INTERACTION STUDY BETWEEN HUNTER-GATHERERS, PASTORALISTS AND FARMING COMMUNITIES IN KHWEBE HILLS,...
Transcript of INTERACTION STUDY BETWEEN HUNTER-GATHERERS, PASTORALISTS AND FARMING COMMUNITIES IN KHWEBE HILLS,...
INTERACTION STUDY BETWEEN HUNTER-GATHERERS,
PASTORALISTS AND FARMING COMMUNITIES IN KHWEBE HILLS,
NORTH WESTERN BOTSWANA
BY
RAMMUTLOA KEFILWE
SUBMITTED IN FULFILMENT OF THE REQUIREMENTS FOR THE
DEGREE
BACHELOR OF ARTS HONOURS
IN THE
FACULTY OF HUMANITIES
UNIVERSITY OF PRETORIA
SUPERVISOR
DR CERI ASHLEY
15 NOVEMBER 2013
This dissertation is dedicated to my grandmother Kefilwe Sylvia Rammutloa
“Ngwana Matsebe Mokoni wa Dintshi Dikgolo”.
DECLARATION
I Kefilwe Rammutloa declare that this dissertation titled “Interaction study between
hunter-gatherers, pastoralists and farming communities in Khwebe Hills, north
western Botswana” is my own work, independently work. I declare that all the
sources used have been cited correctly. This is being submitted for the Bachelor of
Arts Honours Degree in the University of Pretoria. This dissertation has never been
submitted before for any degree or examination at any university.
Rammutloa Kefilwe
Candidate signature
Day of 2013
I
ACKNOWLEDGEMENTS
First I would like to thank Dr Alexander Antonites and Dr Ceri Ashley for introducing
me to this project, which has been a learning curve for me. I am utterly grateful to Dr
Ceri. Z Ashley for granting me an opportunity to work with her and be part of this on-
going research. I am thankful for her assistance, patience, advice and support
throughout the duration of this research project, as well as in my studies. I am also
grateful to Mr Benjamin Saccaggi, Mr Gerhard Jordan and Mrs Jacqueline Jordan for
all the advice towards the research project. I would like to thank the Botswana
National Museum in Gaborone for granting me access to the Khwebe Hills ceramic
collection, as well as the museum staff.
I would like to thank Professor Innocent Pikirayi for his financial assistant. Thank you
to the National Research Foundation (NRF), as well as South African National Park
(SANParks) for financial assistance during the course of the year 2013, particularly
towards the completion of this project and my honours degree.
I am grateful to all those individuals who gave advice towards this project.
II
Glossary
Lekgoba (-Ma) - Slave
XRD - X-ray Diffraction
SNR – Short Neck Recurved
PSNR – Pronounced Short Neck Recurved
ESNR – Everted Short Neck Recurved
SEPN – Short Everted Pronounced Neck
HB – Hemispherical Bowl
EPN- Everted Pronounced Neck
SPN- Straight Pronounced Neck
ERN – Everted Recurved Neck
SRN – Straight Neck Recurved
CN – Collared Neck
III
TABLE OF CONTENTS
CHAPTER 1
INTRODUCTION TO RESEARCH PROJECT 1
THE SITE 1
AIMS 4
SIGNIFICANCE 4
RESEARCH QUESTIONS 4
THESIS STRUCTURE 5
CHAPTER 2
LITERATURE REVIEW 6
INTRODUCTION 7
THE SITE 7
HISTORICAL BACKGROUND 8
SOCIAL INTERACTION 10
BROADER ARCHAEOLOGICAL LITERATURE 11
THE KALAHARI DEBATE 16
CONCLUSION 17
IV
CHAPTER 3
THEORY CHAPTER
INTRODUCTION 19
TRADITIONAL DEFINITIONS 19
COLONIAL AND POST-COLONIAL THEORY 20
APPLICATION OF INTERACTION STUDIES 22
MATERIAL CULTURE AS BOUNDARY 22
SOUTHERN AFRICAN ARCHAEOLOGY 23
CONCLUSION 25
CHAPTER 4
METHODOLOGY CHAPTER
INTRODUCTION 26
PREVIOUS RESEARCH 26
CURRENT INVESTIGATIONS 27
TYPOLOGY 28
ARCHAEOLOGICAL EVIDENCE 34
CONCLUSION 42
V
CHAPTER 5
RESULTS CHAPTER
INTRODUCTION 43
ANALYSIS 43
DATA PRESENTATION 46
PATTERN 59
CONCLUSION 61
CHAPTER 6
INTERPRETATIONS
INTRODUCTION 62
ANALYSIS 62
ARCHAEOLOGICAL EVIDENCE 64
DISCUSSION 66
CONCLUSION 67
CHAPTER 7
SUMMARY 69
BIBLIOGRAPHY 71
APPENDIX
VI
1
CHAPTER 1: INTRODUCTION
This chapter serves as an introduction to the project, which aims at studying the
interaction between hunter-gatherers, pastoralists and farming communities through
ceramics from the Khwebe Hills, of north western region in Botswana. The chapter
entails a description of the site area, a brief narration of the oral history, as well as
background on the previous research done at the site. In addition, this section will
discuss the research questions, as well as give a brief description of the chapters and
the various topics that the research project will focus on.
1.2 SITE
1.2.1 Geography
The Khwebe Hills are situated in the area that was previously known as Ngamiland,
in the North-western region of Botswana (Tlou 1980: 1). This is a land locked
country, surrounded by four countries; which includes Zambia in the northern area,
and South Africa in the southern region. While the eastern area is locked off by
Zimbabwe and the western region is enclosed off by Namibia (see figure 1.1).
According to Dziewiecka (2008: 11) the Khwebe Hills are windy, and the oral history
narrates this by elucidating the name of the hills according to the Khwe mythology.
The hills are named after a wind, which is a supernatural being who was the first
person at hills (Dziewiecka 2008: 11).
“According to San mythology, 'Khwe is the wind, which is a supernatural being. The loud sound of the
wind prophesies evil; it communicates with the creatures of prey where to find the people. These loud
howling sounds are an indication that these beings can approach the huts unheard (Dziewiecka 2008:
11). The wind is a representation of a man who can take a bird form” (Dziewiecka 2008:11).
The Khwebe Hills have their own micro-climate that enables different trees and plants
to grow. This includes Baobabs (adansonia digitata), Mórula (sclerocarya birred) and
2
motsentsela (berchemia discolor) trees. The Shrubs of all kinds of the grewia species
such as mogwana, mokgompatha moretlwa (wild berries), moretologa and
motsontsonyana (Mulberry) - grow there, as do herbs of medical value and wheat type
grass (Dziewiecka 2008: 11-12).
1.2.2 Social dynamics
The hills have hosted people of different ethnicity and diverse economies in last
couple of centuries. The oral history of the site will be discussed in more details in the
literature chapter; however, it is indicates that a type of Bantu-speaking agro-
pastoralists migrated to the Khwebe Hills (Tlou 1980: 38). These new immigrants
were referred to as BaTawana, they dominated and colonised the autochthonous
people, which included the Basarwa and the Bakgalagadi, the latter arrived 800 years
Figure 1.1: A map illustrating Ngamiland, the Khwebe hills where the site is located
and Lake Ngami
3
ago (Tlou 1980: 8). The oral history also narrates that this domination marginalized
the indigenous inhabitants; as such they were left in subjugation.
1.2.3 Earlier research
Previous research at the Khwebe Hills was done in 2008, 2009 and 2010. The 2008
excavation was at the missionary site KWH1, where the British missionary had their
station set up. The excavations in 2009 and 2010 was at site KWH4, a stone walled
site that dates to the 19th
century. This site is the first Tswana town in the region and
associated to the BaTawana who arrived in Ngamiland in the 19th
century from the
south eastern region of Botswana.
There seems to be a huge gap in the archaeological literature addressing questions on
interaction during this period, particularly in southern Africa. Earlier studies by Hall
(2002) and Thorp (2000) have indicated that the Bantu-speaking agriculturalists have
been interacting with the mobile hunter-gatherer bands and pastoral societies.
However, questions have been raised on whether this interaction led to colonisation
and marginalization of the autochthonous people, or if the interaction was more of a
symbiotic relationship. These inquiries have contributed to the Kalahari debate, which
will be discussed in broader details in the next chapter.
The debate questions the role of hunter-gatherers, how they were incorporated and
how interacted with food producing societies (Barnard 2006). Previous studies
(Bollong 1993) and current studies (Diskin & Ashley: in prep) have argued that
ceramics can be used to address questions on interaction. However, there is a gap
within the literature particularly during the 19th
century period. This gap has surfaced
questions on technology, style and typology.
4
1.3 AIMS
The project will address questions on interaction as well as culture contact, using the
19th
century ceramic. The project also aims at creating knowledge about on interaction
in southern Africa between people of distinct ethnicities, as well as different
economies. The archaeological evidence from this site will be used and compared to
the oral history; this is to determine whether or not the archaeological evidence
supports the oral history of the site. According to Huffman (1994: 5-6) there is a gap
in the literature for pastoralists ceramics particularly the historic pottery. As such, this
research project will present a new typology from this 19th
century site, thus trying to
bridge this gap in the archaeological sequence.
1.4 SIGNIFICANCE
Interaction studies are growing in southern Africa, as such this research project will
add on existing knowledge about interaction and culture contact. The typology to be
presented in chapter 4; will also add on existing knowledge on the pastoral ceramics,
hunter-gatherers. As well as explore the possible variations within the 19th
century
Tswana ceramic typology.
1.5 RESEARCH QUESTIONS
The proposed research questions are based on previous research by Diskin and Ashley
(in prep), Sadr (1999) and Sadr (2008). A new ceramic typology will be created for
this site, this typology will be put into the fabric that were categorised by Diskin &
Ashley (in prep). This will serve the purpose of answering the following questions:
Can ceramics illustrate the interaction between hunter-gatherers, herders and
incoming agriculturalists?
5
Was the initial interaction between these people fluid or were the
autochthonous people colonisation?
Are there any vessel variations between these three fabrics?
What variations of vessels forms exist within each fabric?
Does the typology of KWH4 link up with any existing typologies?
1.6 THESIS STRUCTURE:
Chapter 2: Literature - This chapter will give a detailed review of southern African
archaeology, particularly the archaeology of hunter-gatherers, pastoralists and
agriculturalists. This will be used to contextualise the oral history of the Khwebe hills
within the broader archaeology in the landscape and beyond.
Chapter 3: Theory chapter - This chapter will give a broader theoretical review on
interaction studies within archaeology. In addition, two paradigms (colonial and post-
colonial paradigms) will be used to understand the concepts of frontiers and
boundaries.
Chapter 4: Methodology chapter - This chapter will give a detailed description and
discussion of the different methods used to create the typology of KWH4. The chapter
will also present the ceramic typology of site KWH4 and use existing literature and
typologies to contextualise KWH4 ceramic typology
Chapter 5: Data and results – This chapter will present the data and results using
graph, illustrations, and tables.
Chapter 6: Interpretation chapter - This chapter will use the data and results obtained
from the analysis; which were presented in the previous chapter (chapter 5) to
6
understand how this site fits into the broader archaeology. As well as answering the
research questions presented above.
