Content uploaded by Anne Mayor
Author content
All content in this area was uploaded by Anne Mayor on Jan 19, 2016
Content may be subject to copyright.
5
ERTUOAYDFKGHJLZICVN
Ethnoarchaeology, Vol. 2, No. 1 (Spring 2010), pp. 5–48.
Copyright © 2010 Left Coast Press, Inc. All rights reserved.
Cr Trais n
Etcty t Nigr
Bn, W Ac
Anne Mayor, Department of Anthropology, University of Geneva,
Switzerland. anne.mayor@anthro.unige.ch
Arac While anthropologists and historians have clearly underlined
the dynamics of human groups, ethnoarchaeologists have emphasized the
stability of modes of transmission of technical knowledge within ethno-
linguistic groups. Overcoming this apparent opposition by mobilizing and
confronting lines of evidence from three distinct disciplines — ethnoar-
chaeology, ethnohistory, and archaeology— allows me to tackle the material
expression of social identities in the past. Time depth, however, coarsens
the resolution of interpretation, necessitating a shift in focus from the
ethnic group to the linguistic family. Research I conducted with the Swiss
MAESAO team since 1988 in central Mali provides a strong case study for
understanding the complex links between ethnicity and ceramics. I propose
a model of interpretation of archaeological ceramics that takes into account
population dynamics.
This article is copyright protected and may not be used for commercial purposes without the written
permission of the publisher. No forwarding, web posting, or further distribution of this PDF is allowed.
6 Anne Mayor
tm dty n cty, as well as population history,
constitute key issues and challenges for archaeology in general, and African ar-
chaeology in particular. In addition to the traditional epistemological limits of in-
terpretation in archaeology, and poor understanding of the material expression
of ethnological concepts makes the study of relationships between past material
artifacts and their socio-cultural meaning difficult.
Archaeologists have long questioned whether a correlation exists between
material culture and identity and whether artifact variability provides a key to
interpretation of past social boundaries (Shennan 1989; Stark 1998; Jones 1997).
During the last 20 years, ethnoarchaeologists have addressed the subject of iden-
tity and material culture through studies of ceramics, which are one of the best-
represented materials at archaeological sites since the Late Stone Age (see among
others Arthur 2006; Bowser 2000; Costin 2000; David et al. 1988; Dietler and
Herbich 1994; Frank 1998; Gosselain 2002; Kramer 1985; Longacre 1991; Rice
1987; Roux and Corbetta 1990; Skibo et al. 1989; van der Leeuw and Pritschard
1984). In particular, ceramic researchers have focused on describing the learn-
ing processes and diversity in pottery manufacturing processes and variability
in products, in order to understand better the meaning of technical and stylistic
choices.
At present, the challenge is to overcome an apparent contradiction. On one
hand, for the past 20 years anthropologists and historians have debated the com-
plexity of identity construction and maintenance (Amselle and M’Bokolo 1985;
Barth 1969; Jenkins 1997). On the other hand, ethnoarchaeologists have noted
the stability in the transmission of technical knowledge within a group: most
pottery shaping techniques require close and long-lasting relationships between
teacher and student, which are typically familial, explaining the transmission of
knowledge within an ethnolinguistic group and the cultural specificity of craft
traditions (Arnold 1985; Gallay and de Ceuninck 1998; Gosselain 2000).
My objective in this article is to demonstrate that the study of systematic
links established in the present between ceramic traditions and their meaning
enriches the interpretation of the regional archaeological record and makes it
possible to propose models for population history. Specifically, I will propose an
archaeological “interpretative tool” in the domain of ceramics, based on stud-
ies of the present, capable of addressing the material expression of processes of
identity construction and social interactions in ceramic assemblages from the
past. In a situation of cultural continuity, the context of application is primarily
Ceramic Traditions and Ethnicity in the Niger Bend 7
at a regional scale. Nevertheless, archaeologists can use this approach in different
geographic contexts.
To determine whether variability in material culture reflects past socio-cul-
tural boundaries requires the use and comparison of data from three different
disciplines. First, I present an ethnoarchaeological analysis of correlations be-
tween current material features and ethnolinguistic identities, at the level of pro-
duction, use, and diffusion of pottery. Second, I review ethnohistorical sources
situating the development of pre-colonial state formation, the historical depth of
groups, and the main events that influenced the development and relationships
between neighboring groups. ird, I outline an archaeological analysis of sites,
occupation chronology, main techno-economic and cultural features, as well as
ceramic characteristics.
My ethnoarchaeological, ethnohistorical, and archaeological field research
focus on the Inland Niger Delta and Dogon Country in Mali, but my synthesis
takes into account the whole Niger Bend, an immense territory encompassing
650,000 square kilometers including central Mali, southeast Mauritania, south-
west Niger and northern Burkina Faso (Figure 1).
Figure 1. Map of the Niger Bend, showing the sites mentioned in this study. (All images in this
article © Mayor et al. 2005.)
8 Anne Mayor
Most of the ethnoarchaeological materials I present in this paper derive
from the first research project (1988-1996) of the MAESAO (Mission Ethnoar-
chéologique et Archéologique Suisse en Afrique de l’Ouest) (Gallay and Huysec-
om 1989; Gallay et al. 1990, 1996, 1998; Huysecom 1992; Mayor 1996, 1997). is
project include studies of current ceramic traditions among 12 ethnic groups
and of past ceramic assemblages from excavations at the historical sites of Ham-
dallahi and Modjodje, linked to the Fulani Empire of Maasina during the 19th
century. I also collected new ethnoarchaeological, ethnohistorical and archae-
ological material as part of the international program “Human settlement and
paleoclimatic evolution in West Africa,” focusing on cultural and environmental
evolution of the Dogon Country in Mali from Paleolithic to historical periods
(Huysecom 2002; Mayor 2003; Mayor et al. 2005).
Initial studies demonstrated that 19th century ceramic traditions were very
similar to present day ceramic traditions and that it was possible to go further in
the interpretation of historical sites using a contemporary approach, guided by
rigorous methodology. is article presents the results of my doctoral research
(Mayor 2005, 2006, in preparation) and aims to demonstrate that it is possible to
link ceramic technology and style with ethnolinguistic groups of the past pre-dat-
ing the 19th century, when combining ethnohistoric and archaeological studies
with ethnoarchaeological studies conducted in a long-term project and a vast area.
Etnorchaloc Aroac
Geographic Context and Ceramic Production
Today, the Niger Bend is inhabited by many ethnic groups with complementary
ways of life, focused on fishing, herding, or agriculture. ese include the Tuareg,
Maure, Songhay, Fulani, Soninke, Bozo, Somono, Bamana, Dogon, Bwa, Kurum-
ba and Mossi groups (Gallais 1980). Originating in different regions, their lan-
guages belong to several different linguistic families (Blench 2005; Ruhlen 1987,
table 1). In general, anthropologists recognize identity on the basis of economic
specialization, control of resources (water, land, or grassland), habitus, languag-
es, family names, and associated territories, all which change constantly through
time (Bouju 1995; Fay 1997). Current ethnic and linguistic diversity results from
a complex history of migration, alliances, subjection, social differentiation, and
economic and craft specialization. e establishment of most of these popula-
tions occurred primarily over the last 2,000 years by processes of “stacking up” or
Ceramic Traditions and Ethnicity in the Niger Bend 9
superimposition of groups and redefinition of authority and power, rather than
via complete displacement (Fay 1997; Pelzer 2003).
In the Niger Bend region, potters are members of nearly all ethnic groups,
except for the Bozo fishermen. Ceramic production is an activity mostly prac-
ticed by women in the craft castes. Today, a majority of the people of the Niger
Bend region stratify their communities into noble food-producers (cultivators
or herders), endogamous artisanal castes (blacksmiths, weavers, woodcarvers,
jewelers, griots [a West African individual who recounts oral history], etc.) and
former slaves. Potters often marry smiths, but among the Fulani, they are also
the wives of weavers, woodcarvers and leather workers (Gallay et al. 1998; Gardi
1985; LaViolette 2000). Somono potters may belong to the smiths’ craft or not,
and in some Dogon villages, wives of cultivators or smiths may produce ceram-
ics (Bedaux 1986; Gallay and Sauvain-Dugerdil 1981; Gallay et al. 1998). Potters
usually sell their own wares via local village markets or travel by canoe on sales
trips to distant villages.
In contrast to other African regions where ceramic production is nearly
extinct, in central Mali numerous people still use pottery to fill nearly all their
Table 1 Localization and language of the main ethnic groups of the Niger Bend
Ethnic
group
Language and
language sub-family Language familiy Country
Tuareg Tamasheq (Berber) Afro-asiatic Mali, Burkina Faso, Niger, Mauritania,
Algeria, Lybia
Moor Hassaniyya (Arabic) Afro-asiatic Mali, Senegal, Mauritania, Algeria,
Marocco
Songhay Songhay (Nilo-Saharan
or Mande)
Nilo-Saharan or
Niger-Congo
Mali, Niger, Burkina Faso, (Benin,
Ghana, Nigeria)
Fulani Fulfulde (Atlantic) Niger-Congo From Senegal to Sudan
Soninke Soninke (Mande) Niger-Congo Mali, Mauritania, Senegal, (Burkina
Faso, Ivory Coast)
Bozo Bozo (Mande) Niger-Congo Mali
Somono Somono (Mande) Niger-Congo Mali
Bamana Bamanakan (Mande) Niger-Congo Mali, (Burkina Faso)
Dogon Various Dogon (Dogon) Niger-Congo Mali, (Burkina Faso)
Bwa Bwamu (Gur) Niger-Congo Mali, Burkina Faso
Kurumba Kurumba (Gur) Niger-Congo (Mali), Burkina Faso
Mossi More (Gur) Niger-Congo (Mali), Burkina Faso
10 Anne Mayor
domestic vessel needs including water transport and storage, cereal and sauce
cooking, preparation of millet beer, and special functions such as seats, drain-
pipes, toys, braziers, hearths, incense burners, etc. However, people are increas-
ingly replacing ceramic vessels with glass, aluminium, plastic and enamel (Cun-
ningham 2005). In the past few decades, some potter lineages have discontinued
their ceramic production. Many others reduced the amount of decoration on
their pots, and some abandoned certain tools used for specific types of impres-
sions (Mayor 1994, 2005).
Current Variability in Ceramic Traditions
e extensive ethnoarchaeological research conducted by members of the
MAESAO team in central Mali provides the main frame of reference for a model
that links pottery and social meanings. We observed and interviewed 330 potters
and hundreds of pottery market-buyers in 12 different ethnic groups, which in-
cluded 12 markets, 86 villages and abandoned sites, 102 household compounds,
37 pottery workshops, descriptions of 263 pottery shaping sequences, and draw-
ings and documentation of more than 4,000 pots.
