in New perspectives on experimental archaeology: surface treatments
material cultures since the Neolithic, London and New York, and thermal response of the clay cooking pot. Am Antiq 59:197–217
Routledge Skibo JM (1992) Ethnoarchaeology, experimental archaeology and infer-
Manem S (2012) The Bronze Age use of caves in France: reinterpreting ence building in ceramic research. Archaeologia Polona 30:27–38
their functions and the spatial logic of their deposits through the Skibo JM, Schiffer MB, Reid KC (1989) Organic-tempered pottery: an
chaîne opératoire concept. In: Bergsvik KA, Skeates R (eds) Caves experimental study. Am Antiq 54:122–146
in context: the cultural significance of caves and rockshelters in Stark, M. T. (Ed.) (1998) The archaeology of social boundaries,
Europe. Oxbow Books, Oxford, pp 138–152 Washington, D.C., Smithsonian Institution Press
Outram AK (2008) Introduction to experimental archaeology. World Stark MT, Bowser BJ, Horne L (eds) (2008) Cultural transmission and
Archaeology 40:1–6 material culture: breaking down boundaries. University of Arizona
Pierret, A. (1995) Analyse technologique des céramiques archéologiques: Press, Tucson
développements méthodologiques pour l’identification des tech- Wallaert-Pêtre H (1999) Manual laterality apprenticeship as the first
niques de façonnage. Un exemple d'application: le matériel du learning rule prescribed to potters. In: Owen LR, Porr M (eds)
village des Arènes à Levroux (Indre). Paris, Université de Paris I- Ethno-analogy and the reconstruction of prehistoric artefact use
Sorbonne. and production. Mo Vince Verlag, Tübingen, pp 185–206

View publication stats