Chapter 7: Conclusion – This chapter will give a summary of the project and give
future plans where this project.
7
CHAPTER 2: LITERATURE CHAPTER
2.1 INTRODUCTION
This chapter will introduce the site of KWH4 in the Khwebe Hills in details. It will
explore the oral history, with the emphasis on the interaction between hunter-
gatherers, pastoralists and farming communities that occurred in the 19th
century.
Furthermore, the broader literature will be consulted with the purpose of
contextualising the archaeology of Khwebe Hills within southern African
archaeology. As part of the literature chapter, the Kalahari debate will be will be
discussed; this will serve the purpose of elucidating the importance of this project
(site) within the broader archaeological theories on interaction.
2.2 SITE
2.2.1 Khwebe hills
The site of KWH4 is situated in the Khwebe Hills of Ngamiland (Tlou 1980: 1). The
area of Ngamiland is 570,000 km2 in size and stretches from the Lake Ngami shores
to the Kuke territory (Dziewiecka 2008: 11).
The Hills lie approximately 50 km on the
south-west periphery of the modern town of
Maun, and in the frontier area between the
Kalahari Desert and the Okavango Delta
region (Ashley 2009:37), (see figure 2.1). The
site of KWH4 stretches approximately 1
kilometre across the hillside (Ashley 2009:
37). Figure 2.1: A map illustrating where the site
located (Ashley 2009: 36).
8
2.3 HISTORICAL BACKGROUND
2.3.1 Basarwa and Bakgalagadi
According to Dziewiecka (2008: 12), the original community of the Khwebe Hills had
no borders, rulers, or chieftaincies during the 17th century and 18th century. At that
time the hills were inhabited by a hunter-gatherer community who called themselves
Khwe, which means people in Nharo (Dziewiecka 2008: 12). According to
Dziewiecka (2008:12) this is where the name of the hills derives from. The
archaeological evidence indicates that the hunter-gatherer community have occupied
this area for approximately 10000 years (Tlou 1980: 12). This community can be
further divided into two groups; those that lived by the sand belts and exploited the
riverine mammals; they were referred to as “Banoka”, the river people (Tlou 1980:8).
The other hunter-gatherer group were known as the Basarwa; they inhabited the arid
areas and maintained subsistence through hunting wild game and gathering wild
berries (Tlou 1980:8). The hunter-gatherer bands co-existed with another group (Tlou
1980: 12). This group arrived 800 years ago, they were known as Bakgalagadi, a
pastoral community whose origin is unclear. The Oral history narrates that the
ambiguous origin is a result of internal rivals that cause the group to split into
multiple, and culturally distinct communities, which are spread across Botswana
today (Chambanne & Monaka 2008: 133-135).
Kuper (1970: 45) also discusses this unclear origin, as his research indicates that the
Ngaloga, Saga and the Kgwatlheng are classified and identified as Bakgalagadi, who
all have different origins. Earlier linguistic investigations by Okihiro (cited in Mautle
1986: 20) suggested that the Bakgalagadi were a fragment of Barolong, a Tswana
speaking group. However, this has been discredited by current on-going linguistic
9
research across the Botswana country (Chambanne & Monaka 2008: 133-135; Tlou &
Campbell 1997:33). Furthermore, this new evidence suggests that Sekgalagadi, the
language, is a separate member of Sotho-Tswana language group (Westphal 1944
cited in Mautle 1986: 20).
Nevertheless, the Bakgalagadi origins cannot be traced clearly; however, they resided
close to permanent waterholes in the Khwebe Hills (Tlou 1980: 8). These waterholes
gave them the opportunity to herd small stocks of goats and sheep, and at times cattle
(Tlou 1980: 8). They also enabled them to cultivate land at a small-scale growing
beans, melons and sorghum (Mautle 1986: 24).
2.3.2 Agro-pastoralists
By the end of the 18th
century and the beginning of the 19th
century there seems to be
an inflow of Bantu-speaking agro-pastoralists in Ngamiland. The first groups to arrive
are the Hambukushu and Bayei, both agro-pastoralists; they settled and fished
alongside the Banoka (Tlou 1980: 12). According to Potten (1976: 63) the Bayei and
Hambukushu people moved gradually towards the Okavango Delta from south-west
Zambia, through the Caprivi Strip, Kwando and Linyanti Rivers. With their arrival,
the Bayei introduced more advanced fishing and hunting methods, as well as the
practice of diverse agriculture to the region. While the Hambukushu were equipped
with iron ornaments, as such they introduced the production and usage of the iron
implements to the region (Potten 1976: 63).
During this period another agro-pastoralist group that this research project is
particularly interested in moved to the hills. They are BaTawana, a Setswana speaking
group that arrived in Ngamiland in 1795 (Wilmsen 2003: 333), and settled at the
Khwebe Hills 1805 (Tlou 1980: 38-42). The origin of their name is clear, as most
10
Tswana-speaking groups are named after their leaders. The BaTawana group name
derives from their leader Kgosi Tawana (small lion).
According to the oral tradition they are an offshoot of the BaNgwato from Shoshong
in south eastern region of Botswana (Potten 1976: 63). They migrated to the hills after
the BaNgwato state was torn apart by a succession dispute (Tlou 1980:38). This
disagreement resulted from Kgosi Mathiba marriages; he married a Mokwena woman
from his maternal kin (Bakwena) and who bore a son called Tawana (Tlou 1980: 38).
The BaNgwato had strong customs, and one of them was that Tawana could not
inherit the throne because of his maternal ancestry. Even with these strong customs,
Kgosi Mathiba chose Tawana instead of KgamaІ, who was the son of a MoNgwato
Mohumagadi (the king’s great wife). This created a chaos in the throne and led to
civil war. As such, Kgosi Mathibe turned to the Bakwena (his maternal kin) who
supported Tawana. The Bakwena suggested that Tawana flee to the Khwebe Hills, as
they claimed that the hills were their hunting grounds (Tlou 1980: 38-45).
2.4 SOCIAL INTERACTION
2.4.1 The site
The oral histories and the origins of all the inhabitants, both indigenous and non-
indigenous are interesting. They highlight that the Khwebe Hills hosted diverse
inhabitants that regularly interacted with each other. The oral history speaks of social
interaction between these groups, particularly in commodity exchange (Tlou 1980:
28). These include pelts, karosses and ostrich egg-shell beads, fish and grain.
However, Tlou (1980: 28) discusses how the intangible services were also exchanged.
This includes knowledge on herbs and medicinal skills; these were exchanged from
hunter-gatherers to agro-pastoralists, as they were prominent in curing snake bites,
11
and were knowledgeable about divining (Tlou 1980: 28). Other social interaction
included symbiotic clientship between the Basarwa hunter-gatherers and Bakgalagadi
pastoralists. The hunter-gatherers would herd the Bakgalagadi’s livestock, and at
times hunted for them in return of grain and milk (Tlou 1980: 28).
However, Dziewiecka (2008: 12) describes that the arrival of the BaTawana was
invasive; it shifted the symbiotic clientship relationship, thus causing havoc within the
social structure of the hills (Tlou 1980: 39). Bowen (cited in Mautle 1986: 24) notes
that the BaTawana domination resulted in Bakgalagadi subjugation. This prohibited
Bakgalagadi from owning property or keeping any proportion of their surplus produce
(Mautle 1986: 24). According to the historical accounts, whenever the Batswana
people (this includes the BaTawana and other Tswana speaking groups) interacted
with Basarwa or Bakgalagadi, it would be invasive (Mautle 1986: 22). This is visible
in the oral history that Tlou (1980) narrates; a Mosarwa or Mokgalagadi was turned
into a lekgobe (slave) by a Motswana (Mautle 1986: 24). This had led Dziewiecka
(2008: 12) argue that the BaTawana invaded the Khwebe Hills and colonised the
autochthonous people, thus imposing their way of life and disturbing the indigenous
lifestyle as well as creating long term subjugation.
2.5 BROADER ARCHAEOLOGY
The oral history demonstrates an interesting narration on the interaction between the
indigenous people and the immigrants. It points out that the interaction was more of a
colonisations scenario rather than culture contact. For the purpose of contextualising
the site and its archaeology, the broader archaeological literature is presented below.
This serves the purpose of identifying and understanding the archaeological signature
12
of hunter-gatherers, pastoralists and Tswana agro-pastoralists and the type of
interaction reported elsewhere.
2.5.1 Hunter-gatherers
Previous anthropological and archaeological studies were aimed at exploring the
identity of hunter-gatherers; their socio-economic behaviour, as well as their reaction
to external contact settings (Kent 1992: 10). Schrire (1992: 62) argues that
archaeologist tend to expect hunter-gatherers to show a clear cultural package, that is
documented in the archaeological record. This cultural package should include small
ostrich eggshell beads, a reasonably high percentage of microlithic tools, low density
of ceramics (if any) and a prevalence of hunting wild animals, mainly the small
bovid(s) (Schrire 1992: 62).
This is quite a clear picture of what is expected from hunter-gatherer sites, however, it
is not as simple as it seems and the preservation in southern Africa is always
problematic. Bollong & Sampson (1997: 269) have argued that the 19th
century
ethnographic studies indicate that the southern African hunter-gatherers learnt making
pottery from their contact with Khoekhoe herders. This cultural package approach
seem to homogenise the hunter-gatherers in southern Africa, thus it ignores and
eliminates their experience throughout time and space.
2.5.2 Pastoralists
Early evidence of a pastoral economy dates to 2000 years ago in southern Africa
(Sadr 2003: 195). This early archaeological evidence is from the Bambata cave in
Zimbabwe (Bousman 1998: 135), and it indicates that these migrating pastoralists
were moving into southern Africa with domesticated Orvis-Capris (sheep and goat)
13
from the east Africa region (Reid et al 1998: 81). These earlier pastoralists who were
moving into southern Africa are referred to as proto-Khoe (Ehret 2008: 7-9). Thus, the
term Khoekhoe is used in southern Africa to refer to the pastoralists during the
historic periods of 18th
and 19th
century (Fauvelle-Aymar 2008: 1-2).
Schrire (1992: 63) argues that the Khoekhoe herders have a high density of ceramics
when compared to San hunter gatherers. With that said, the Khoekhoe pastoralist’s
ceramics have diagnostic features such as the bagged shape, spouts, as well as lugs
(Rudner 1979: 11). Earlier studies by Schapera 1933 (cited in Rudner 1979: 8)
indicate that the spouts and lugs were sometimes rubbed with red ochre. According to
Rudner (1979: 8-10), it is these diagnostic features that make the Khoekhoe
pastoralist’s ceramics dissimilar to those produced and used by San hunter-gatherers.
However, Fauvelle-Aymar (2008:77-78) argues that some archaeologists understand
the 19th
Khoekhoe pastoralists as a type of San hunter-gatherer with sheep, ceramic
and at times with cattle, thus, making their archaeological signature somehow similar
to those of hunter-gatherer. This includes artefacts such as large ostrich eggshell
beads, few re-touched stone tools and both domesticated and undomesticated faunal
remains (Schrire 1992: 62-62).