In a first stage, we identified 13 ceramic traditions from this array of doc-
umentation (Gallay et al. 1996, 1998; Huysecom and Mayor 1993). A ceramic
tradition is a group of technical, morphological, and decorative characteristics
specific to the production of potters belonging to a social group defined by a
matrimonial network, whether this is a group of cultivators or an endogamous
craft caste. We characterize these traditions by one or more chaînes opératoires,
differentiated by shaping techniques used to make the preform and by the differ-
ent tools used to support their work. Four broad classes of techniques to shape
pottery preform exist: pounding in a concave form (also called tamper and con-
cave anvil technique), molding over a convex form, hollowing and drawing of
a lump of clay, and molding in a concave form over a mobile plate (Figure 2).
Potters in the region use an array of different decorating techniques like rolled,
dragged or direct impressions, incisions, paintings and added plastic elements.
In a second stage, I refined this qualitative description of ceramic variability
with quantitative analysis of 33 technological, morphological, and ornamental
variables (Mayor 2005, in preparation). GIS maps (Figure 3) show the ethnic dis-
tribution of potters who produced the ceramics (production level), as well as
the spatial distribution of the pottery used in household compounds after being
purchased by consumers (consumption level).
Ceramic Traditions and Ethnicity in the Niger Bend 11
Regularities and Mechanisms
e MAESAO researchers identified several recurrent or law-like patterns,
called “regularities” in French ethnoarchaeology, as well as their explanatory
mechanisms, through analysis of the links between material objects and their
meaning. ree regularities, detailed below, are particularly significant for the
present commentary, and make it possible to select the most relevant criteria in
terms of cultural identity.
Figure 2. Chaînes opératoires for the four main ceramic shaping techniques in the Niger Bend.
12 Anne Mayor
Pattern 1: Shaping techniques and aesthetic properties of pottery define tradi-
tions that reflect the identity of the producers. e most significant elements for
the ceramics-ethnicity relationship concern the technical choices made during
preform shaping and the choice of decorating tools, particularly manufactured
Figure 3. Current distribution map, in the Inland Niger Delta, for the technique of molding in
a concave form over a mobile plate, showing the ethnic distribution of its production, the geo-
graphic distribution of its consumption, and the territory occupied by the Somono.
Ceramic Traditions and Ethnicity in the Niger Bend 13
tools made for rolled or direct impression. In contrast, potters usually make
the upper section of a vessel with coils. e clay paste shows little variation,
as it always contains grog temper, sometimes complemented with an organic
temper (cereal husk, dung) in the Inland Niger Delta or mineral temper (sand,
laterite grains) in the Dogon Country (Gallay et al. 1998). Potters typically fire
their pots in direct contact with the fuel and we observed the use of kilns only
among groups living in the emerging margins of the Delta (Mayor 1999). Each
tradition uses only one or two of the main techniques for preform shaping.
Manufactured tools for ceramic decoration, such as braided cord roulettes,
braided strip roulettes, carved wooden cylinders, sculpted wooden awls and
stripped organic cobs each require specific knowledge and correlate to differ-
ent ceramic traditions (Mayor 2005).
e mechanism that stimulates Pattern 1 is the ceramic learning tradition.
Learning shaping techniques is a long process and requires acquisition of motor
skills and expertise, which implies close and durable familial relationships be-
tween teacher and student. is process explains the transmission of knowledge
within an ethnolinguistic group and the cultural specificity of the traditions.
Technical borrowing is present but uncommon, and as a consequence the ho-
mogenization of traditions is slow (Gallay and de Ceuninck 1998). An historical
dynamic over several centuries marked by several kinds of borrowing explains
the current combination of techniques and ethnolinguistic units. However, de-
spite several centuries of interethnic contact and different processes of technical
borrowing, ceramic traditions remain quite distinct today.
Pattern 2: Pottery within a housing unit reflects the identity of the occupants.
While ceramic assemblages are well-differentiated at the level of production, the
situation is less clear at the level of consumption. Ethnoarchaeological research
demonstrates that the blending of ceramic traditions within compounds is less
important when the compound is located in a village outside the zone of ethnic
superposition and when a family of potters inhabits the compound (Gallay 1992;
Gallay and de Ceuninck 1998). Despite this blending, statistically the ceramic
assemblages ethnoarchaeologists found in households generally reflected the
identity of consumers.
e mechanism that underlies Pattern 2 is ethnicity or group membership.
In a multiethnic context, the market economy encourages the acquisition of pot-
tery produced by potters from other ethnic groups, but statistically, except for
14 Anne Mayor
the Fulani case, users tend to buy most pottery produced by craftspeople of their
own group.
Pattern 3: e spatial distribution of a ceramic tradition reflects the settlement
structure of the producing group. Based on ceramic inventories documented in
the compounds of many villages, I created maps to show the territories of settle-
ment units using material elements accessible to the archaeologist. e maps,
constructed with a GIS system, indicate the spatial distribution of each variable
and its meaning in cultural terms (Figure 3).
e cumulative effect of socio-economic mechanisms of production, cir-
culation, and acquisition of pottery create the spatial structure of a ceramic
tradition. In other words, when women from endogamous and patrilocal craft
classes produced ceramics, the ceramic production zone corresponds to the
matrimonial network of potters and thus, by extension, to the zone occupied by
their ethnic group. In rural economies, the village market transactions generate
a consumption zone for the circulation of pottery beyond the original produc-
tion zone, thus extending beyond the limits of the ethnic group (Burri 2003; de
Ceuninck 2000; Gallay and de Ceuninck 1998).
Dsci
e ethnoarchaeological approach demonstrates that several ethnolinguistic
groups and shaping techniques exist in the Niger Bend today. eir correlation
is particularly clear in zones where settlement territories are distinct and overlap
only slightly, but is less so in mixed settlement zones, due to phenomena of tech-
nical borrowing. e Inland Niger Delta is typical of a mixed settlement zone,
and intense contacts between potters of different groups may be at the origin of
technical innovation and diversity.
Borrowing occurs in cases of contact between potters from different groups
(for example at markets), marriages between crafts people of different castes,
social restructuring or geographical isolation of part of a group that is no lon-
ger included in the previous matrimonial network. In this context, some simple
techniques of shaping, such as molding over a convex form, are more widely
spread than others that require more advanced psycho-motor skills, acquired
since infancy.
But despite the existence of interethnic contacts over several centuries,
we can identify quite distinct production groups today. is demonstrates the
Ceramic Traditions and Ethnicity in the Niger Bend 15
strength of the relationship between the shaping technique practiced and the
ethnolinguistic identity of the craftsperson, and the long period necessary for all
groups to adopt a technique uniformly.
Recognition of these shaping techniques on small potsherds is not simple.
But in the Niger Bend, gestures and the tools used during the chaîne opératoire
leave traces on the internal and external surfaces of the pots that are specific
to each shaping technique, such that it is possible to observe them on ancient
pottery. Following an initial description by Huysecom (1994), several studies in
other African cultural contexts have contributed to building a frame of reference
for the identification of pottery shaping techniques, very useful for archaeolo-
gists (Gelbert 2003; Livingstone Smith 2001).
In contrast, morphological elements are not well correlated ethnically and
geographically, since most of the traditions of central Mali are characterized by a
majority of pots having a round base and non-carinated body, and by a minority
of vessels that are footed or have a carinated body. Morphology and dimensions
are indeed very informative for determining the function of the pots (de Ceun-
inck 1994), but not for determining cultural identities.
Decorating techniques have varying ethnic and geographic correlations.
Rolled and dragged impressions, made with manufactured tools show a good cor-
relation with the different traditions. With the exception of twisted cord impres-
sions, which potters use in nearly all traditions, each roulette impression marks
only one or two traditions out of the 13 identified, with minor borrowing found
in one or two other traditions. Direct impressions using specific tools, such as
sculpted wooden awls, combs, nets or mats, also show clear ethnic correlation.
Like roulette impressions, they are each specific to only one or two traditions of
the 13, again with minor borrowing observed in a few other traditions. Incisions
are generally somewhat less ethnically correlated than impressions, while paint-
ing and added plastic decorations show poor ethnic correlation. Some decora-
tions are common, decorating most of the vessels within a tradition, while others
are rare, used only on a small percentage of the vessels.
ere are few unequivocal associations between a decoration technique and
a tradition because stylistic borrowing of decorations is easier and more com-
mon than borrowing of shaping techniques. Despite this, the situation in the
Niger Bend, after several centuries of cohabitation, still indicates strong ethnic
correlations for some decoration types made with manufactured tools such as
roulettes.
16 Anne Mayor
Etnrc Aroac
e second frame of reference useful to archaeological interpretation of ceramic
traditions employs historical data of the Niger Bend, with a special focus on the
Inland Niger Delta and the Dogon Country. Unlike other regions of Africa, the
Sahelian zone, including the Niger Bend region benefits from the contribution of
numerous Arabic written sources, whether external or internal, which describe
the history of sovereigns, cities and conflicts (Levtzion and Hopkins 1981). But
vast regions pass for “countries without history,” for which history must be recon-
structed using other sources, primarily oral tradition. In these non-centralized
societies, historical data go back only two or three centuries, and oral memory
can reach remote periods only in the context of myths and legends.
Ethnic groups in their current formation result from different histories, di-
rectly or indirectly associated with the development of the precolonial empires of
the Niger Bend (Figure 4). ese precolonial states mainly included Ghana (be-
ginning between the 5th and the 8th century, ending between the 11th and the
13th century, Dieterlen and Sylla 1992; Togola 1996), Mali (13th-16th century,
Ly-Tall 1977, 1985), Songhay (15th-16th century, Cissoko 1975, 1985; Dramani-
Issifou 1993; Holder 2001; Insoll 1996; Lange 1994), the Moroccan occupation
(end of the 16th century, Abitbol 1979; Dramani-Issifou 1982), the Mossi king-
doms (13th-19th century, Izard 1985), the Bamana kingdom of Segou (18th-19th
century, Konare Ba 1987; Roberts 1981), the Diina of Maasina (19th century, Ba
and Daget 1984; Sanankoua 1990), and the Futanke hegemony (19th century,
Ly-Tall 1996; Robinson 1988), ending with French colonization (end of 19th cen-
tury-1960, Roberts 2005).
Historical sources, read carefully and taking into account the filters em-
ployed in viewing, manifesting and creating ethnicity depending upon the au-
dience (Jenkins 1997), indicate that current groups have historical depth that
varies greatly. eir ethno-genesis depends on the historical dynamics of succes-
sive migrations and diverse power relationships (alliances, dependencies, subjec-
tions, etc.), that contributed to creating or reinforcing economic specialization
and cultural distinctions (Fay 1997, Kassibo 1994).
e following ethnohistorical synthesis gives an idea of the complexity of
population history, as well as the processes of ethno-professional differentiation,
and chronologically orders the main events responsible for the current distribu-
tion of the groups (Figure 5). Ethnohistorical accounts suggest in different re-
Ceramic Traditions and Ethnicity in the Niger Bend 17
gions the presence of an early substrate population, considered autochthonous
but not precisely dated. e autochthonous populations include the Bozo/Nono,
a fishing and rice growing group affiliated with the Soninke in the Inland Ni-
ger Delta (Mande linguistic family), the Bwa, essentially cultivators and hunters
on the dry lands of the southeast margin (Gur linguistic family), and the Proto-
Songhay, fishermen, hunters and cultivators along the Niger River to the east
(Nilo-Saharan or Mande linguistic family).