2.5.3 Ethnicity debate
Although the archaeological signature of both Khoekhoe pastoralists and San hunter-
gatherer has been untangled above, their ethnicity is still under debate. The Cape
travellers in the 18th
century identified San hunter-gatherer as Khoekhoe herder
without livestock (Marks 1972: 57). These early writers have argued that the
distinction between San hunter-gatherer and Khoekhoe pastoralist was minimal: “a
San with stock was a Khoekhoe and vice versa” (Marks 1972: 57). This has prompted
14
ethnographic studies to argue that Khoekhoe and San are related people (Barnard
2008:61). Historical investigations have also indicated that San hunter-gatherers
sometimes kept stock, and sometimes lost it, their subsistence depended on changing
historical and environmental conditions (Schrire 1980:10).
According to Barnard (2008: 61-62) the Khoekhoe and San people have unstable
ethnicities; they keep shifting back and forth with the acquisition and loss of livestock
and other material cultures. These fluid identities and unstable ethnicities have led to
the collective usage of the term Khoisan as a reference to both these groups. However,
Bollong (1993: 44-45) has argued that archaeologically we can distinguish Khoekhoe
herder ceramics from those made by hunter-gatherer societies. The Khoekhoe
pastoralist’s ceramics have tampering agent such as sand and fine mineral, while the
hunter-gatherer ceramics have been tempered with organic material such as fibre and
bone as tempering agents (Bollong 1993: 44-45).
2.5.5 Tswana farmers
For the past couple of decades anthropologists and ethnographers have been interested
in different Bantu-speaking communities in southern Africa. One of these groups
includes the Sotho-Tswana groups; this group is made up of communities that have
similar language. This section below presents the broader literature on the
archaeology of the western Sotho-Tswana people, as well as their history.
According to Campbell and Tlou 1984 (cited in van Waarden 1998: 147) the oral
history narrates that the Tswana speaking people in Botswana arrived in two groups;
which is the Bakgalagadi and the Baphofu. van Waarden (1998: 147) narrates that
prior to AD 1400 the Batswana people lived between the upper Limpopo River,
Notwane River, the Molopo and the Vaal River. Studies by Mason 1983 (cited in
15
Boeyens 2003:64) indicate that the Later Iron Age Tswana farmers originate locally in
the South African interior, during the Early Iron Age period. With such evidence van
Waarden (1998: 147) states that around AD 1450 the Bakgalagadi moved across the
Notwane River and settled in Dithejwane Hills west of Molepolole. Subsequently the
Bakwena people in Botswana split from Baphofu during 16th
century, however, the
reason for the split cannot be clearly established (van Warden 1998: 147). The
Bakwena are among the oldest Sotho-Tswana speaking people in southern Africa, and
are spread across certain parts of South Africa and Botswana (Thebe 1996:1).
The 16th
century split within the Baphofu people was not a major political change
compared the one that occurred in the 19th
century. According to the oral history and
the archaeological evidence, it is during the 19th
century that the Tswana polities
underwent social and political transformation (Reid at al 1997: 373). They began to
expand their territories, thus consolidating small neighbouring communities into their
chiefdoms (Reid et al 1997: 373-374). This is evident in megasites such as
Molokwane in north-west province of South Africa (Hall 2012: 301-303) and
Dithejwane Hills in the eastern region of Botswana (Van Waarden 1998: 150). These
archaeological sites have large dry stone wall structures that indicate expansion of
power and political control over territories.
Huffman (2000: 6-9) has used ceramic at these sites and their surrounding territories
to give evidence for the political control around these sites. The Sotho-Tswana
ceramics style has been labelled as Moloko (Huffman 2000: 6-9); Hall (1998: 247)
has also identified variations within Moloko in terms of style, which he also argues as
evidence the political transformation. The oral history of the Khwebe Hills also
narrates the same pattern of the expansion of Tswana polities during the 19th
century.
According to Tlou (1980: 30) the BaTawana were the first people to establish a
16
centralised state in the Khwebe Hills. They incorporated the autochthonous
inhabitants of the region into their state, thus spreading the Tawana cultures and
institutions all across the whole region (Tlou 1980: 30).
2.6 KALAHARI DEBATE
The previous section has discussed the research on the hunter-gatherers, pastoralists
and the Tswana farming community in southern Africa. Furthermore, the first part of
this chapter has discussed that some archaeologists seem to understand hunter-
gatherers as isolated homogenised people (Schrire 1992: 62). This idea of hunter-
gatherers as isolated people has been discussed in the Kalahari debated by both
anthropologists and archaeologists.
The Kalahari debate discusses the role of hunter-gatherers and their relationship with
the outside world (Barnard 1992: 298). This discussion takes place between the
traditional anthropologists (Lee and Solway 1990) and revisionist (Wilmsen and
Denbow 1990). The traditionalist argue that the San Bushmen or the Basarwa are
exponents of hunting and gathering culture and remained isolated until recent times
(Barnard 2006: 2). This isolation restricted these foraging communities from
practicing pastoralism, cultivating land, trading with food producing societies and also
constrained them from having food storage facilities (Headland et al 1989:43).
The revisionist’s regarded southern African hunter-gatherers as an underclass
(Barnard 2006: 2). Denbow and Wilmsen (1986 cited in Sadr 1999:105) have
suggested that all the Kalahari people in the previous two millennia have been linked
to wider social and economic networks; this includes herding and the cultivation of
land. These relations left hunter-gatherer community captured in the political and
economic hierarchy of the Bantu-speaking agro-pastoralists (Sadr 1999: 105). The
17
hunter-gatherers were in contact with the outside world, and worked as herdsmen, for
possibly 150 years (Smith 2001: 16). Their contact was through the political and
social hierarchy, thus leading them into the Mafisa system (Smith 1998: 207). This
system is what Mautle (1986: 24) highlights where Bakgalagadi pastoralists being
subjects that produce food for the Batswana.
The debate also asks questions on whether or not certain features of prehistoric
hunter-gatherers way of life could have survived into the twentieth century (Grauer
2007: 1). Archaeological research based on the Kalahari debate has aimed at
exploring the relation between hunter-gatherer communities and Bantu-speaking agro-
pastoralists (Sadr 1999: 105). This includes research done by Hall (2002) in the
Moloko type settlements in the northwest province, as well as smith’s (1998)
research.
What is interesting about Hall’s (2002) and Smith’s (1998) researches is that is seems
to follow the revisionists ideology of marginalised hunter-gatherer communities.
Smith (1998: 207) argues that the Hunter-gatherers that interacted with Bantu-
speaking agro-pastoralists were trapped in a social and political hierarchy. These
social and political hierarchies include system such as mafisa (see above).
2.7 CONCLUSION
This chapter has illustrated that the oral history narrates the interaction between
hunter-gatherers, pastoralists and farming communities to be an invasive colonisation.
However, the archaeological evidence in southern Africa describes both colonisation,
as well as a more fluid interaction between people of different ethnicity and distinct
economies. This has broadened avenues to further look at interaction from a
18
theoretical perspective. In the next coming chapter, the project will look at how the
theoretical conception of fluidity in interaction is understood.
19
Chapter 3: Theory chapter
3.1 INTRODUCTION
In the last several decades the archaeology discipline has witnessed the rebirth of
theoretical interpretation of intersocietal interaction (Schortman & Urban 2010: 3).
This has shaped the conceptualisation of frontiers, boundaries, colonisation and
culture contact (Lightfoot & Martinez 1995: 471). These terms on their own seem to
create debates within the archaeological discipline. As such, the aim of this chapter is
to define these concepts, to understand how the colonial theory and post-colonial
approaches are applied to interaction studies.
3.2 TRADITIONAL DEFINITIONS
The word frontier has been defined by different social science scholars. According to
Humphrey (1988: 3) frontiers can be described as regions where regular interaction
takes place between two or multiple culturally distinct communities. During this
interaction, one of the communities will attempt to take control over the others
(Humphrey 1988: 3). This usually happens when an occupied territory expands over
its horizons, thus increasing the settlement density and competition (Hudson 1969:
365). The expansion into these territories leads to colonization; whereby the new
people invade a new area, extending its habitation beyond the borders of their
environment (Hudson 1969: 366-367). This process produces consistency in
settlement pattern where the indigenous people remain the minority in that space
(Hudson 1969: 365).
However, according to Silliman (2005: 55-59) archaeological research on interaction
seem to be solely understood as colonisation. It is conceptualised that frontiers serve
20
as territory makers; they define colonial territories from the remote hinterlands of
indigenous people (also see Lightfoot & Martinez 1995: 471-472). The English
dictionary defines frontiers as a synonym of a borderline, which is explained as an
ambiguous in-between state. These frontier states are situated literally on the
separating line between two countries (Kopytoff 1987: 9).
In early European anthropology reports, frontiers were defined as territories controlled
by a white man to eliminate and maintain a boundary between him and the
unoccupied land (Naum 2010: 101). While at the same time frontiers can be
understood as geographical barriers (Naum 2010: 101). These geographical barricades
also limit any social and political contact on outlying areas (Naum 2010: 101-102).
Brown and Gibson 1983 (cited in Rockman 2003: 14-15) argue that frontiers can
range from being totally infiltrate boundaries to strainer boundaries. They could also
be uninhabitable or restrict any movement from one area to the other (Rockman 2003:
14-15). Nevertheless, frontiers develop, change and dissolve over time (Rockman
2003: 14-15). Thus, they can develop to become isolated independent political units
(Naum 2010: 101-1-2). Once the frontiers are developed as isolated political entities,
they become visible in the landscape, subsequently asserting that they were
established to control, separate and demarcate territories (Naum 2010: 102-103).
3.3 COLONIAL AND POSTCOLONIAL THEORY
Archaeological studies have historically understood interaction using the colonial
paradigm (Naum 2012:101). This colonial framework suggests that interaction
between people of difference will create frontiers. Anthropologists who use this
theoretical framework claim that frontiers develop when colonial people or core
polities migrate into open air or sparsely developed lands that are close or far in
21
distance (Lightfoot & Martinez 1995: 472). These territories that are now frontiers
will become territorial markers, which demarcate boundaries of colonial expansion.
They will be maintained with the purpose of creating permanent segregation between
the new comers and indigenous inhabitants (Lightfoot and Martinez 1995: 472).
However, interaction is arguably more complex and multifaceted than what the
colonial approach presents. The complexity of interactions between people of distinct
ethnicity has been highlighted by the post-colonial theory (Gosden 2004: 18). This
theoretical framework derives from the decline of colonial structure (Gosden 2004:
18). It readdressed interaction studies and the ideas on identity transformation; post-
colonial theory analyses how culture contact has affected the identity of both the
colonised and the coloniser (Young 2001: 342-357). It takes into account that
colonists can change the identity of the colonised, while at the same time the
colonised can also change the colonialists identity (Young 2001: 342-357). Bhabha
(1994:171) states that the post-colonial theory questions the beliefs that colonialism
instituted, and this is applicable in interaction studies in archaeology.
Lightfoot and Martinez (1995: 472) argue that the colonial paradigm dismisses the
idea of frontiers being fluid. As such post-colonial theory in archaeology suggests that
frontiers should be reconceptualised as regions where social networks are entered and
explored by people of different ethnicity (Lightfoot & Martinez 1995: 472). However,
van Dommelen (2005: 112) argues that post-colonial theory is still western way of
thinking about the presentation of colonial structures and situations. Thus, post-
colonial theory in archaeology still views culture contact between the different people
as colonisation. Where one community takes control over the other, and identities of
both will be transformed.
22
3.4 APPLICATION IN ARCHAEOLOGICAL INTERACTION STUDIES
The colonial theory has suggests the concept of frontiers as impermeable cultural
barriers that restrict social interaction, filter information, and limit the movement of
material culture between people that are on opposite sides of the frontier (Lightfood &
Martinez 1995: 473). However, according to Naum (2010: 102) post-colonial theory
in archaeology proposes the idea that frontiers remained fluid zones, where tensions,
battles and friendly relations take place.
Thus, Frontiers emerge as fragmented landscapes that are fluid social and cultural
regions (Naum 2010: 102). The argument brought forward by post-colonial
archaeology is that we should discontinue this conception of understanding frontiers
as physical cultural boundaries that constrain intercultural relationships (Schortman &
Urban 1992: 427). Frontiers can act as geographical markers; however, they should be
seen as zones that permit interaction between people from different homelands
(Schortman & Urban 1992: 10-11).
Schortman & Urban (1992: 427) have argued that when polities expand beyond their
territories the people who lie on the other side of the periphery or frontier are not
necessarily passive like the colonial theorists make them to be. They are rather active
participants in interregional processes of cultural exchange and culture transformation
(Schortman & Urban 1992: 427).
3.5 MATERIAL CULTURES AS BOUNDARY
Thomas (1999: 182) has argued that material culture and technologies have been used
as indicator of culture transformation, particularly in the colonisation phase.
Archaeologists working around the world have been interested in studying social
23
boundaries through the distributional of material culture (Stark et al at 2000: 269).
Earlier anthropological studies have highlighted external forces influencing the
change, while neglecting the importance of internal social, political, and economic
processes (Stein 2002: 93). Stein (2002: 93-94) has argued that power and political
control have been misinterpreted as one-way flow from complex polities to less
developed societies (Stein 2002: 93-94).
In southern Africa it has been observed that when hunter-gatherer communities
interact with Bantu-speaking agriculturalists the material culture of these hunter-
gatherer bands transform (Thorp 2010: 113). Sadr (2002: 44) argues that hunter-
gatherer sites close to farming community settlements seem to have high amount of
farmer’s artefacts. However, this ideology of understanding political control as one-
way situation is not applicable in southern Africa. As Sadr (2002:44) has argued that
sites of the farming communities that are close to hunter-gatherer camps also tend
have hunter-gatherer material culture.
3.5 SOUTHERN AFRICAN ARCHAEOLOGY
Until recently, southern African archaeologists have distinguished Later Stone Age
(LSA) Early Iron Age (EIA) sites and people according to their material culture. This
differentiating ordered phases, industries and complexes into successive sequences
(Phillipson 2005: 4). Thus, it led to the creation artificial compartments that continue
both in time and in space, (Phillipson 2005: 4), consequently creating impermeable
boundaries between the LSA and EIA. However, Alexander (1984: 12) argues that the
EIA farmers in Southern Africa had moving frontiers that permit them to explore and
exploit hunter-gatherers inhabitants and economy. These frontiers were static; they
24
developed over a long term relation between these farmers and the hunter-gatherers
(Alexander 1984:13).
3.5.2 Proposed approach
The southern African archaeology does not neglect the idea of colonisation; however
it also takes note of a symbiotic culture contact. Frontiers can be understood as
abstract area where people of different ethnicity and economies meet. This approach
takes into account that interaction influences changes in frontiers, in a way that it
exposes the cultural process, such as the maintenance or the integration of alternative
culture (Emberling 1997: 316). Social boundaries can be fluid allowing a variety of
interactions, which are often overlapping, such as exchange, competition,
colonization, and conquest.).
Sadr (2008: 104) for example, argues that the frontiers between Later Stone Age and
Early Iron Age people and their economies were more fluid and permeable. These
permeable boundaries allowed people to incorporate different economies and material
culture into their own, and move in and out of different economies to sustain their
diets during dry seasons or voluntarily. As such, it is suggested that the site of KWH4
is a frontier between the BaTawana, the Basarwa and the Bakgalagadi. This site
indicates the birth of a Tawana polity and its expansion; however, I argue that the hills
experienced culture contact rather than colonisation. The autochthonous people were
not passive recipients of the Tawana culture and institutions; rather they were active
donor in the cultural transformation of the Tawana polity.
25
3.6 CONCLUSION
Frontiers should not been seen as geographical boundaries that keep people of
different cultures separate. The post-colonial theory seems to understand that people
can interact where by changes will occur. These changes could be from the core polity
to its surroundings periphery and vice versa. However, the initial stages of interaction
should not be seen as colonisation where one group conquers the other. Early stages
of interaction should be understood as culture contact, where people can freely
exchange commodities and intangible material. Southern African archaeology should
re-address the early research interpretations on sites that hunter-gatherers and farmers
interacted. Archaeologists who focus on interaction studies should be aware of the
difference between culture contact and its changes, compared to colonialism and its
changes.
26
CHAPTER 4: METHODOLOGY CHAPTER
4.1 INTRODUCTION
The site of KHW4 is located on the foothill of one of the Khwebe Hills in Ngamiland
(Ashley 2009: 37). It is a stone wall site associated with the BaTawana polity during
the 19th
century. The excavations were conducted over three seasons; these were in
2008, 2009 and 2010. A total number of nine features were excavated and all features
had multiple layers (contexts). Feature 297, which is a midden, yielded a large
proportion of the ceramics that was analysed for the dissertation. The aim of this
chapter is to discuss the methods used in the analysis of the ceramics from KWH4. In
addition a new typology will be presented; this will be compared with existing
ceramic typologies of the Tswana agro-pastoralists and Khoekhoe pastoralists in
southern Africa. This will assist to contextualise the archaeology of site KWH4 to the
broader southern African archaeology.
4.2 PREVIOUS RESEARCH
The initial research conducted in 2008 at the site of KWH4, was aimed at exploring
the impact of the BaTawana migration to the hills on the autochthonous hunter-
gatherer bands, as well as the pastoral community (Diskin & Ashley: in prep). This
was later followed by visual and analytical investigations. The visual analysis
indicated three fabrics and two distinct stylistic groups. This visual analysis was also
used collectively with the XRD (X-ray Diffraction) analysis. This analysis is a
spectroscope technique used to identify the crystalline material and the provenance of
raw material (Druc 2001: V). The analysis also identifies the arrangement pattern of
the atoms in the minerals (Druc 2001: V).
27
The XRD analysis was conducted on thirty four sherds and it established two
provenances. The XRD analysis also characterised the ceramics fabric in the
following, Fabric 1 clay mineralogy comprised of quartz and plagioclase along with
hydroxylapatite and bone (Diskin & Ashley in prep). The presence of such organic
temper was association with a nomadic hunter-gatherer community (Diskin & Ashley
in prep). Fabric 2 comprised of different quartz, plagioclase and k-feldspar along with
muscovite which sometimes dehydroxylted and this fabric was associated with
pastoralists. Fibre 3 comprised different quantities of quartz and albite and this fabric
is associated with the BaTawana (Diskin & Ashley: in prep).
The XRD analysis indicated that the provenance of raw material for Fabric 3 is totally
different to that of Fabric 1 and Fabric 2. As such, Diskin and Ashley (in prep)
research has created avenues that need to be explored in terms of understanding the
stylistic variability between and within these fabrics.
Figure 4.1: An image of site KWH4 and the stone walling
28
4.3 CURRENT INVESTIGATIONS
4.3.1 Methods
Current studies were conducted with the aim to create a new ceramic typology for the
Khwebe Hills. The ceramic sherds were first divided into diagnostics and non-
diagnostics. The diagnostics sherds with clear rims profiles were analysed and this
includes both decorated and undecorated sherds. The non-diagnostic sherds were all
body sherds, and these were analysed for the presence or the absence of surface
treatments, however, the analysis for the body sherds was kept to a minimal.
The rims sherds were firstly placed on a flat surface in order to verify the orientation
of the lip and vessel form. Once this was established, these rims were placed on a
printed rim chart, whereby an approximate diameter of the vessel mouth was
established.
Decorations were document on all the rim sherds; these included identifying
decorations on the lip, the interior of the lip, the rim, neck area, shoulder area, as well
as the body. Features such as spouts, lugs and perforations were also noted, in
addition to this, surface treatment such as burnish was also recorded. Each analysed
ceramic sherd was documented into an excel sheet database (See table 4.1).
4.4 TYPOLOGY
4.4.1 Vessel shapes
The collection has two main types of vessel shapes, bowls and jars. Within the bowl
shape category only the hemispherical bowl (HB) was noted. The jar category
produced two distinct sub-categories, the long neck jars and the short neck shape jars,
and these vary both in size and neck orientation. Jars that fall within the short neck
29
jars category include the Short Neck Recurved (SNR), Pronounced Short Neck
Recurved (PSNR), Everted Short Neck Recurved (ESNR), and Short Everted
Pronounced Neck (SEPN) (see figure 4.1).
Tab
le 4
.1:
an i
llust
rati
on o
f w
hat
the
KW
H4 c
eram
ic d
atab
ase
looks
like.
Note
the
dif
fere
nt
attr
ibu
tes
that
wer
e re
cord
ed f
or
each
sher
d
30
While the long necked jars category has five variations, this includes the Everted
Pronounced neck (EPN), the Straight Pronounced Neck (SPN), the Everted Recurved
Neck (ERN), Straight Recurved Neck (SRN), as well as Collared Incurved Neck
(CN). See figure 4.2 for the rim profiles of the Long Necked vessels.
Short Necked
Jars
Short Neck
Recurved
(SNR)
Pronounced
Short Neck
Recurved
(PSNR)
Everted Short
Neck Recurved
(ESNR)
Short Everted
Pronounced
Neck (SEPN)
Profiles
Long
Necked
Jars
Everted
Pronounced
Neck (EPN)
Straight
Recurved
Neck (SRN)
Everted
Recurved
Neck
(ERN)
Straight
Recurved
Neck
(SRN)
Collared
Incurved
Neck (CN)
Profiles
Figure 4.1: Illustration of rim profiles of Short Necked
Jars rim profiles
Figure 4.2: Illustrations the rim profiles of the Long Necked Jars
31
4.4.2 Lips
The analysis indicated that within the collection we have five generic types of lip
formations (see table 4.4). It must be noted that there are variants within the lip form,
there is internally bevelled and externally bevelled. Other variations of the lip forms
include the square rounded that is over-hanging, as well as a round over-hanging lip.
Rounded
Square
Bevelled
Square
rounded
Overhanging
4.4.3 Decorations
All sherds were analysed for any surface treatment this includes the diagnostic and
non-diagnostics. The decoration was order into three categories, which includes the
impression motifs, the incision motifs and the paint motif.
4.4.3.1 Impression motif
Punctates Round Punctates Punctate with
incision
Punctate with
redslip (band)
Punctate with
motif
Table 4.4: Lip forms from the KWH4 collection
Table 4.5: A record of the all the different impressed motifs within the KWH4 collection.