Over time, this substrate was overlain by successive incursions of popula-
tions (Soninke, Mandinka, Fulani, Dogon, Mossi, Bamana, Tuareg, some compo-
nents of Songhay, etc.). Contacts between them led to the ethnic differentiations
and economic specializations observed today. In the Inland Niger Delta, stability
in the overall socio-economic structure developed after the arrival of the Arbe, a
Fulani group, at the end of the 14th century, and this structure remained globally
the same until the end of the French colonial period (Fay 1997).
In the Dogon Country, new populations of cultivators arrived in the 13th-
Figure 4. Maximum spatial and chronological extension of the main pre-colonial states in the
Niger Bend. Crescents indicate the approximate date of introduction of Islam.
18 Anne Mayor
15th centuries (Gallay et al. 1995; Mayor et al. 2005). In the 15th century, the
Bamana cultivators from the south-west, colonized new lands in Mema and the
northern Inland Niger Delta (Togola 1996), the Fulani migrated from west to east
and south of the Niger bend (Fay 1997), the Tuareg herders migrated from the
north to Gourma and Oudalan (Pelzer 2003), and the Mossi kingdoms emerged
in Yatenga, dominating or driving back pre-established groups (Izard 1985).
Figure 5. Synthetic map of the main historical population movements leading to the establish-
ment of current ethnic groups in central Mali (after Quensière ed. 1994, supplemented).
Ceramic Traditions and Ethnicity in the Niger Bend 19
Alongside the history of food-producers, the history of migrations of crafts
people importantly contributes elements to understanding the evolution of ce-
ramic traditions. e migration of crafts people occurred after the formation of
endogamous castes probably in the 13th century (Tamari 1991). e origin and
history of the castes remains, however, poorly known. Initially, artisans engaged
in ceramic production probably belonged to families characterized by a tendency
towards endogamy, which through time became strictly endogamous castes. e
phenomenon of endogamy among artisans may be a factor in the transcultural
diffusion of techniques with new rules of marriage having required the creation
of new networks of relationships (Gosselain 2000; Sterner and David 2003). Nev-
ertheless, in the Niger Bend, the relationship of interdependence between craft
castes and the ethnic groups of nobles is strong, and marriages between artisans
from castes of distinct groups remain rare.
is brief summary shows that some populations have roots for which their
historical depth goes back several centuries, more than a millennium for some in-
digenous populations, even if they have not come through history in linear, mono-
lithic or independent fashion. Ethnohistory also suggests that the main break in the
population history of the Niger Bend took place between the 13th and 15th cen-
turies during the Mali Empire, after the arrival of new populations of herders and
cultivators, who for the most part superimposed themselves on preceding peoples.
e complex history of the Niger Bend makes it a challenge to reconstruct a
history of ceramic traditions in line with ethnolinguistic entities of the past, but
deep local roots of some populations as well as the process of superimposition
of people through time makes it possible, unlike other regions characterized by
a total replacement of groups.
Archaloc Aroac
e final frame of reference is that of archaeology. My synthesis of the archaeol-
ogy of the Niger Bend during the period of pre-colonial empires aims to show
the broad trends of population settlement by summarizing elements that are rel-
evant to the issue of population history and the evolution of ceramic traditions
(Mayor 2005, in preparation). I have gathered archaeological data from excava-
tions (Figure 1) conducted over the last 30 years in Dhar Nema (Mauritania),
Mema, the Inland Niger Delta, the Lakes Region, the Middle Niger between
Tombouctou and Niamey (Mali and Niger), Oudalan (Burkina Faso), Yatenga
20 Anne Mayor
(Burkina Faso) and the Dogon Country (Gourma-des-Monts, Cliff, Plateau and
Seno Plain in Mali). e Niger Bend is unquestionably an African region that has
been the subject of much archaeological research, which has contributed funda-
mental elements to understanding complex societies (Insoll 2000; MacDonald
1994; Magnavita et al. 2002; McIntosh S. 1999). e Swiss team conducted some
of these excavations in the Inland Niger Delta (Curdy 1982; Gallay et al. 1990;
Mayor 1996) and the Dogon Country (Gallay and Sauvain-Dugerdil 1981; Huy-
secom 2002; Mayor et al. 2005; Ozainne et al. 2009).
In the current state of research, a discontinuity characterizes settlement dur-
ing the first millennium B.C.E. in most Sahelian regions, obscuring the transition
between the Neolithic and the first populations who mastered iron metallurgy, and
reflecting economic change as well as the establishment of new groups in previ-
ously uninhabited regions (Breunig and Neumann 2002; Mayor et al. 2005). Next,
between the first millennium B.C.E. and the first millennium C.E., two migrations
were probably responsible for a large component of the protohistoric populations
in the Niger Bend, one from the northwest with the complex societies of the Dhars
of Mauritania (MacDonald 1996; MacDonald et al. 2004, 2009) and the other
from the northeast with populations previously mastering iron metallurgy in Ni-
ger (Vernet 1996). Archaeologists interpret these movements south as retreats in
response to the strong aridity marking this period. Populations practicing comple-
mentary economies including fishing, hunting, herding, agriculture, iron metallur-
gy, interregional (or even long-distance) commerce, and sometimes-hierarchical
social organization became established across the Niger Bend.
e period of optimum humidity covering the second half of the first millen-
nium C.E. included the development of urbanization, intensive iron production
between local needs and the emergence of the first empires (Ghana to the north-
west and the proto-Songhay dynasties of the Ja to the east, along the Middle
Niger river valley).
Between the 13th and 15th centuries, people abandoned the sedentary tells
that they had occupied for several centuries or millennia in the Inland Niger
Delta, Mema, Middle Niger region and northern Burkina Faso (Arazi 1999; von
Czerniewicz 2004; MacDonald 1994; Magnavita et al. 2002; McIntosh R. 1998;
Raimbault and Sanogo 1991). At the same period, new populations of herders
and cultivators migrated and colonized most of the regions of the Niger Bend,
adopting new patterns of territorial occupation in a period of progressive drought
(Pelzer 2003; Togola 1996).
Ceramic Traditions and Ethnicity in the Niger Bend 21
During the second millennium, different regions of the Niger Bend simul-
taneously show the use of several different funerary practices, ceramic shaping
techniques, and types of roulette impressions, revealing clear cultural diversity.
In view of the socio-economic complexity and population diversity demon-
strated by archaeological discoveries, I argue that mechanisms of production,
consumption, and diffusion of pottery, observed in the present, were already ac-
tive during the protohistoric and historical periods in the Niger Bend and may
explain variability in ceramic assemblages from archaeological sites.
Nevertheless, there are limits in the comparative study of archaeological
pottery. Based on the ethnoarchaeological studies discussed above, the most rel-
evant elements for determining the identity of ceramic producers are those that
relate to shaping techniques and certain decorative techniques, mostly roulette
impressions. But archaeological studies of shaping techniques are often lacking
due to several factors including fragmentation of the ceramics, mixing of strati-
graphic units due to the tell-type sites excavated, impressive quantity of sherds
that are not often associated with specific structures, and the analytic focus on
rims rather than the body of the vessel. Finally, problems of identification, de-
scription, and illustration of the ceramic decoration in Africa are so common
that they render comparisons between sites and regions difficult. Hopefully, the
publication of a field identification manual for African impressed pottery will
reduce these problems (Haour et al., in press).
Despite these limitations, I synthesized the abundant archaeological data
available and reconstructed the spatial-temporal distribution for some criteria
selected for their cultural accuracy in the Niger Bend (Figure 6; Mayor 2005,
2006, in preparation).
What are the Correlations between Ceramic Features
and Ethnic Groups?
By intersecting data from the three frames of reference — ethnoarchaeology, eth-
nohistory, and archaeology — it is possible to retrace the historical depth and
cultural evolution of shaping techniques and several roulette impression decora-
tions by moving from the present to the past. Given the source data available, it
is clear that these are only hypotheses to guide research and to be re-evaluated in
light of future archaeological discoveries.
22 Anne Mayor
Shaping Techniques
Today, there are 4 different ceramic shaping techniques, including molding over
a convex form, hollowing a lump of clay, molding in a concave form, and pound-
ing in a concave form (Figure 2), which can be traced historically through archae-
ology and oral history and correlated to various ethno-linguistic groups through
time and space.
e technique of molding over a convex form (Figure 7)
e technique of molding over a convex form is one of the least ethnically corre-
lated today in the study area, since it is practiced by the Soninke and the southern
Bamana exclusively, by the northern Bamana, the Dogon tradition C and the
southern Fulani alongside other techniques, and by some Songhay, northern Fu-
lani, and southern Somono on an occasional basis. However, several local studies
Figure 6. Archaeological distribution map of the dierent techniques for shaping ceramic
preforms in the Niger Bend.
Ceramic Traditions and Ethnicity in the Niger Bend 23
(Gallay 1970; Gelbert 2003; Raimbault 1980) and more general studies on the
African continent (Drost 1967; Gosselain 2000, 2001) show that this technique is
preferentially associated with Mande-speaking groups. Nevertheless, some Gur-
speakers like the Gurmantche and the Nuna of Burkina Faso and some Senoufo
of the Ivory Coast also use this technique. A study of the phenomenon of cul-
tural borrowing in the Senegal River Valley (among Soninke and Tukulor groups)
shows that potters easily learn this technique and rapidly pass it from one group
to another through simple contact (Gelbert 2001, 2003). is phenomenon
Figure 7. Cultural history in the Niger Bend for the techniques of molding over a convex form,
hollowing of a lump of clay in a wooden plate, hollowing of a lump of clay in a clay mould and
molding in a concave terracotta mould over a mobile plate.
24 Anne Mayor
explains the current distribution of this technique among several non-Mande-
speaking ethnic groups neighboring the Soninke and the Bamana.
is technique has a history of at least 4,000 years in the Niger Bend. Its his-
tory and relationship to the formation of ethnolinguistic groups in the Niger Bend
followed several stages. First, an early common origin probably came out of the Sa-
haran Neolithic during significant north-south population movements following
the onset of current aridity as indicated by the site of Kobadi in Mali (Gallin 2001-
2002). Second, there is a progressive correlation with groups of Mande speakers
(proto-Soninke), who differentiated from this substrate west of the Niger Bend.
e first inhabitants of the Lakes Region and the Inland Niger Delta can be associ-
ated with this ethnogenesis. ird, there is a possible correlation with indigenous
groups perhaps belonging to the Gur linguistic group, which separated from the
earlier substrate south and east of the Niger Bend. Finally, because this technique
is easily learned, potters often borrowed the technique resulting in its cultural dis-
persal and its progressively weakening correlation with Mande speakers.
e technique of hollowing a lump of clay (Figure 7)
e technique of hollowing a lump of clay is practiced in the study zone only by
the Bwa (indigenous Gur-speaking group) and by the wives of Dogon smiths of
the Irin clan who practice the Dogon D tradition (Gallay et al. 1998). More gener-
ally, this technique is used by several southern West African groups in the Niger-
Congo family, including Mande (some Bamana), Gur (Lobi, some Senoufo) and
Kwa (Baoule, Aboure) speakers. Gosselain (2000) suggests that this technique
may represent the earliest tradition in the Mande area as the Mande “proto-tech-
nique”. Later people replaced this technique with the technique of molding over
a convex form in a geographic distribution that roughly corresponds to that of
the Mali Empire, which historians also assume was responsible for the spread of
endogamous craft castes system.
Unfortunately, archaeological evidence of the technique of hollowing a lump
of clay is still rare outside the Dogon Country, where it seems to be present since
the 4th-3rd centuries B.C.E. on the Cliff, as well as in northern Burkina Faso, where
it seems to have been used at least since the first century C.E. Current Dogon oral
traditions frequently attribute abandoned sites to the Bwa. I thus propose the hy-
pothesis that the ancestors of the current Bwa were among the first inhabitants
Ceramic Traditions and Ethnicity in the Niger Bend 25
mastering iron metallurgy in the Dogon Country, who were then gradually pushed
back to the southwest periphery of the rocky massif. e Jeme-irin blacksmiths,
whose wives are currently responsible for the production of ceramics of the Do-
gon D tradition, settled on the Bandiagara Plateau later, probably during the 16th
century (Robion-Brunner 2008). ese potters probably borrowed their technique
from the neighboring Bwa technique, and adapted it introducing the use of a mas-
sive raw clay anvil as support instead of a wooden plate.
If this technique was early in West Africa, only the study of new Neolithic
and early protohistoric assemblages can provide support for it. Based on our
current state of knowledge, this technique has a long history in the Niger Bend,
but there is no evidence for it either past or present in the Inland Niger Delta.
us, I consider the hypothesis of generalized replacement of the convex mold-
ing technique by the hollowing of a lump of clay technique during to the Mali
Empire unlikely. Instead, I offer the hypothesis of co-existence of two techniques
among the populations of the Niger-Congo linguistic family, one to the north in
connection with Mande populations, the other to the south in connection with
Gur and other populations. is technique was present in a large zone including
northern Burkina Faso and southern Mali, as well as the Dogon Country.
e technique of molding in a concave form (Figure 7)
e technique of molding in a concave form over a mobile plate correlates well
with a specific ethnic identity since only the Somono potters, who may or may
not be part of the smith caste, practice the technique. e very rare borrowings
of this technique occur in the cases of mixed marriages with the Bamana or the
Fulani. Looking at a broader geographic range, it is of interest to note that this
elaborate technique requiring a mobile support, a mold, and very specialized
tools and producing high quality pottery, is known nowhere else in sub-Saharan
Africa.
Archaeologically, the technique also appears to be highly geographically
limited and centered on the southern Inland Niger Delta, where it is clearly iden-
tifiable on large funerary jars; but potters also used the technique to make other
ceramic forms. e Mema potters also may have practiced this technique, but re-
search has not yet confirmed this. is technique is more recent than the others.
Archaeological evidence suggests potters used this technique as early as the 5th
26 Anne Mayor
century and definitely by the 11th century on (Bedaux et al. 1978; Curdy 1982;
McIntosh S. 1995; McIntosh and McIntosh 1980). ese dates seem to be prior
to the rapid development of trans-Saharan commerce and render the hypothesis
of introduction from the Maghreb unlikely. ese dates are also prior to the cre-
ation and spread of endogamous craft castes during the Mali Empire according
to Tamari’s (1991) hypothesis.
I must thus consider the possibility of local evolution during the devel-
opment of techno-economic specialization and urbanization, two phenomena
that took place during the first millennium at the time of the Ghana Empire.
is situation raises the issue concerning the identity of these artisans. Ac-
cording to most historical sources, it seems that the Somono identity appeared
after the 16th century from an aggregate of local populations of multiple ori-
gins (Bozo, Dogon, Marka, Bwa, artisans, etc.) to serve the local imperial pow-
ers, particularly as boatmen in the context of the Bamana kingdom of Segou
(Gallais 1980; Roberts 1981). Considering this, the technique of molding in a
concave mould over a mobile plate was thus created before the formation of
the Somono group, by artisans living in the southern Inland Niger Delta and
making pottery used by Bozo fishermen and Nono rice growers. eir descen-
dants were probably massively integrated, around four centuries ago, into the
newly formed Somono group. It is, however, difficult to know whether these
artisans belonged to a group of Mande speakers, like the Bozo and the Nono,
or whether they were ancestors of the Bwa, whom Soninke immigrants in oral
histories claim were the original inhabitants at the time of the foundation of
the new Djenne. According to this last hypothesis, these artisans developed
their technique of hollowing a lump of clay by replacing the wooden plate with
terracotta moulds of various sizes, and by inventing the highly mobile plate
device. Only detailed technical study of early pottery in the Inland Niger Delta
and Mema will allow better understanding of the emergence of this particular
technique.
e technique of pounding in a concave form (Figure 8)
e technique of pounding in a concave form today is widespread by potters
living across the Sahelian band, with the Inland Niger Delta forming its extreme
western extension (Huysecom 1996). According to Sterner and David (2003),
this technique originates between Chad and Sudan, as an early form integrating
Ceramic Traditions and Ethnicity in the Niger Bend 27
Figure 8. Cultural history in the Niger Bend for the techniques of pounding in a concave form
with various supports (type 1 pestles: truncated conical; type 2 pestles: stumpy cylindrical).
28 Anne Mayor
a simple depression hollowed in the ground, covered or not by a mat, and the use
of a stone pestle. Nilo-Saharan potters, migrating east and west of the Lake Chad
Basin, then began using a truncated conical clay pestle (type 1) during the first
millennium B.C.E. Sterner and David assume that it arrived in the Niger River
region with Songhay populations during the first millennium C.E. and would
have gradually spread to their neighbors during the second millennium. Other
groups invented the stumpy cylindrical clay pestle (type 2).
Today, in the Niger Bend region, the technique of pounding in a concave
form is widespread and potters use at least six different supports on which
they pound the clay and three kinds of pestles. Mainly Songhay but also some
northern Fulani and northern Bamana potters use the type 1 pestle on a diago-
nal mat of palm leaves and practice the technique using a wooden anvil, while
some northern Somono potters use it on a rounded raw clay anvil. In contrast,
wives of Dogon cultivators (A tradition) use a stone pestle and a straight mat
made of corded strands of baobab fibers along the Bandiagara cliff, while wives
of smiths among the Mossi, the Kurumba and the Dogon (B tradition) use the
type 2 clay pestle with a fired anvil or a hollow in the ground in Yatenga and
the Seno Plain.
According to the present state of archaeological knowledge in the Niger
Bend, the evolution of this technique is as follows. It is likely that practice of the
technique of pounding in a concave form in its early form with a stone pestle
over a simple hollow in the ground or on a mat, dates to the second millen-
nium B.C.E. in the Tilemsi Valley, Mali (Gaussen and Gaussen 1988). In Ouda-
lan (von Czerniewicz 2004, Magnavita et al. 2002) indigenous farming groups,
who were probably included in the proto-Songhay group made pottery using
V-shape mats since the first millennium B.C.E. and with a type 1 clay pestle at
least from the beginning of the first millennium C.E. e practice of this tech-
nique using a diagonal mat and a type 1 clay pestle, attested in Marandet (Ni-
ger) from the 6th century C.E. (Magnavita et al. 2007), is likely to be found in
the cultural context of proto-Songhay core. e diffusion of this technique up
the Niger River to Lake Debo by Songhay populations was accompanied by the
diagonal mat and the type 1 pestle and several changes in support (wooden and
raw clay anvil), with transcultural borrowing in the Inland Niger Delta by the
northern and southern Fulani, the northern Bamana and northern Somono. In
contrast, its horizontal spread across the Niger Bend was probably associated
with Gur populations, with changes in support (straight mat of baobab fibers,
Ceramic Traditions and Ethnicity in the Niger Bend 29
terracotta anvils or hollows in the ground) and the adoption of the stumpy
cylindrical clay pestle (type 2). e first archaeological evidences date back at
least from the 4th century C.E. in Gourma (Bedaux et al. 2003) and Seno plain
(Guindo 2006). Finally, Dogon artisans adopted this technique, with some us-
ing stone or type 2 pestles and straight mats (A tradition) and others using type
2 pestles and terracotta anvils or hollows in the ground (B tradition). e com-
plex processes of cultural and technical-transfer still remain relatively difficult
to explain in detail for this technique.
Common Roulette Impressions
In addition to production techniques, roulette techniques also can be traced
historically and correlated to various ethno-linguistic groups through space
and time. Ethnoarchaeological and historical studies by Gosselain (2000)
and Livingstone Smith (2007) demonstrate that roulettes first appeared in
the southwest Sahara and gradually spread from west to east along a savanna
corridor between the desert and forest zones, in contexts from the Late Neo-
lithic to the Late Iron Age. ey argue decorations with twine roulettes and
engraved roulettes were distributed across a large portion of Africa, indepen-
dent of variations in language, social organization, and economy. However, my
ethnoarchaeological studies show that some roulette decorations, studied in a
more restricted geographic zone, and type by type, can be good indicators of
ethnolinguistic entities. e observation of the presence or absence of roulette
decorations, considered as a whole, has little meaning, while comparison of the
frequencies of each different roulette type at each site and within each popula-
tion reveals interesting boundaries from historical and cultural points of view.
As an example, I recount below the historical depth and cultural evolution of
four common roulette impression types, often used to decorate pottery in the
Niger Bend.
Twisted cord roulette impressions (Figure 9)
Twisted cord roulette impressions do not show ethnolinguistic correlations, as
they are being used today in 11 traditions out of 13. e Somono and Dogon C
and D traditions use this type of impression to decorate entire pots. Fulani tradi-
tions use only the form of a single impressed row over the joint between the body
30 Anne Mayor
and the rim. Bwa, Songhay, Mossi and Dogon A and B traditions only rarely show
this decoration.
Archaeologically, twisted cord roulette impressions appear beginning with
the Neolithic at nearly all sites and in all regions of West Africa (southeast
Mauritania, Mema, Delta, Lakes Region, Middle Niger, Oudalan and Gourma).
ese ceramics are present more frequently in assemblages dating to the first
millennium C.E. and less so during the second millennium, in the Inland Niger
Delta (Bedaux et al. 2005; McIntosh S. 1995; Raimbault and Sanogo 1991), in
Tellem ritual caves on the Cliff (Bedaux and Lange 1983), and at Yatenga sites
(Andah 1978). I conclude that this decoration was part of the common Neo-
lithic substrate across the entire region, and Neolithic peoples transmitted it
to nearly all of the regional ethnolinguistic groups with a decreasing frequency
from northwest to southeast, and from the first to the second millennium. is
decoration is thus a poor cultural indicator in itself unless taken further, refin-
ing its description by taking into account the position of the impression on the
vessel or the diameter of the cord.