32
4.4.3.2 Incision motif
4.4.3.3 Paint motifs
The decorations have three groups (techniques) of decorations. The first group is the
impressed decorations; this group includes punctates (that vary from single band to
double band), the round punctates that also come in a variation of single to double
band. Other types of impression decorations include punctate with slip, this includes
either red band or black band of slip. This group also has decorations such as punctate
with incision, as well as punctate with motif (see table 4.5).
The incision category illustrate single band and double bands are used as decorations.
Grooves are also applied as decoration motif and these also vary from single grooves
to double grooves. Other types of decorations include incision or grooves with
punctates, as well as incised comb (Table 4.6). The collection also indicates that paint
was also applied as decorations. These vary from red band or black band of slip, slip
Incisions Grooves Incisions/grooves
with punctate
Incised comb
Band of slip
(red/black)
Slip motif Diagonal bands of
slip (red/black)
Chevron Motif
Table 4.6: A record of the all the different incision motifs within KWH4 collection.
Table 4.7: A record of the different paint applied decorations within KWH4 collection.
33
motif, to diagonal bands of slip either red in colour or black, as well as chevron motif
(see table 4.7). This last category of paint decoration can be used with impressed
motifs or the incisions.
4.4.4 Lugs
The KWH collection also has lugs within it, and three types of lugs were recorded
include the square lug, the round lug and the elongated lug. The square and the round
lug indicate that they are perforated, either vertically or horizontally. While the
elongated lugs remain imperforated, however, they tend to have decorations as part of
their surface treatment (see table 4.8 and figure 4.2).
4.5 Archaeological evidence
Square lug Round lug Elongated lug
Table 4.8: A record of all three different lugs found within the KWH4 collection.
Figure 4.2: An illustration of a sherd with an elongated lug decorated in a
band of punctate, as well as incision.
34
4.5. Archaeological evidence
Pastoralist’s ceramics
The ceramic typology from KWH4 seems to connect with the work done by Sadr and
Sampson (1999). According to Rudner (1986: 13-20) the Khoekhoe pastoralists
ceramics have a distinct attribute, which is the amphoral in shape. Sadr and Sampson
(1999) have identified different wares from the Seacow collection.
Spouted Ware, Incised typed (SPINC)
According to Sadr and Sampson (1997: 6) the SPINC ware appear archaeologically
before AD 400 and this ware continues to AD 600. The SPINC vessels are
characterised by short neck with no lugs, however, spouts on the shoulder are present.
The neck shape varies from flared, vertical to convergent. The lip form is usually
bevelled with a slightly thickened mouth, which is approximately 100 mm in
diameter. The neck is decorated with horizontal incised lines that may be fine or
broad. The decorations may also extend from rim to the body, and these occur
together with short diagonal, herringbone, or cross dashes on the lip of the spout and
the mouth (Sadr and Sampson 1999: 6). (See figure 5.3).
Spouted Ware, Impression type (COMB Ware)
In the Western Cape region the SPINC was replaced by impression only motif after
AD 650 (Sadr & Sampson 1999: 6). The COMB ware retained the vessel forms as
those of SPINC; however, the decorations seem to be restricted on the lip of the
mouth and the spout. The decoration is made by impressing a notched implement,
which Sadr & Sampson (1999: 6) suggest that it is possibly a white mussel shell.
35
The Seacow collections also indicate that the vessel basses were at times decorated
with such a motif. The notches impressions vary from small notches to thick in size; it
is argued that these were possibly made by ostrich eggshell fragments (Sadr &
Sampson 1999: 6). This created a row of short comb-stamping and this was applied as
a single band. The comb-stamping was not applied on the spouts and the basses of the
vessels (Sadr & Sampson 1999: 6). The mouth radius of this ware can be divided into
two classes, a small class with radius less than 100 mm and a wide-mouth class that
has radius that 100 mm and more. This Ware illustrates that vertical necks were
preferred; however, the small mouthed vessels class have definitive bevelled lips.
While the wide mouth-vessels also had vertical necks with simple round lips (Sadr &
Sampson 1999: 6).
Lugged Ware, Undecorated Type (LUND)
The Archaeological record of south-western in the Western Cape Province indicates
that the LUND ware appears from AD 850 and lasts until Ad 1250. During the period
of AD850 the spouts ware disappears and it is replaced by an undecorated vessels
type, which has a long neck with variable rim orientations (Sadr & Sampson 1999: 7-
8). The mouth of the LUND vessels is wide, and it has a flat topped lip. This ware has
no presence of spouts; however, two horizontally pierced lugs occur on opposite sides
of the shoulder area (Sadr & Sampson 1999: 8).
Lugged Ware, Incised Type (LINC)
The Lugged Incised type vessels appear in the archaeological record AD 1250,
particularly in the Western Cape Province. This ware has the same lugs and tall necks
as the LUND ware. However, the lips are more thickened with flat lips occurring
throughout this ware. The neck of ware is decorated with horizontal incisions, which
36
are often zoned on intersection where the neck and the shoulder meet. These
horizontal incisions at times extend over the shoulder to the body. Sadr & Sampson
(199: 8) state that the seven sherds from the Seacow collection had a combination of
both horizontal incisions as well as punctates. A small percentage of the LINC ware
showed that diagonal lines on the shoulder area and the body are present (Sadr &
Sampson 1999: 8).
Untyped Body Sherds with Punctates
This Category includes non-diagnostic body sherds that are decorated with a single
band of punctate impression.
Contextualisation
There seems to be a huge gap in the literature of the Khoekhoe pastoralist’s ceramics,
particularly the 19th
century typology. There is no sufficient information and evidence
Figure 4.3: An illustration of the amphorae shapes and the different Khoekhoe wares
from Seacow in the Western Cape Province, South Africa (Sadr & Sampson 1999: 8)
37
that indicates the possible dates where the LINC ware discontinues in the
archaeological record. Nevertheless, the typology presented by Sadr & Sampson
(1999) cannot be fully applied to the Khwebe Hills, as there are dissimilarities in both
time and space.
The KWH4 collection is not similar to the SPINC, as no spouts were recorded within
this collection. The Spouted Ware, Impression type (COMB Ware) is also absent
within this collection. However, the collections closely shows pattern with the two
categories that Sadr & Sampson (1997: 8) have identified, which are the Lugged
Ware, Undecorated Type (LUND), the Lugged Ware, Incision Type (LINC). The non-
diagnostic body sherds also indicate Untyped Body Sherds with Punctates, which
Sadr and Sampson (1999: 8) have identified within the Seacow ceramic collection.
The archaeological research by Huffman (1994) from the same region as the Khwebe
Hills indicates a clear pattern than the Seacow typology. This site of Toteng yielded
62 ceramic sherds that are associated to the Khoekhoe pastoralists. These vessels were
short neck jars that had red-buff chevron; Huffman (1994: 4) has associated this to
historical herders. The vessels also indicated that they were bagged shaped with
pierced lugs. They had thickened jar rims that were decorated with crosshatching
(Huffman 1994: 4), (see figure 4.4 and 4.5)
Figure 4.4: An illustration of the plain bowl with perforations on the
surface (Huffman 1994).
38
The ceramic vessels that date to the historical periods seem to have short straight
necks with flat lips. Huffman (1994: 4) argues that lugs are possibly part of these
vessels; the decorations vary from punctates, red bands, or red chevron motifs and
small applique lumps on shoulders. This ceramic typology from Toteng indicates the
same pattern as the Khwebe Hills collection. The Khwebe Hills collection has a
variation of long neck and short neck, perforated lugs are present, and decorations
such as punctates, chevron motif and bands of red slip have been noted.
Tswana ceramics
The literature indicates that the earliest type of Tswana ceramic facies of Moloko is
Icon (Hanisch 1979; cited in Huffman 2007: 429). Huffman (2007: 420) points out
that this type of pottery occurs in regions such as Limpopo Province and Mpumalanga
Province in South Africa, as well as in Botswana. The Icon facies date from AD 1300
to AD 1500 and the archaeological evidence indicate that in some regions the Icon
incorporates earlier elements of Eiland (Huffman 2007: 9). The Icon facies features
jars and bowls, all these type of vessels have distinct multiple decorations themes, on
the rim and neck. These decorations come in a form of colouring as chevron and
lozenges motifs (Huffman 2002: 9).
Figure 4.5: An illustration of thickened jar rim with crosshatching as
decoration (Huffman 1994).
39
Within Moloko style pottery there are multiple facies, the first phase has been
discussed above. However, it the transition between phase A, phase B and phase C
that archaeologists have been interested in. From figure 4.6 it is noticeable that Icon
and Ntsuatsatsi have different origins, the later coming from KwaZulu-Natal and the
other from the northern regions of south Africa (Huffman 2002: 9). New evidence
provided by Huffman (2007: 186) illustrate new facies within the Moloko style, it also
does not take note on Ntsuatsatsi. The new sequence by Huffman (2007: 186)
indicates that Madikwe is replaced by Marataneng facies (See figure 4.7).
Figure 4.6: An illustration of the different facies within Moloko Branch (Huffman
2002: 7)
40
According to Mason (1986: 723) the Tswana ceramic underwent three transitions
(Hall 1998:249).The first phase and second phase ceramics are highly decorated and
more elaborate than the third phase (Mason 1986: 723-724). Hall (1998: 249) used the
multivariate system that Huffman 1980 (cited in Hall 1998: 249) proposed to re-
analyse the Tswana ceramic typology. This system incorporates the analysis of the
vessels shapes, the decorations and the position where the decorations lie. Hall (1998:
249-252) documented all the fifteen ceramic classes with their stylistic variation (see
table 4.9).
4.5.2 Decorations
According to Hall (1998: 248-250) phase A and Phase B are more elaborated, phase A
has class 1-13 and it has more hemispherical bowls that are elaborately decorated than
phase B and Phase C and on the constricted pot . Phase B consist of class 1, class 2,
class 3, class 5, class 6, class 9, class 9, class 11, class 14 and class 15 (see table 4.9)
Hall (1998: 249-253) points out that the Tswana ceramics post AD1700 are not highly
? Thabeng Buispoort Marateng
Letsibogo Olifantspoort Madikwe
Icon
Figure 4.7: An illustration of the sequence of Moloko type ceramic (Huffman
2007:186).
41
decorated, and he has identifies that only class 1, 2, 3, 6 and 13 feature in this third
phase (see figure 4.8).