Folded strip roulette impressions (Figure 9)
Potters no longer make impressions of folded strip roulettes in the Niger Bend. Al-
though, archaeologists found this type in contexts from the first millennium B.C.E.
in Mema, Dia, in the Faita tradition sites, and in some caves of the Cliff derived from
late Neolithic traditions at Tichitt in southeast Mauritania (Bedaux and Lange 1983;
Bedaux et al. 2005; MacDonald and Schmidt 2004; McIntosh S. 1995; Togola 2008).
ese impressions next appear in very high percentages in the Inland Niger Delta
in phase I/II occupations (3rd century B.C.E.- 3rd century C.E.), where they deco-
rate more than 70 percent of the vessels. ey also are present on ceramics at some
sites in the Lakes Region (Raimbault and Sanogo 1991). is decoration also is found
in the Dogon Country during the first millennium C.E. (Mayor et al. 2005) and at
Koumbi Saleh from the 9th century on (Berthier 1997). Its percentage decreases in
the Inland Niger Delta during phases III and IV, becoming minor after the 11th cen-
tury. is decoration seems to be rare in the Middle Niger and Yatenga, and absent
in Oudalan. us, it is present in the northwest and central zones of the Niger Bend
slightly before and during the Ghana Empire before potters abandoned it during the
Mali Empire. It can be associated with a high degree of certainty to Mande artisans
linked to the proto-Soninke and then the Soninke, as well as the Bozo and Nono of
Ceramic Traditions and Ethnicity in the Niger Bend 31
the Delta. In the Lakes Region, its relatively important presence at the site of Mouys-
sam, dated between the 4th and 7th centuries, and its complete absence from the
neighboring site of Kawinza, dated between the 7th and 11th centuries, clearly in-
dicates a change in ceramic producers over the course of the Ghana Empire. e
presence of this decoration in high percentages at certain sites in the Dogon Country
dated to the first millennium suggests its local production by Mande artisan families
who had emigrated from the Delta.
Braided cord roulette impressions (Figure 9)
Braided cord roulette impressions, simple or double, are today very well-correlat-
ed ethnically, since Somono potters uniquely make them with some rare borrow-
ing by southern Bamana potters in an ethnically diverse village. is decoration,
which is in severe decline, first appears archaeologically in the Late Neolithic of
Mema during the second millennium B.C.E. and next is found in low percent-
ages in the first phase of occupation of the Inland Niger Delta (Arazi 2005; Be-
Figure 9. Cultural history in the Niger Bend for the principal common roulette impressions.
32 Anne Mayor
daux et al. 2001, 2005; MacDonald 1994; McIntosh 1995; Togola 2008). Between
the 4th and 14th centuries, the frequency of this decoration strongly increased
decorating 40 to 50 percent of the ceramics at southern Delta sites from the 10th
century on. In the Lakes Region, an important percentage of ceramics have this
decoration dating to the first phase of establishment between the 4th and 6th
centuries, but much less after (Raimbault and Sanogo 1991). In the Dogon Coun-
try, archaeologists find it in low percentages at many sites on the Plateau and the
Seno Plain after the first millennium (Bedaux and Lange 1993; Mayor 2005). is
would reflect either importation from the Delta at a time when the Plateau was
within the consumption zone of this tradition, or the local production of some
emigrants from the Delta.
e evolution of the proportion of braided cord roulette impressions is thus
the opposite of that for folded strip roulette impressions. ese two decorations
are, in addition, associated not only in the same assemblages, but also on the
same pots in certain places of the Lakes Region and the Delta. Interestingly, an
increase in the frequency of the braided cord roulette impressions in the In-
land Niger Delta coincides with the appearance of the technique of molding in
a concave form over a mobile plate, and also with the appearance of 5th cen-
tury funerary jars during full urbanization and at the onset of influence from the
Ghana Empire. I infer from this that the potters responsible for this decorative
technique had ancestors coming out of the local Neolithic substrate in Mema,
were linked with the Bozo and the Nono occupying the Mema, the Delta and
the Lakes Region during the Ghana and Mali Empires, and had descendants who
formed the principal crafts component of the Somono group beginning in the
16th century. ey probably belonged to the Mande ethnolinguistic group, like
those responsible for folded strip roulette impressions.
Braided strip roulette impressions (Figure 9)
Simple or alternate braided strip roulette impressions show a rather good eth-
nic correlation today, since the Bwa and the Bamana south of Bani in contact
with the Bwa use this decoration technique. It appears that Bamana potters in
southern Mali also use this decoration type (Frank 2007). In addition, the Mossi
state that they used this tool in the past, before its replacement by a metallic
spring dragged along the pottery surface. is decoration is found in archaeo-
logical contexts from the 3rd century B.C.E. in caves of the Cliff, but this must
Ceramic Traditions and Ethnicity in the Niger Bend 33
be confirmed given the lack of stratigraphy (Bedaux and Lange 1983). It is next
observed during the first millennium C.E. on the Bandiagara Plateau and in the
Seno plain (Mayor 2005), from the 11th century in Oudalan (von Czerniewicz
2004) and around Segou during the Bamana Empire (Haour et al. in press). It also
has an important presence in Yatenga at a date yet unknown (Marchal 1978). By
contrast, it is absent in the Delta and the Lakes Region.
In contrast with the two preceding types of roulette impressions, the geographic
distribution of braided strip roulette centers on the Dogon Country, southern Mali
and Yatenga. Historically potters from the Gur linguistic group, to which belong
the Mossi, the Bwa and the Kurumba, the latter considered to be descendants of
the Tellem (the group preceding the Dogon on the Bandiagara Cliff), probably used
this decoration type. e Bamana recently may have borrowed this decoration from
these neighboring groups. e late appearance of this decoration in Oudalan (von
Czerniewicz 2004) certainly indicates the installation of a few potter families from
another ethnic group, probably from Yatenga. While the few roulette impressed pot-
sherds observed along the Niger River (Bedaux et al. 1978) are probably evidence of
occasional acquisition during commercial exchanges. In light of the often poor iden-
tification of this decoration, a new examination of the archaeological material from
numerous sites of the Niger Bend would be useful for verification.
Syts Crs n Cr Idty
e elements forming a ceramic tradition do not all evolve at the same time, and
each element can tell a different story. e independent analysis of the different
shaping techniques and decorations shows, however, that the spatial distribution
of ceramic characteristics before the 15th century forms groups that are in large
part superimposed and that it is possible to historically associate them with dif-
ferent linguistic families (Figure 10). Over a common substrate originating from
the Neolithic, that combine the technique of molding over a convex form and the
twisted cord roulette impressions, I have identified three principal groups that
have differentiated since the first millennium B.C.E.:
1. A group located in the northeast of the Inland Niger Delta, linked to
Mande speakers, produced ceramics using the molding over a convex
form technique and folded strip impressions. Potters replaced this
forming technique between the 5th and 11th centuries C.E. in the
Inland Niger Delta with molding in a concave form, and the braided
34 Anne Mayor
cord roulette impressions progressively supplanted the folded strip
roulette impressions.
2. A group, located in the east of the Niger Bend linked to Songhay speak-
ers combining indigenous Gur elements, produced ceramics marked by
use of the technique of pounding in a concave form over a mat.
3. A group located in the Dogon Country and south of the Niger Bend,
linked to Gur speakers, produced ceramics using the technique of
hollowing a lump of clay and braided strip roulette impressions.
e ceramic traditions associated with current populations said to be au-
tochthonous, notably those associated with the Bozo/Somono in the Inland Ni-
ger Delta and the Bwa on the southern deltaic margin, can be associated with
groups 1 and 3 and have deep roots.
During the 13th to 15th centuries, significant population movements result-
ed in the current formation and distribution of numerous ethnic groups. After
this period, historical records and ethnohistory make it possible to correlate more
precisely changes in ceramic traditions with population dynamics (Figure 11):
Figure 10. Synthesis of relationships between ceramic traditions and linguistic groups in the
Niger Bend prior to the 15th century.
Ceramic Traditions and Ethnicity in the Niger Bend 35
Figure 11. Current distribution of shaping techniques and historical population movements
leading to their development in the Inland Niger Delta.
36 Anne Mayor
e progression of the Songhay along the Niger to Lake Debo in the 14th-
15th centuries introduced the technique of pounding in a concave form over a di-
agonal mat support in the northern part of the Inland Niger Delta. Over the cen-
turies, the Fulani, the northern Bamana and the northern Somono borrowed this
technique and transformed it with the adoption of wooden and raw clay anvils.
e colonization of Bamana cultivators in the 15th-17th centuries from the
High Valley of the Bani River to Mema and Lake Debo, allowed for the spread
of the technique of molding over a convex form. In contact with the Bwa, the
Bamana south of Bani borrowed the braided strip roulette.
After their supposed arrival from Mande between the 13th and 15th cen-
turies, the Dogon adopted the technique of pounding in a concave form over a
straight mat (tradition A) or in a massive terracotta anvil (tradition B), already
practiced locally by pre-Dogon populations. After the 16th century, two new
smith clans were active in the Dogon Country, leading to Dogon ceramic tradi-
tions C and D.
I have created this historical scenario of the evolution of ceramic traditions
during the period of pre-colonial empires based on the confrontation of ethnoar-
chaeological, historical, and archaeological studies. Some of the current ceramic
traditions have significant historical depth, traceable back to the Iron Age, while
others have become differentiated over the centuries, in parallel with the process
of individualization of ethnic groups or through transcultural borrowing.
Prior to the 15th century, it is possible to associate some ceramic features
with linguistic families, while after the 15th century, it becomes possible to es-
tablish correlations with currently known ethnic groups, even though their cul-
tural attributes, social structure, and way of life have certainly evolved since this
period. e parallel use of different shaping techniques and roulette decorations
in a given region from the Iron Age on provides evidence for the coexistence of
several different ethnic groups from this time.
e principal technical and stylistic traditions were formed just before or dur-
ing the Ghana Empire, and thus before the presumed arrival of craft castes in the
context of the Mali Empire. State formations had a clear influence on all of the
populations of the region by generating changes in their socio-economic structure
with the introduction of the caste system, but this evolution is not directly visible
in the ceramic traditions of the Niger Bend. It appears that potters transmitted
technical knowledge from generation to generation within artisan families show-
ing a tendency towards endogamy, which then became strictly endogamous castes.
Ceramic Traditions and Ethnicity in the Niger Bend 37
e innumerable political and military events for the most part did not affect
the transmission of specific technical knowledge within the different groups. e
influence of the empires was felt, in fact, primarily through the migrations that
they caused, for economic, political, or religious motives. e important contri-
bution of new populations and foreign artisans during the 13th-15th centuries
in different regions of the Niger Bend, in the context of the Mali and Songhay
Empires, is reflected by a diversification in ceramic traditions, while the flow of
the Dogon toward the Cliff under pressure from the Mossi kingdom of Yatenga is
seen in the spatial contraction of the Dogon A tradition (Gallay 1994).
Cnci
My ethnoarchaeological study of the correlations between ceramics and ethno-
linguistic identities represents a challenge to assumptions made by anthropol-
ogists and historians about the fluidity and permeability of ethnic groups and
ceramic traditions in West Africa. e results of this longitudinal research in
the same region underline the consistency of patterns emerging from the ethno-
archaeological data, despite the contemporary ethnic complexity of the region.
Based on solid field documentation, this study demonstrates that the explo-
ration of ethnicity or, more generally, identity, in African archaeology is possible
under certain conditions. As expected, there is no simple unequivocal correlation
between a material feature and an ethnic group, but a cluster of dynamic relation-
ships that evolve under the influence of different factors that must be decoded.
is approach to historical reconstruction was possible through comparison of
different kinds of data obtained in disciplines such as ethnoarchaeology, history,
and archaeology, the selection of relevant ceramic features with respect to identity
in the present, and consideration of a geographic zone sufficiently large as to be
able to compare settlement zones. Moreover, by combining historical elements,
expressing the dynamic dimension of societies, and psychomotor mechanisms ob-
served in the present, which reflect more stable elements of identity, I was able to
take into account the transmission of technical knowledge and the evolution of
traditions over a long time period. In this process, I had to shift from the notion of
ethnic group to that of linguistic family when going further back in time.
It is nevertheless important to keep in mind the vulnerability of interpreta-
tive constructs and the approximate aspect of such models. It is also important
to carry on with archaeological excavations to test the hypotheses, work which is
38 Anne Mayor
now in progress in the Dogon Country, a region of great complexity at the cross-
ing of several different cultural areas.
At a regional level, I hope that in the future, this study will serve to reevalu-
ate and refine the interpretation of various ceramic assemblages of the Niger
Bend, especially those excavated in settlement mounts of the Inland Niger Delta
and Oudalan. Interpretations of archaeological typological variability, based
mainly on chronology, should now take into account the question of identities
and population dynamics raised by ethnoarchaological work.
At a more general level, I hope to show that ethnoarchaeological studies
carried out with archaeological questions, and with the goal of identifying ex-
planatory mechanisms for the observed patterns, are useful for enriching ar-
chaeological interpretations, as already shown by the work of numerous other
ethnoarchaeologists (Arthur 2006; Longacre 1991; Roux 2007; Skibo et al. 1989).
is is worth stating, because after the debates of the past thirty years be-
tween processualists and post-processualists, general and direct historical analo-
gies, or ecological and historic-cultural determinisms, ethnoarchaeology is now
presented as eclectic (Cunningham 2003; David and Kramer 2001). Extending
the suggestions of including ethnoarchaeology in material culture studies (Hod-
der 1986), Gosselain even provocatively proposes that we entirely abandon the
ethnoarchaeological approach (Gosselain in press). However, a revival of inter-
est by ethnoarchaeologists for historical analogies is currently underway, with
the idea that historical continuity offers a greater likelihood that the same struc-
tures and explanatory processes affect both modern and archaeological contexts
(Cunningham 2003; Gosselain in press; MacEachern 1996).
Developed beyond these theoretical currents, studies by the Swiss team of
the ceramic traditions in Mali (Gallay et al. 1998), and particularly the final step
presented here (Mayor 2005, in preparation), combine technological analyses of
production sequences and identification of explanatory mechanisms, use both
general and direct historical analogies, and test the applicability of the results
obtained to different archaeological assemblages. I hope to have demonstrated
here some benefits of such a complete, but non-eclectic, approach.
Acknowledgments
e content of this paper was presented at the SAFA Conference in Calgary in
June 2006 and at the SAA symposium on ethnic ambiguity in African archaeol-
Ceramic Traditions and Ethnicity in the Niger Bend 39
ogy held in Austin in April 2007. I would like to thank François Richard and
Kevin MacDonald for inviting me to join this interesting symposium.
is research has been supported by the Swiss National Foundation for Sci-
entific Research (FNS), the Swiss-Liechtenstein Foundation for Archaeological
Research abroad (SLSA) and the Faculty of Science of the University of Geneva.
Our research in Mali benefits from the collaboration of the Cultural Mission of
Bandiagara, the Institute of Human Sciences in Bamako, the University of Ba-
mako, the Swiss Consulate in Bamako, as well as the inhabitants of numerous vil-
lages in the Inland Niger Delta and the Dogon Country. Finally, I thank Rebecca
Miller for translating this paper into English and the reviewers for their useful
comments.
References
Abitbol, Michel. 1979. Tombouctou et les Arma : de la conquête marocaine du Soudan
nigérien en 1591 à l’hégémonie de l’empire Peuhl du Macina en 1833. Paris: Maison-
neuve et Larose.
Amselle, Jean-Loup, and Elikia M’Bokolo, eds. 1985. Au coeur de l’ethnie: ethnies, tribal-
isme et état en Afrique. Paris: La Découverte.
Andah, Bassey W. 1978. Excavations at Rim, Upper Volta. West African Journal of Ar-
chaeology 8:75-138.
Arazi, Noemie. 1999. An archaeological survey in the Songhay heartland of Mali. Nyame
Akuma 52:25-43.
_____.2005. Tracing history in Dia, in the Inland Niger Delta of Mali: Archaeology, oral
traditions and written sources. PhD Diss., University College of London.
Arnold, Dean E. 1985. Ceramic theory and cultural process. Cambridge: Cambridge
University Press.
Arthur, John. 2006. Living with pottery: Ethnoarchaeology among the Gamo of southwest
Ethiopia. Salt Lake City: University of Utah Press.
Bâ, Amadou Hampaté, and Jacques Daget. 1984 (1955). L’empire peul du Macina
(1818–1853). Abidjan: Ecole des hautes études en sciences sociales.
Barth, Fredrik. 1969. Ethnic groups and boundaries: e social organization of culture
difference. London: Brown and Co.
Bedaux, Rogier. 1986. Pottery variation in present day Dogon compounds (Mali):
Preliminary results. In Variation, culture and evolution in African populations, eds.
Ronald Singer and John K. Lundy, 241–248. Johannesburg: Witwatersrand University
Press.
Bedaux, Rogier, and A. G. Lange. 1983. Tellem, reconnaissance archéologique d’une
culture de l’Ouest africain au Moyen-Age: la poterie. Journal des Africanistes
54(1–2):5–59.
40 Anne Mayor
Bedaux, Rogier, Jean Polet, Kléna Sanogo, and Annette Schmidt, eds. 2005. Recherches
archéologiques à Dia dans le Delta intérieur du Niger (Mali): Bilan des saisons de
fouilles 1998–2003. Leiden: CNWS.
Bedaux, Rogier, Trinette S. Constandse-Westermann, Louwrens Hacquebord, A. G.
Lange, and Johannes Diderik Van Der Waals. 1978. Recherches archéologiques dans
le Delta intérieur du Niger (Mali). Palaeohistoria 20:91–220.
Bedaux, Rogier, Kevin C. MacDonald, Alain Person, Jean Polet, Kléna Sanogo, Annette
M. Schmidt, and Samuel Sidibé. 2001. e Dia archaeological project: Rescuing cul-
tural heritage in the Inland Niger Delta (Mali). Antiquity 75(290):837–848.
Bedaux, Rogier, Alain Gallay, and Kevin C. MacDonald. 2003. L’archéologie des Dogon
de l’Est: Nokara, Sarnyéré and Douentza. In Regards sur les Dogon du Mali, eds.
Rogier Bedaux and Johannes Diderik Van der Waals, 40–47. Leyde: Rijksmuseum
voor Volkenkund.
Berthier, Sophie. 1997. Recherches archéologiques sur la capitale de l’empire de Ghana:
étude d’un secteur d’habitat à Koumbi Saleh, Mauritanie: Campagnes II-III-IV-V
(1975–1976)-(1980–1981). BAR International series, 680. Oxford: British Archaeo-
logical Reports.
Blench, Roger. 2005. A survey of Dogon languages in Mali: Overview. Ogmios newslet-
ter/Foundation for endangered languages 3(02):2–3.
Bouju, Jacky. 1995. Qu’est-ce que l’ethnie dogon? Cahiers des Sciences Humaines
31(2):329–363.
Bowser, Brenda. 2000. From pottery to politics: An ethnoarchaeological study of politi-
cal factionalism, ethnicity and domestic pottery style in the Ecuadorian Amazon.
Journal of Archaeological Method and eory 7(3):219–248.
Breunig, Peter and Katharina Neumann. 2002. Continuity or discontinuity? e 1st
millenium BC-crisis in West African prehistory. In Tides of the desert-Gezeiten der
Wüste: Contributions to archaeology and environmental history of Africa in honour
of Rudolph Kuper, ed. Tilman Lenssen-Erz, Ursula Tegtmeier, and Stefan Kröpelin,
491–505. Köln: Heinrich-Barth-Institut.
Burri, Elena. 2003. Cartographie des composantes stylistiques de la céramique dans le
Delta intérieur du Niger (Mali). Bulletin du Centre Genevois d’Anthropologie 6:69–93.
(de) Ceuninck, Grégoire. 1994. Forme, fonction, ethnie: Approche ethnoarchéologique
des céramiques du Delta intérieur du Niger (Mali). In Terre cuite et société: la
céramique, document technique, économique, culturel, 161–177. Juan-les-Pins: Edi-
tions APDCA.
_____. 2000. La circulation des poteries dans le delta intérieur du Niger (Mali). In Arts
du feu et productions artisanales, eds. Pierre Pétrequin, Philippe Fluzin, Jacques
iriot, and Paul Benoit, 221–236. Juan-les-Pins: Editions APDCA.
Cissoko, Sékéné Mody. 1975. Tombouctou et l’empire songhay : épanouissement du Sou-
dan Nigérien aux XVe – XVIe siècles. Dakar, Abidjan: Nouvelles éditions africaines.
_____. 1985. Les Songhay du XIIe au XVIe siècle. In Histoire générale de l’Afrique, 4:
l’Afrique du XIIe au XVIe siècle. ed. D.T. Niane, 213–236. Paris: UNESCO / Nouvelles
éditions africaines.
Ceramic Traditions and Ethnicity in the Niger Bend 41
Cunningham, Jerimy J. 2003. Transcending the “Obnoxious Spectator”: A case for
processual pluralism in ethnoarchaeology. Journal of Anthropological Archaeology
22(4):389–410.