Class
number
Description of the class Phase
1 Recurved jar with a single band immediately below the A, B, C
2 Recurved jar decorated in one zone of the neck A, B, C
3 Recurved jar with a band immediately below the rim and a band
below the neck
A, B, C
4 Recurved jar with a band immediately below the rim, a band on
the neck, and a band of decoration extending below the neck onto
the body
A
5 Recurved jar with multiple bands extending to the base of the
neck, a band or arcades on the shoulder, extending onto the body,
and isolated triangles low of the body
B
6 Open bowel with a single band of decorations immediately below
the rim
A, B, C
7 Open bowl with a band of decoration immediately below the rim
and a band on top of the rim
A
8 Open bowl with decoration in one zone of the body A
9 Open bowl with a band of decoration immediately below the rim
and a band on the body
A, B
10 Open bowl with decoration on the top of the rim, a band of
decorations immediately below the rim and a band on the body
A
11 Open bowl with multiple bands on the body with a band of
chevrons or arcades extending very low of the body
A, B
12 Open bowl with a wide band of decoration below the rim and
which is decorated in the same zone internally, or has alternating
bands of graphite and ochre all the way to the bottom
A
13 Constricted pot with a wide band of decoration below the rim A, C
14 Constricted pot with a band of decoration immediately below the
rim, followed by a wide of decoration down onto the upper body
B
15 Constricted with single band of decoration immediately below the
rim
B
Table 4.9: A detailed record and description of the different classes found in Moloko
ceramic phase (Hall 1998: 250)
42
4.6 Conclusion
The ceramic collection from KWH 4 seems to fit in with some aspects of t Sadr and
Sampson (1999) LUND ware from Seacow in the Western Cape Province, South
Africa. However, the Toteng ceramic typology shows more similarity in style,
particularly the Khoekhoe typology and the historic typology. The typology of the
Khwebe Hills also depicts that it falls into the third phase of the Tswana ceramic that
Hall (1998) has identified. Thus, the next chapter will explore the KWH4 collection
using analytical data to further understand and verify if the KWH4 collection falls
within these existing typologies.
1 2 3
6 13
Figure 4.8: Ceramic profiles and their decorations within the third phase of the Tswana
ceramic sequence (Hall 1998: 253)
43
Chapter 5: Result Chapter
5.1 INTRODUCTION
The previous chapter has discussed and illustrated that Diskin and Ashley (in prep)
have analysed the collection visually and through XRD analysis. This was to
distinguish the different fabrics within the KWH4 ceramic collection and their
provenance. Their analysis has been supplemented by a typology which is also
presented in the previous methodology chapter. This chapter aims at presenting the
data that illustrate the relationship between the three fabrics and the typology from
chapter 4.
5.2 ANALYSIS
The analysis was done on all 132 rim sherds from the KWH4 ceramic collection and
no representative sample was selected.
5.2.1 Fabrics
5.2.1.1 Fabric 1
According to Diskin and Ashley (in prep) fabric 1 is possibly associated with the
hunter-gatherer community in the Khwebe Hills. The mineralogy of this fabric
includes quartz, plagioclase as well as hydroxylapatite bone (see figure 5.1 and 5.2).
Bollong et al (1993: 44-45) have argued that organic temper such as bone is
associated with nomadic hunter-gatherer communities.
44
5.2.1.2 Fabric 2
The XRD analysis has indicated that F2 mineralogy consists of quartz, plagioclase
and K-feldspar along with muscovite and at times with dehydroxylted (Diskin &
Ashley: in prep). The visual analysis has indicated that F2 clay is not well sorted, thus
it has coarse inclusion and bristly feeling to it. F2 also indicates poor quality in its
manufacturing, particularly during firing process (see figure 5.3 and 5.4). Diskin &
Ashley (in prep) have associated this fabric to the pastoral society in the community
in the Khwebe Hills.
Figure 5.1: A picture of F1 with calcite
and bone temper and F3 attributes
Figure 5.2: An illustration of a sherd section
with clear bone temper within F1
Figure 5.3: An illustration of Fabric 2,
note how poorly fired the sherd is.
Figure 5.4: A picture illustrating the
coarse inclusion of Fabric 2.
45
5.2.1.3 Fabric 3
The visual and XRD analysis have indicated that F3 is dissimilar in mineralogy when
compared with F1 and F2. This fabric mineralogy comprised of different quantities of
quartz and albite (Diskin and Ashley: in prep). Unlike F1 and F2, the clay of F3 is
well sorted, it has a smooth feeling on the surface and it is very hard, thus indicating
that the vessels were well fired (see figure 5.5 and 5.6). Diskin and Ashley (in prep)
have associated F3 to the incoming agro-pastoralists BaTawana.
5.2.1.4 ANOMALIES
The collection also comprises of anomalous fabrics that show multiple characteristics
associated either with F1, F2 or F3. For example some of the ceramic sherds that
illustrate characteristics of F2 (coarse inclusions) have bone as a tempering agent.
Within F3 there are sherds that have coarse inclusion, which is associated with the
pastoral economy and society in the hills. Other F3 ceramic sherds illustrate that bone
Figure 5.5: A pictures of F3 fabric and
the different decorations within this
fabric
Figure 5.6: A picture of the rim
sherds within fabric 3
46
has been added as a tempering agent. However, these anomalies have not been treated
as separate entities. They have been incorporated and analysed as part of the generic
fabrics that Diskin and Ashley have created.
5.3 DATA PRESENTATIONS
The table below indicates the total number of rim sherds from KWH4, it also illustrate
how many rim sherds form part of the different fabrics. A total number of seven rim
sherds out of the 132 were too small for any type of analysis. In addition to this, four
rim sherds were miscellaneous, thus they could not be categorised either as F1, F2 or
F3.
Fabric Total number of sherds
F3 82
F2 22
F1 14
UNIDENTIFIABLE 7
TOO SMALL FOR ANALYSIS 7
TOTAL: 132
The data from the table was converted into a pie chart to understand the ratio of
fabrics when compared to each other (see figure 5.7). As the graph illustrates fabric 3
makes up (n=82) 62% of the 132 sherds, thus it forms the largest part of this
collection. Fabric 2 contributes (n=22) 18% to the entire collection and this if
followed by Fabric 1 contributing (n=14) 10%. The sherds that were too small for
analysis and those that could not put into specific categories contributed (n=7) 10% to
the collection.
Table 5.1: Documentation of all the different fabric with KWH4 ceramic collection
47
5.3.1 Fabric 1
5.3.1 Fabric 1
As discussed earlier, Fabric 1 forms the smallest part of the collection and this
includes both the bowls and jars. Within the jar category there is a variation of short
neck jars and the long neck jars. Table 5.2 below indicates that F1 has more short
neck vessels than the bowls and the long neck jars.
Vessel form Total number
Bowls 1
Short Neck Vessels 8
Long Neck Vessels 4
The graph below illustrates the variations of short necks and long neck vessels within
fabric 1. The short neck jars seems to be limited to two types, and this includes the
Short Neck Recurved (SNR) and the Pronounced Short Neck Recurved (PSNR).
While the long neck jars shape vary from the Collared Incurved Neck (CN), Everted
F3
F2
F1
UNIDENTIFIABLE
TOO SMALL FOR
ANALYSIS
Figure 5.7: Pie chart graph illustrating the different fabrics and its ratios compared to
the others.
Table 5.2: Record of vessel shapes within fabric 1
48
Recurved Neck (ERN), and Straight Pronounced Neck (SPN). Three of the sherds that
were analysed from F1 were body sherds, thus the vessel shape could not be
identified. From F1 it can be clearly noted that only one hemispherical (HB) forms
part of this fabric.
Rim diameter
The table below (see table 5.3) presents the lip range of the different vessel types
within F1, as well as the average rim size. To calculate the average size of the rims
the following formula was used:
∑
Where ∑ is the sum of all the data values in the sample, n is the number of the
observation in the sample.
Table 5.3 below indicates that the rim size of both the short neck jars and the long
neck vessels almost have the same. The rim diameter of the Collared Incurved Neck
(CN) and the Hemispherical Bowl (HB) could not be identified as they were small for
0
1
2
3
4
5
6
SNR PSNR CN ERN SPN N/A HB
SNR
PSNR
CN
ERN
SPN
N/A
HB
Figure 5.8: An illustration of the long necked and short neck jars and their
occurrence
49
this type of analysis. The Short Neck Recurved (SNR) rim size ranges from 22cm-
34cm, while the Pronounced Short Neck (PSNR) ranges from 14cm-21cm. The long
neck jars indicate that there are jars with small mouths, such as the Straight
Pronounced Neck (SPN) the rim size of 14cm.
Vessel form Rim Range Average Diameter
Short Neck Recurved (SNR) 22-34cm 17.6cm
Pronounced Short Neck
Recurved (PSNR)
14-21cm 18cm
Hemispherical Bowl unidentifiable -
Collared Incurved Neck (CN) Unidentifiable -
Everted Recurved Neck (ERN) 17 mm 17cm
Straight Pronounced Neck (SPN) 14-36cm 25cm
Decorations and surface treatments
Decorations keys
P Punctate IC Incised Comb
RP Round Punctate RS Red slip
PS Punctate Slip RBS Red Band Slip
PM Punctate Motif BB Black Band
PI Punctate with incision CM Chevron Motif
CS Comb Stamping B Burnish
I Incision
G Grooves
Table 5.3: Illustrating the average size of the vessels within fabric 1
Table 5.4: A record of keys used to indicate the different decorations within the
collection.
50
Lugs
Shapes of lugs
RL Round Lug
EL Elongated Lug
SL Square Lug
VP Vertically Perforated
HP Horizontally Perforated
The analysis has indicated that there are three types of lugs, the round lug, the square
lug and the elongated lug (see table 5.5). The analysis has also indicated that the lugs
are perforated, either horizontally or vertically, and the analysis has indicates that lugs
are absent within fabric 1.
From the table above (see table 5.6) it clear that ceramic sherds from F1 are not
densely decorated. They seems to be limited to incisions as well as red slip, while
burnishing as a surface treatment does not occurs regularly. The SNR seem to only
have incision as decorations. While burnish as a surface treatment only occurs in the
PSNR and grooves are used for SPN jars.
Vessel
form
P RP PS P PM PI CB I G RS RBS BB C S B
SNR X
PSNR X
HB
CN
ERN
SPN X
Table 5.6: A record of the different vessels shapes and the various decorations
Table 5.5: Keys to be used to indicate the different lugs in the collection.
51
Lips
The graph below (see figure 5.9) illustrates the frequency of these lip forms within the
F1 fabric. The analysis indicates that there are six square rounded (S, R) lips, five
Bevelled (B). There also seems to be an equal proportion of the square and the round
lip forms in this fabric, while there was only one sherd with a round overhanging lip.
The graph also indicates that the Square Round lip is preferred when compared to the
other lips.
5.3.2 Fabric 2
Fabric 2 contributes (n=22) 18% of the 132 sherds that were analysed. The Fabric 2
data was approached the same as Fabric. In F2 there seems to be a preference for the
long neck jars than the short neck jars. This is shown in table 5.4 that illustrate that
the number of long neck jars is almost twice as that of the short neck jars. It is also
quite clear that within F2 there is no abundance of bowls due to the fact that no bowls
were identified.
0
1
2
3
4
5
6
7
B S R R,O S,R
B
S
R
R,O
S,R
Figure 5.9: illustration of the different lip forms and their frequency within
fabric 1
52
Vessel type Total number
Bowls -
Short neck jars 9
Long neck jars 4
Not-Applicable (body) 6
Unidentifiable 3
The table above was converted into a graph with the aim of indicating the variations
of vessel forms with fabric 2 (see figure 5.10). Note how the short neck jars are
limited to the SNR and the PSNR, while the long neck jars show a variation, this
includes the EPN, SPN, ERN and the SRN. All three of the long neck jars (SPN, ERN
and SRN) occur equally and only one EPN was recorded.
0
1
2
3
4
5
6
7
SNR PSNR EPN SPN ERN SRN N/A UN-ID
SNR
PSNR
EPN
SPN
ERN
SRN
N/A
UN-ID
Table 5.7: A record illustrating the number of vessels type within fabric 2
Figure 5.10: The occurrence of long and short neck vessels within fabric 2
53
Rim Diameter
∑
Where ∑ is the sum of all the data values in the sample, n is the number of the
observation in the sample.
Table 5.8 illustrate that the short neck jars have larger rim mouths than the long neck
jars. The size range of the the long necked jars (EPN, ERN, SPN and SRN) could not
be recorded because the diameter of other sherds within thie vessel form could not be
determined.
Decorations
The analysis for the decoration indicates that the multiple decorations were used
within fabric 2. The Short Neck Recurved (SNR) jars demonstrate that incisions and
red slip are the main decorations used. While the Pronounced Short Neck Recurved
indicated that grooves are the decoration motifs. The long neck jars such as the
Everted Recurved Neck illustrate that punctates forms part of the decorations; this
Vessel form Rim range Average Size
Short Neck Recurved (SNR) 16-23cm 17cm
Pronounced Short Neck Recurved (PSNR) 15 mm 15cm
Everted Pronounced Neck (EPN) 18cm 18cm
Everted Recurved Neck (ERN) 21 mm 19.5cm
Straight Pronounced Neck (SPN) 20 mm 10cm
Straight Recurved Neck (SRN) 20 mm 10cm
Table 5.8: A record of the vessel forms with their size range and the average size of
the rims.
54
also includes incisions, black slip, as well as burnish. The Straight Pronounced Neck
(SPN) indicates that grooves and red band of slip is used, however, no decorations
were recorded on the Straight Recurved Neck.
Lug
Table 5.10 indicates the within F2 has the abundance of lug when compared to F1.
From the table and the data indicates that the Short Neck Recurved (SNR) jars have
round lugs that are vertically perforated. However, the long neck jars indicate that
lugs play a greater role than in the short neck jars. The Everted Recurved Neck has
Vessel form P RP PS P PM PI CB I G RS RBS BB BS C S B
SNR X X
PSNR X
ERN X X X X
SPN X X
SRN
N/A (Body) X
Vessel form Square
Lug
Round
Lug
Elongated lug Vertically
perforated
Horizontally
perforated
SNR X X
PSNR
ERN X
SPN X X X X
SRN
N/A (Body) X X X X
Table 5.9: A record of the different decorations within specific categories of the vessel
forms
Table 5.10: A record of lugs occurring in fabric 2
55
elongated neck, however, this lug is not perforated rather it is decorated with incision
and punctates (see figure 5.11). The Straight Pronounced Neck(s) have square lug(s)
and the round lug(s) that are vertically perforated (see figure 5.12 and 5.13). All three
lugs (round, square and elongated) were noted on the body sherds, of which all are
vertically perforated.
Figure 5.11: An illustration of F2 rim
sherd with an elongated imperforated
lug, which has punctates and incision
as decorations.
Figure 5.12: an Illustration of a SPN
jar in F2 with a round lug that is
vertically perforated.
Figure 5.13: An illustration of an
ERN jar with a square lug that is
vertical perforated.
56
Lip forms
The lip form analysis from fabric 2 indicates that there lip forms vary from Bevelled
(B), Round (R), Square Rounded (R, S), Square and Round Overhanging (SR, O).
Their distribution within fabric 2 indicates that there was a preference of bevelled lips
as there are 6 bevelled lips, which it twice the amount of each lip form.
5.3.3 Fabric 3
Vessel Forms
Fabric 3 forms the largest part of this collection and it shows that both bowls and jars
are used. The jar type vessels indicate that both the short neck jars and long neck jars
form part of F3. Table 5.11 indicates that the short neck vessels form the largest part
of the F3 collection. Seven hemispherical (HB) were recorded, while seven rims
sherds could not be categorised into specific vessel forms (see table 5.11).
0
1
2
3
4
5
6
7
B R R,O S S,R N/A (Body)
B
R
R,O
S
S,R
N/A (Body)
Figure 5.14: Occurrences of different lip forms within fabric 2.
57
Vessel type Total number
Bowls 7
Short neck jars 48
Long neck jars 6
N/A (body sherds) 13
Unidentifiable 7
The table above was converted into the graph below with the aim of illustrating which
vessels occur more in fabric 3. From the graph it is clear that the Everted Pronounced
Neck (EPN) is not part of the collection. The Everted Recurved Neck, Collared
Incurved Neck, both long neck jars, and the Short Everted Pronounced Neck, seem to
occur at a very low level. While the Short Neck Recurved (SNR) jars highly features
in fabric 3.
0
5
10
15
20
25
30
35
40
45
SNR PSNR ESNR SEPN HB EPN SPN ERN SRN CN N/A UN-ID
SNR
PSNR
ESNR
SEPN
HB
EPN
SPN
ERN
SRN
CN
N/A
UN-ID
Table 5.11: A record of the different vessel forms within F3 and their proportion
Figure 5.15: A graph illustration of the different vessels forms and the rate they
occur within fabric 3
58
Rim diameter
∑
Where ∑ is the sum of all the data values in the sample, n is the number of the
observation in the sample. The formular above has been used to determine the mean
(avarage) rim size of the different bowls and jars within fabric 3.
Vessel Form Rim size range Average
Short Neck Recurved (SNR) 14cm-28cm 21cm
Pronounced Short Neck Recurved (PSNR) 30 cm 7cm
Everted Short Recurved Neck (ESNR) 18cm-25cm 21.5cm
Short Everted Pronounced Neck (SEPN) 24 cm 24cm
Hemispherical Bowl (HB) 49 cm 16cm
Everted Pronounced Neck (EPN) - -
Straight Pronounced Neck (SPN) 16-17cm 16.5cm
Everted Recurved Neck (ERN) 18 mm 18cm
Straight Recurved Neck (SRN) 16-28cm 22cm
Collared Incurved Neck (CN) 28 mm 28cm
Vessel form P RP PS P PM PI CB I G RS RBS BB BS C S B
SNR x x x x
PSNR x x x x x
ESNR
SEPN
HB x x
EPN
SPN x x
ERN
CN x x x
SRN
N/A (Body) x x x x x x x
Table 5.12: A record of all the different vessel forms within F3, their rim size average
and size range.
Table 5.14: A record of all the decorations on the different vessels forms within F3.
59
Lips
Within fabric 3 there seems to be a preference for square lip forms than the square
rounded. The bevelled lip form seems to be the second largest lip type preferred in
fabric 3.
0
5
10
15
20
25
30
35
B B, O S R SR SR,O O N/A
B
B, O
S
R
SR
SR,O
O
N/A
Un-
Identifiable
x x x
Figure 5.16: An illustration of a
hemispherical bowl from F3.
Figure 5.17: An illustration of a
hemispherical bowl from F3
Figure 5.18: A graph illustrating the proportion if the different lips within F3.
60
5.4 Pattern
The data indicates that all three fabrics (F1, F2 and F3) are using the short neck jars
more than the long neck jars and the bowls. Within the short neck jars, the Short Neck
Recurved (SNR) jars seem to be used quite extensively in all three fabrics. The data
also clearly indicates that there are some similarities in rim size within these fabrics.
For example, the average rim size of the Short Neck Recurved (SNR) jars in fabric 1
is the same the ones in fabric 2 (average of 17cm). While the rim size for F1 ranges
from 22cm-34cm and F2 rim size for the SNR ranges from 16cm-23cm. The SNR jars
rims have an average size of 21cm, while the sizes range from 14cm-28cm. This
indicates that the SNR jars in F1 are much larger than those in F2 and F3.
The long neck jars within fabric 1 indicate a smaller rim size compared to the short
neck jars. However, F2 does not indicate the same pattern; the long neck jars within
this fabric have wider rims than the short neck vessels. The average size of the long
neck size ranges from 19cm-20cm for all the long neck jars. While the rim average for
this category is around 20cm. Fabric 3 indicates that the long neck jars rims generally
range from 16cm-20cm and more.
The collection also indicates pattern with regards to the different decorations. The
short neck jars within F1 seem to be decorated with incision and they have burnish as
a surface treatment. While the long neck jars seem to be limited to only red slip. This
different from the F2 collection, for example, the short neck jars indicate that
incisions, grooves and red slip form part of the decoration motifs in this fabric. The
long neck jars seem to have punctates, incisions, black slip, grooves, red band of slip,
as well as burnish as surface treatment.
61
Fabric 3 in general indicates that different decoration styles are being used. The short
neck jars are decorated in punctates, punctate with slip, red slip, incisions, red band of
slip, chevron motif and burnish as surface material. The long neck jars also indicates
almost the same decorations as the short neck jar. The decorations include punctates,
round punctates, motifs, punctate with red slip, grooves, incisions, black slip, as well
as burnish as a surface treatment.
The data presented above clearly indicates that F3 is highly decorated when compared
to F1 and F2. It is also clear that punctates occur only in F2 and F3, while F1 does not
have any type of punctates as decorations. At the same time the chevron motif and
punctate with motif seem to only occur in F3 and not in any of the other fabrics.
5.5 Conclusion
The graphs, tables, pictures and illustrations we can understand that within each fabric
there are long neck and short neck vessels and at times with bowls. Thus each vessel
within fabric 1, fabric 2 and fabric 3 may have decorations that are distinct within that
fabric and vessel form. As such, the data will be further discussed and contextualised
in the interpretation chapter, with the purpose of answering research questions that
were raised in the introduction chapter.
62
Chapter 6: Interpretation
6.1 Introduction
The previous chapter has presented analytical data; it looked at any possible pattern
between the three fabrics through exploring vessel shape variations, decorations, area
of decorations, rim size average, ranges of rim sizes as well as lip forms. The
historical literature of the site narrated that the BaTawana come and colonised the
autochthonous people of the Khwebe Hills (Tlou 1980 & Dziewiecka 2002). These
narrations seem ignore the idea on culture contact and fluid interaction. As such, the
aim off this chapter is to discuss the data from the preceding chapter in details and
explore any possible implication for our understanding of contact during this time
period.
6.2 Analysis
6.2.1 Fabrics
Fabric 1
Fabric 1 that has been associated with mobile hunter-gatherer community has shown
interesting attributes. Although fabric 1 was the smallest fabric and it contributed 10%
(n=14) to the collection. The F1 collection has bowls, short neck vessels and long
neck vessels; however, this fabric indicates the abundance of short neck vessels. The
short neck jars also indicate that their rim diameter is much larger than those of the
long neck jars. The data indicates that F1 is not highly decorated, due to the fact that
only three types of decoration were recorded, the red slip, the incision and burnish.
63
Fabric 2
Fabric 2 contributes (n=22) 18% of the 132 sherds that were analysed. The analysis
indicated that this fabric only has jars in its collection, which are the short neck jars
and the long neck jars. According to the data F2 has more short neck vessels than long
neck vessels. The average rim sizes for the short neck jars is approximately 15cm-
23cm, the larger rims are within the Short Neck Recurved (SNR) jars. The size range
of the rims within the long neck jars is 18cm-20cm, with an average size ranging from
10cm-19cm.