_____. 2005. Household vessel exchange and consumption in the Inland Niger Delta of
Mali: An ethnoarchaeological study. PhD diss., McGill University, Montreal.
Costin, Cathy Lynne. 2000. e use of ethnoarchaeology for the archaeological study of
ceramic production. Journal of Archaeological Method and eory 7(4):377–403.
Curdy, Philippe. 1982. Tiébala (Mali): Un complexe céramique du 6e siècle après J.-C.
Archives Suisses d’Anthropologie Générale (Genève) 46(2):183–198.
(von) Czerniewicz, Maya. 2004. Studien zur Chronologie des Eisenzeit in der Sahel-
Zone von Burkina Faso/Westafrika. PhD diss., Johann-Wolfgang-Goethe University,
Frankfurt-am-Main.
David, Nicholas, and Judy Sterner, Kojo Gavua. 1988. Why pots are decorated. Current
Anthropology 29(3):365–389.
David, Nicholas, and Carol Kramer. 2001. Ethnoarchaeology in action. Cambridge: Cam-
bridge University Press.
Dieterlen, Germaine, and Diarra Sylla. 1992. L’empire du Ghana: Le Wagadou et les
traditions de Yéréré. Paris: Karthala/Arsan.
Dietler, Michael, and Ingrid Herbich. 1994. Ceramics and ethnic identity: Ethnoar-
chaeological observations on the distribution of pottery styles and the relationship
between the social contexts of production and consumption. In Terre Cuite et Société:
La céramique, document technique, économique, culturel, 459–472. Juan-les-Pins:
Editions APDCA.
Dramani-Issifou, Zakari. 1982. L’Afrique noire dans les relations internationales au XVIe
siècle: Analyse de la crise entre le Maroc et le Sonrhaï. Paris: Karthala.
_____. 1993. Les Songhay: Dimension historique. In Vallées du Niger, 151–161. Paris: Ed.
de la Réunion des musées nationaux.
Drost, Dietrich. 1967. Töpferei in Afrika: Technologie. Berlin: Akademie Verlag.
Fay, Claude. 1997. Les derniers seront les premiers: Peuplement et pouvoirs mandingues
et peuls au Maasina (Mali). In Peuls et Mandingues: Dialectique des constructions
identitaires, eds. Mirjam De Bruijn and Han Van Dijk, 165–191. Leiden: Afrika-
Studiecentrum.
Frank, Barbara. 1998. Mande potters and leatherworkers: Art and heritage in West Af-
rica. Washington and London: Smithsonian Institution Press.
_____.2007. Marks of identity: Potters of the Folona (Mali) and their “mothers.” African
Arts 40(1):30–41.
Gallais, Jean. 1980. Le delta intérieur du Niger et ses bordures: Etude de géographie régio-
nale. Lisse: Swets et Zeitlinger.
Gallay, Alain. 1970. La poterie en pays Sarakolé (Mali, Afrique occidentale): Étude de
technologie traditionnelle. Journal de la Société des Africanistes 40(1):7–84.
_____. 1992. A propos de la céramique actuelle du delta intérieur du Niger (Mali): Ap-
proche ethnoarchéologique et règles transculturelles. In Ethnoarchéologie: Justifica-
tion, problèmes, limites, 67–89. Juan-les-Pins: Editions APDCA.
42 Anne Mayor
_____. 1994. Sociétés englobées et traditions céramiques: Le cas du Pays dogon (Mali)
depuis le 13ème siècle. In Terre cuite et société: La céramique, document technique,
économique, culturel, 435–457. Juan-les-Pins: Editions APDCA.
Gallay, Alain, and Grégoire de Ceuninck. 1998. Les jarres de mariage decorationées du
Delta intérieur du Niger (Mali): Approche ethnoarchéologique d’un bien de prestige.
In Tradition und innovation: Prähistorische archäologie als historische wissenschaft:
festschrift für Christian Strahm, eds. Barbara Fritsch, Margot Maute, Irenäus Matus-
chik, Johannes Müller and Claus Wolf, 13–30. Rahden: M. Leidorf.
Gallay, Alain, and Eric Huysecom. 1989. Ethnoarchéologie africaine: Un programme
d’étude de la céramique récente du Delta Intérieur du Niger (Mali, Afrique de l’Ouest).
Genève: Département d’anthropologie et d’écologie.
Gallay, Alain, Eric Huysecom, Mathieu Honegger, and Anne Mayor. 1990. Hamdallahi,
capitale de l’Empire peul du Massina, Mali: Première fouille archéologique, études
historiques et ethnoarchéologiques. Stuttgart: F. Steiner.
Gallay, Alain, Eric Huysecom, and Anne Mayor. 1995. Archéologie, histoire et traditions
orales: Trois clés pour découvrir le passé Dogon. In Die Kunst der Dogon, ed. Lorenz
Homberger, 19–43. Zürich: Museum Rietberg.
Gallay, Alain, Eric Huysecom, and Anne Mayor. 1998. Peuples et céramiques du Delta
intérieur du Niger (Mali): Un bilan de cinq années de missions (1988–1993). Mainz:
P. von Zabern.
Gallay, Alain, Eric Huysecom, Anne Mayor, and Grégoire de Ceuninck. 1996. Hier et
aujourd’hui, des poteries et des femmes: Céramiques traditionnelles du Mali. Genève:
Département d’anthropologie et d’écologie de l’Université.
Gallay, Alain, and Claudine Sauvain-Dugerdil. 1981. Le Sarnyéré Dogon: Archéologie d’un
isolat, Mali. Paris: Editions ADPF.
Gallin, Annabelle. 2001–2002. Proposition d’une étude stylistique de la céramique im-
primée de Kobadi. Préhistoire, Anthropologie Méditerranéennes 10(11):117–133.
Gardi, Bernhard. 1985. Ein Markt wie Mopti: Handwerkkerkasten und traditionnelle
Technologie in Mali. Basel: Wepf.
Gaussen, Jean, and Michel Gaussen.1988. Le Tilemsi préhistorique et ses abords: Sahara
et Sahel malien. Bordeaux: Editions du CNRS.
Gelbert, Agnès. 2001. Ethnoarchaeological study of ceramic borrowings: A new meth-
odological approach applied in the middle and upper valleys of the Senegal River. In
Ethno-archaeology and its transfers, eds. Sylvie Beyries and Pierre Pétrequin, 81–94.
BAR International series, 983. Oxford: British Archaeological Reports.
_____. 2003. Traditions céramiques et emprunts techniques: Etude ethnoarchéologique
dans les haute et moyenne vallées du fleuve Sénégal. Paris: Editions de la Maison des
sciences de l’homme.
Gosselain, Olivier P. 2000. Materializing identities: An African perspective. Journal of
Archaeological Method and eory 7(3):187–217.
_____. 2001. Globalizing local pottery studies. In Ethno-archaeology and its transfers,
eds. Sylvie Beyries and Pierre Pétrequin, 95–111. BAR International series, 983.
Oxford: British Archaeological Reports.
Ceramic Traditions and Ethnicity in the Niger Bend 43
_____. 2002. Poteries du Cameroun méridional: Styles, techniques et rapports à l’identité.
Paris: Editions du CNRS.
_____.In press. A quoi bon l’ethnoarchéologie? In L’archéologie comme discipline?, ed.
Philippe Boissinot. Paris: Editions du Seuil.
Guindo, Nema. 2006. La Plaine du Séno: Plus de deux millénaires d’histoire du peuple-
ment. Etudes Maliennes 65:173–181.
Haour, Anne, Katie Manning, Noemie Arazi, Olivier P. Gosselain, Sokhna Gueye,
Daouda Keita, Alexander Livingstone Smith, Kevin C. MacDonald, Anne Mayor, Su-
san K. McIntosh, and Robert Vernet eds. In press. African impressed pottery: Roulette
decoration past and present. London: Oxbow.
Hodder, Ian. 1986. Reading the past: Current approaches to interpretation in archaeol-
ogy. Cambridge: Cambridge University Press.
Holder, Gilles. 2001. Poussière, ô poussière!: La cité-état sama du pays dogon (Mali).
Nanterre : Société d’ethnologie.
Huysecom, Eric. 1992. Vers une ethnoarchéologie appliquée: Exemples Africains. In
Ethnoarchéologie: Justification, problèmes, limites, 91–102. Juan-les-Pins: Editions
APDCA.
_____. 1994. Identification technique des céramiques Africaines. In Terre cuite et société:
La céramique, document technique, économique, culturel, 31–44. Juan-les-Pins: Edi-Juan-les-Pins: Edi-
tions APDCA.
_____. 1996. Iron Age terracotta pestles in the Sahel area: An ethnoarchaeological
approach. In Interregional contacts in the later prehistory of Northeastern Africa,
eds. Lech Krzyzaniak, Karla Kroeper and Michael Kobusiewicz, 419–458. Poznan:
Archaeological Museum.
_____. 2002. Palaeoenvironment and human population in West Africa: An international
research project in Mali. Antiquity 76(292):335–336.
Huysecom, Eric and Anne Mayor. 1993. Les traditions céramiques du delta intérieur du
Niger: Present et passé. In Vallées du Niger, 297–313. Paris: Editions de la réunion
des musées nationaux.
Insoll, Timothy. 1996. Islam, archaeology and history: Gao region (Mali) ca. AD 900–
1250. BAR International series, S647. Oxford: British archaeological reports.
_____. 2000. Urbanism, archaeology and trade: Further observations on the Gao region
(Mali): e 1996 field season results. BAR International series, S829. Oxford: British
Archaeological Reports.
Izard, Michel. 1985. Le Yatenga précolonial: Un ancien royaume du Burkina. Paris:
Karthala.
Jenkins, Richard P. 1997. Rethinking ethnicity: Arguments and explorations. London:
Sage Publications Ltd.
Jones, Sian. 1997. e archaeology of ethnicity: Constructing identities in the past and
present. London: Routledge.
Kassibo, Bréhima. 1994. Histoire du peuplement humain. In La pêche dans le delta
central du Niger: Approche pluridisciplinaire d’un système de production halieutique,
ed. Jacques Quensière, 81–97. Paris: ORSTOM, Karthala.
44 Anne Mayor
Konare Ba, Adam. 1987. L’épopée de Segu : Da Monzon: un pouvoir guerrier. Lausanne:
P.M. Favre.
Kramer, Carol. 1985. Ceramic ethnoarchaeology. Annual Review of Anthropology
14:77–102.
Lange, Dierk. 1994. From Mande to Songhay: Towards a political and ethnic history of
Medieval Gao. Journal of African History 35(2):275–301.
Laviolette, Adria. 2000. Ethno-archaeology in Jenné, Mali: Craft and status among
smiths, potters and masons. Oxford: Cambridge Monographs in African Archaeology
49, British Archaeological Reports International Series.
Levtzion, Nehemia and John F.P. Hopkins, eds. 1981. Corpus of early Arabic sources for
West African history. Cambridge: Cambridge Historiae Africanae.
Livingstone Smith, Alexander. 2001. Chaîne opératoire de la poterie: Références ethno-
graphiques, analyses et reconstitution. PhD diss., Université libre de Bruxelles.
_____. 2007. Histoire du décor à la roulette en Afrique subsaharienne. Journal of African
Archaeology 5(2):189–216.
Longacre, William A., ed. 1991. Ceramic ethnoarchaeology. Tucson: University of Ari-
zona Press.
Ly-Tall, Madina 1977. Contribution à l’histoire de l’Empire du Mali (XIIIe–XVIe siècles):
Limites, principales provinces, institutions politiques. Dakar, Abidjan: Nouvelles édi-
tions africaines.
_____. 1985. Le déclin de l’Empire du Mali. In Histoire générale de l’Afrique IV: L’Afrique
du XIIe au XVIe siècle, ed. Djibril Tamsir Niane, 197–212. Paris: Nouvelles éditions
Africaines.
_____. 1996. Le Macina et l’Empire torodbe (tukuloor) jusqu’en 1878. In Histoire générale
de l’Afrique VI: L’Afrique au XIXe siècle jusque vers les années 1880, ed. Jacob F. Ade
Ajayi, 647–682. Paris: UNESCO.
MacDonald, Kevin C. 1994. Socio-economic diversity and the origin of cultural
complexity along the Middle Niger (2000 BC to AD 300). PhD diss., University of
Cambridge.
_____. 1996. Tichitt-Walata and the Middle-Niger: Evidence for cultural contact in the
second millenium BC. In Aspects of African archaeology, ed. Gilbert Pwiti and Robert
Soper, 429–440. Harare: University of Zimbabwe Publ.
MacDonald, Kevin C., and Annette M. Schmidt. 2004. e Faïta Facies and the “Ob-
scure Millennium” of the Middle Niger. In Congrès de l’Association panafricaine de
préhistoire et disciplines assimilées, eds. Kléna Sanogo and Tereba Togola, 222–229.
Bamako: Institut des sciences humaines.
MacDonald, Kevin C., and Robert Vernet, Marcos Martinon-Tores, and Dorian Q.
Fuller. 2009. Dhar Néma: From early agriculture to metallurgy in southeastern Mau-
ritania. Azania: Archaeological Research in Africa 44(1):3–48.
MacEachern, Scott. 1996. Foreign countries: e development of ethnoarchaeology in
Sub-Saharan Africa. Journal of World Prehistory 10(3):243–304.
Ceramic Traditions and Ethnicity in the Niger Bend 45
Magnavita, Sonja, Abdoulaye Maga, Carlos Magnavita, and O. Idé. 2007. New studies on
Marandet (central Niger) and its trade connections: An interim report. Zeitschrift für
Archäologie Aussereuropäischer Kulturen 2:147–165.
Magnavita, Sonja, Maya Hallier, Christoph Peltzer, Stephanie Kahlheber, and Veerle
Linseele. 2002. Nobles, guerriers, paysans: Une nécropole de l’Age de Fer et son
emplacement dans l’Oudalan pré- et protohistorique. Beiträge zur Allgemeinen und
Vergleichenden Archäologie 22:21–64
Marchal, Jean-Yves. 1978. Vestiges d’occupation ancienne au Yatenga (Haute-Volta):
Une reconnaissance du pays Kibga. Cahiers de l’Orstom, série sciences humaines
15(4):449–484.
Mayor, Anne. 1994. Durée de vie des céramiques africaines: Facteurs responsables et
implications archéologiques. In Terre cuite et société: La céramique, document tech-
nique, économique, culturel, 179–198. Juan-les-Pins: Editions APDCA.
_____. 1996. Hamdallahi, capital of the Fulani Empire of Macina, Mali: A multidisci-
plinary approach. In Aspects of African archaeology, eds. Gilbert Pwiti and Robert
Soper, 671–680. Harare: University of Zimbabwe.
_____. 1997. Les rapports entre la Diina peule du Maasina et les populations du Delta
intérieur du Niger, vus au travers des traditions historiques et des fouilles archéo-
logiques. In Peuls et Mandingues: Dialectique des constructions identitaires, eds.
Mirjam De Bruijn and Han Van Dijk, 33–60. Paris: Kathala, Leiden: Afrika-Studie-
centrum.
_____. 1999. Quand le feu, l’air et la terre s’unissent pour créer . . . : L’exemple de la cuis-
son des poteries au Mali. In Le feu: Feu dévorant, feu domestique, feu sacré, ed. Eric
Golay, 95–106. Genève: Musée d’ethnographie.
_____. 2003. L’abri-sous-roche de Dangandouloun en pays Dogon (Mali): Quand et
comment? In Constellasion: Hommage à Alain Gallay, eds. Marie Besse, Laurence-
Isaline Stahl Gretsch and Philippe Curdy, 353–374. Lausanne: Cahiers d’archéologie
romande.
_____. 2005. Traditions céramiques et histoire du peuplement dans la Boucle du Niger
(Mali) au temps des empires précoloniaux. PhD diss., University of Geneva, Switzer-PhD diss., University of Geneva, Switzer-
land.
_____. 2006. Integrating ethnoarchaeological, historical and archaeological data:
Towards a history of pottery traditions in the Niger Bend. On-line proceedings of
the 18th Biennial Meeting of the Society of Africanist Archaeologists, University of
Calgary, http://cohesion.rice.edu/centersandinst/safa/news.cfm?doc_id=9495.
_____. In preparation. Histoire des traditions céramiques et du peuplement dans la
Boucle du Niger au temps des empires précoloniaux. Frankfurt-am-Main: Journal of
African Archaeology, monographs series.
Mayor, Anne, Eric Huysecom, Alain Gallay, Michel Rasse, and Aziz Ballouche. 2005.
Population dynamics and paleoclimate over the past 3000 years in the Dogon Coun-
try, Mali. Journal of Anthropological Archaeology 24(1):25–61.
McIntosh, Rodrick J. 1998. e peoples of the Middle Niger: e island of gold. Malden
Mass, Oxford: Blackwell.
46 Anne Mayor
McIntosh, Susan K., ed. 1995. Excavations at Jenné-Jeno, Hambarketolo, and Kaniana
(Inland Niger Delta, Mali): e 1981 season. Los Angeles: University of California
Press.
_____. 1999. Beyond chiefdoms: Pathways to complexity in Africa. Cambridge: Cam-
bridge University Press.
McIntosh, Susan K., and Rodrick J. McIntosh. 1980. Prehistoric investigations at Jenne,
Mali. BAR International series, 89(1–2). Oxford: British Archaeological Reports.
Ozainne, Sylvain, Eric Huysecom, Anne Mayor, Caroline Robion-Brunner, and Sylvain
Soriano. 2009. Une chronologie pour le peuplement et le climat du pays dogon : la
sequence culturelle et environnementale du gisement d’Ounjougou (Mali). Antropo
18:37–46. www.didac.ehu.es/antropo.
Pelzer, Christoph. 2003. La “nomadisation” du Sahel burkinabè: Le cas des provinces de
l’Oudalan, du Séno et du Soum (1500–1900). In Histoire du peuplement et relations
interethniques au Burkina Faso, eds. Richard Kuba, Carola Lenz and Claude Nuru-Richard Kuba, Carola Lenz and Claude Nuru-
kyor Somda, 203–221. Paris: Karthala.
Quensière, Jacques, ed. 1994. La pêche dans le Delta central du Niger: Approche pluridis-
ciplinaire d’un système de production halieutique. Paris: ORSTOM, Karthala.
Raimbault, Michel. 1980. La poterie traditionnelle au service de l’archéologie: Les ateliers
de Kalabougou (cercle de Ségou, Mali). Bulletin de l’Institut Fondamental d’Afrique
Noire, Série Sciences Humaines 42(3):444–474.
Raimbault, Michel, and Kléna Sanogo. 1991. Recherches archéologiques au Mali: Les sites
protohistoriques de la zone lacustre. Paris: ACCT, Karthala.
Rice, Prudence M. 1987. Pottery analysis: A sourcebook. Chicago: University of Chicago
Press.
Roberts, Richard. 1981. Fishing for the state: e Somono and the political economy
of the Middle Valley of the Niger. In Modes of production in Africa: e precolonial
area, ed. Donald Crumney and Charles Stewart, 174–203. London : Sage Publica-
tions.
_____. 2005. Litigants and households. African dispute and colonial courts in the French
Soudan, 1895–1912. Social History of Africa. Portsmouth, NH: Heinemann.
Robinson, David. 1988. La guerre sainte d’Al-Hajj Umar: Le Soudan occidental au milieu
du XIXe siècle. Paris: Karthala.
Robion-Brunner, Caroline. 2008. Vers une histoire de la production du fer sur le plateau
de Bandiagara (pays dogon, Mali) durant les empires précoloniaux: Peuplement des
forgerons et traditions sidérurgiques. PhD diss., University of Geneva, Switzerland.
Roux, Valentine. 2007. Ethnoarchaeology: A non historical science of reference nec-
essary for interpreting the past. Journal of Archaeological Method and eory
14(2):153–178.
Roux, Valentine, and Daniela Corbetta. 1990. Le tour du potier: Spécialisation artisanale
et compétences techniques. Paris: CNRS.
Ruhlen, Merritt. 1987. A guide to the world’s languages, 1: Classification. Caufield Est: E.
Arnold.
Ceramic Traditions and Ethnicity in the Niger Bend 47
Sanankoua, Bintou. 1990. Un empire peul au XIXe siècle: La Diina du Maasina. Paris:
Karthala, ACCT.
Shennan, Stephen J., ed. 1989. Archaeological approaches to cultural identity. London:
Unwin Hyman.
Skibo, James M., Michael B. Schiffer, and Nancy Kowalski. 1989. Ceramic style analysis
in archaeology and ethnoarchaeology: Bridging the analytical gap. Journal of Anthro-
pological Archaeology 8(4):388–409.
Stark, Miriam, ed. 1998. e archaeology of social boundaries. Washington: Smithsonian
Institution Press.
Sterner, Judy, and Nicholas David. 2003. Action on matter: e history of the uniquely
African tamper and concave anvil pot-forming technique. Journal of African Archae-
ology 1(1):3–38.
Tamari, Tal. 1991. e development of caste systems in West Africa. Journal of African
History 32(2):221–250.
Togola, Tereba. 1996. Iron Age occupation in the Méma region, Mali. African Archaeo-
logical Review 13(2):91–110.
_____. 2008. Archaeological investigations of Iron Age sites in the Mema Region, Mali
(West Africa). BAR International series, 1736. Oxford: British Archaeological Re-
ports.
Van Der Leeuw, Sander E., and Alison C. Pritschard, eds. 1984. e many dimensions of
pottery: Ceramics in anthropology and archaeology. Amsterdam: University Press.
Vernet, Robert. 1996. Le Sud-Ouest du Niger de la préhistoire au début de l’histoire.
Paris: Sépia.