The data indicates that the short neck jars with decoration were possibly limited to
incisions, grooves and red slip. While the long neck jars were decorated in punctates,
incision, grooves, red band of slip, black slip, as well as burnishing as surface
treatment. The analysis also indicates the possibilities that the short neck jars did not
have the abundance of lugs like the long neck jars. Only one short neck jar (SNR)
indicated a round lug that was vertically perforated. The long jars seem to have lugs
that are different. Three types of lugs were recorded; this includes the square lug, the
round lug and the elongated lug. The round and the square lugs indicate variation of
vertically and horizontally perforated, while the elongated lugs seem not to
perforation as a feature, but rather punctates and groves.
Fabric 3
Fabric 3 forms the largest part of the ceramic collection from KWH4. The fabric has
bowls and the variation of the long neck jars and the short neck jars. The data
indicates that there is a prevalence of the short neck jars than the long neck jars and
the bowls. The bowls seem to be large with a rim size of 49cm (although some rims
were broken in a way that disabled to do rim size analysis). The rims sizes for short
64
neck jars seem to generally range from 14cm-28cm. The data also indicated the same
pattern for the long neck jars. The rim size for the long neck jars ranges from 16cm-
28cm.
The analysis seems to suggest that F3 is more decorated than the other fabrics. The
decoration for this fabric includes punctates, round punctates, punctate with slip,
punctate with motif, incisions, grooves, red slip, red band of slip, chevrons, as well as
burnishing as surface treatment. It also seems like the long neck jars are highly
decorated than the short neck jars.
6.3 Archaeological evidence
6.3.1 Pastoralist’s ceramics typology
Diskin and Ashley (in prep) have associated F2 to the pastoralists in the Khwebe
Hills. As such, the typology and data of this fabric is somehow similar to the LUND
and LINC that Sadr & Sampson (1999) have identified in the Seacow valley in South
Africa. The jars from F1 are mostly undecorated and they have lugs rather than spouts
and incisions and seem to be used as a decoration motif. Although the time period and
space are different, KWH4 collection indicates the almost same characteristics of the
Seacow valley typology.
The research in Toteng by Huffman (1994) on Khoekhoe pastoralists illustrate a clear
pattern and link with the KWH4 typology. The Toteng typology comprised of short
neck jars that had red-buff chevron, they were bagged shaped with pierced lugs. They
had thickened jar rims that were decorated with crosshatching (Huffman 1994: 4).
According to Huffman (1994:4-5) Khoekhoe pastoralists ceramics that are more
historic seem to have short necks. This KWH4 F2 ceramics indicate this same pattern,
65
whereby it the number of the short neck jars is higher than those of the long neck jars.
The Toteng typology also reviled that bowls form part of the pastoralist’s ceramic
typology and these bowls are perforated. This evidence of bowls from Toteng does
not correlate with the F2 data, as no bowls were recorded within this fabric, however,
this does not mean that that bowls were never made or used by the F2 makers.
However, the Toteng collection reviled that lugs were important in this collection.
Research by Schapera (cited in Rudner 1979: 11) indicated that the historic Khoekhoe
pastoralist’s ceramics had lugs, and these were rubbed ochre. The Toteng typology
also indicates variations of decorations, this includes punctates, red bands, red
chevron motifs and small applique lumps on the shoulder area (Huffman 1994:4)., and
some of these decorations were recorded within F2.
6.3.2 Tswana typology
The site of KWH4 dates to the 19th
century as such, according to the archaeological
literature it should illustrate the third phase of the Tswana ceramics, which dates to
post AD 1700. According to Hall (1998:249) this third phase present a single or a
double band of punctates (Hall 1998: 249). Huffman (2000: 12) has also documented
that the 19th century Tswana ceramics have triangles and chevrons motifs filled with
stabs and finger nail impressions. The vessel shapes within fabric 3 includes
hemispherical bowl, which Hall (1998: 250) has also recorded.
The hemispherical bowls from fabric 3 all have the same decoration, which is a red
band of slip on the rim and at times with punctates. . Hall (1998: 250-253) has
identified that the Tswana ceramic sequence post AD 1700 has recurved jars,
however, he does not make a clear distinction within these recurved jars if they are
short neck or long neck jars. This sequence has decorations with a single band
66
immediately below the rim, a zone of decorations on the neck (chevron), a constricted
pot with a wide of decoration on the body (Hall 1998: 250-252). The data indicates
that the body sherds from fabric 3 illustrate this pattern, as chevron motifs were
recorded.
The fabric that is associated with the Khoekhoe pastoralists, matches with what
Rudner (1976) has identified, as well as, what Sadr & Sampson (1999) have recorded
in the Seacow. The rim sizes of these sherds are smaller than what Sadr & Smith
(1991: 107-109) for the Kasteelberg. However, the decorations such as the punctates
seem to occur in the Western Cape Khoekhoe ceramics. Sadr & Smith (1991) and
Sadr & Sampson (1997) have identified bands of punctates as one of the decoration
for the Khoekhoe ceramics. With regards to the lugs Schapera 1933 (Rudner 1979:
11) early research indicated that the lugs were rubbed with red ochre. Some lugs from
this collection indicated that they have red slip as surface treatment.
Discussion
The oral history of the Khwebe Hills suggests that the inhabitants interacted with each
other. Dziewiecka (2008: 12) gives a narration that the arrival of the BaTawana was
invasive, it shifted the symbiotic clientship relationship, causing havoc within the
social structure of the hills (Tlou 1980: 39). Bowen (cited in Mautle 1986: 24) has
argued that this led BaTawana to dominate over Bakgalagadi and the Basarwa. This
domination led to the prohibition of Bakgalagadi to own property or keep any
proportion of their surplus produce (Mautle 1986: 24). As such, Bakgalagadi become
Makgobe (slaves) and were always referred to as Mokgalagadi wa lolwapa (Mautle
1986: 24-26). However, Mautle (1986: 24) narration seems to ignore the oral history
that Van Waarden (1998: 147) presents, which indicates that the Bakgalagadi are an
67
ancestry group of the Batswana people. Nevertheless, according to the oral history the
Bakgalagadi and the Basarwa of the Khwebe Hills were colonised by the BaTawana
when they arrived at the hills in 1805 (Tlou 1990; Dziewiecka 2008).
The data does not indicate such an interaction of colonisation, rather it seems that the
Khwebe Hills inhabitants experience culture contact and the frontier between them
was fluid. The anomalies within the fabrics indicate this idea, as the analysis indicates
that some F3 sherds are tempered with bone and according to Bollong et al (1993: 44-
45) has argued that such organic temper is associated to mobile hunter-gatherer
communities. The analysis by Diskin and Ashley (in prep) also indicates that F3 has
coarse inclusions that are characterise F2, which is associated with the Bakgalagadi
pastoralists. Even with these anomalies, the provenance of the raw material for these
fabrics seems to be different.
The oral history also indicates that the Tawana institutes and cultures were passed
onto the autochthonous people. As such, the Bakgalagadi pastoralist’s ceramics would
have lost its distinct features such as the lugs, and these lugs are vertically perforated
and the elongated lug is decorated with punctates and grooves, and red slip. This
correlates with Schapera (cited in Rudner 1979: 11) as he noted that the pastoralist’s
ceramics were rubbed with red ochre.
If the oral history is taken into consideration then the deco rations of both F1 and F2
should indicate such a relationship. The literature review also indicates that not much
is known about the hunter-gatherer ceramics particularly in the 19th
century. As such,
an interpretation cannot be given for F1. Decorations such as punctates, incisions,
grooves, red slip have been noted by Sadr and Sampson (1999) and Huffman (1994).
In addition, F2 still shows the characterisation of Short neck jars.
68
6.5 Conclusion
This chapter has given possible interpretation from the KWH4 data. This indicates
culture contact rather than colonisation. Although, the Short Neck Recurved (SNR)
jars occur frequently in all the fabrics, this does not mean that people were colonised.
The archaeological evidence indicates that the autochthonous people were not
colonised, however, that the shared information with the BaTawana. It also indicates
possible permeable boundaries between the indigenous people of Khwebe Hills and
the immigrants. It is possible that people moved in and out of boundaries that have
been set out, where certain ceramic types and temper are associated with certain
communities.
69
Chapter 7: Conclusion
This chapter will give a brief summary on the research and the project in general.
This study has illustrated the significance of ceramic analysis and research in
interaction studies. Where previously it was assumed and narrated that interaction
between hunter-gatherers, pastoralists and agriculturalists was invasive, this study has
shown that interaction are more complex than previously thought. This project has
built on the existing ceramic typologies of both pastoralists and Tswana
agriculturalists, particularly work by Diskin and Ashley (in prep). I have illustrated
that interaction between people of different ethnicity and economies can interact. This
interaction should not be approached from a colonial perspective where the
indigenous people are colonised by the core polities. The post-colonial paradigm was
also criticised for the lack of understanding that frontiers are fluid, people make the
choices of moving in and out of other frontiers. This was shown through the ceramic
data and results from KWH4, where vessel shapes, and tempering material were
considered to be agents. What is thought to be colonisation by Tlou (1980) and
Dziewiecka (2008) the archaeology illustrates a different story. In this scenario
colonisation verses culture contact. It is suggested that the interaction between the
Khwebe Hills inhabitant was culture contact, where Bakgalagadi and Basarwa choose
to adopt certain cultures (this includes the Moloko type of ceramics) from the
BaTawana. The same with the BaTawana, they adopted certain styles of making
pottery that is more associated with mobile hunter-gatherer bands than the Sotho-
Tswana ceramic sequence.
70
Future research
This research project has revealed an enormous amount of questions surrounding
previous and current investigations on interaction. As such, this project will lead to
Masters Degree that will address such questions; however, this will be done at a
regional scale in South Africa.
71
Bibliography
Alexander, J. 1984. Early frontiers in southern Africa. In Hall, M., Avery, G. Avery,
D.M, Wilson, M. L & Humphrey, A. J. B. (eds), Frontiers: southern African
archaeology today. Great Britain: Bar international series
Ashley, C. 2009. Migration, missionaries and contact: recent archaeological
research in Khwebe hills, Botswana. In (ed) Whitehouse, R, Archaeology
International institute of archaeology. University of Collage London, London:
36-41
Barnard, A. 1992. Hunters and herders of southern Africa: A comparative
ethnography of the Khoisan peoples. Cambridge: Cambridge University Press
Barnard, A. 2006. Kalahari revisionism, Vienna and the indigenous people debate.
Social Anthropology, 14: 1-16
Barnard, A. Ethnographic Analogy and Reconstruction of Early Khoekhoe Society.
South African Humanities, 10: 61-75
Bhabha, H. K. 1996. Cultures in between, in Hall, S and du Gay, P (eds). Questions of
cultural identity pp. 53-60
Bollong, C. A., Vogel, J. C., Jacobson, L., Van Der Westhuizen, W. A &
Sampson, A. 1993. Direct dating and identifying of fibre temper in pre-
contact bushman (Basarwa) pottery. Journal
Archaeological science, 20: 44-55
72
Bollong, C. A., Sampson, G. C & Smith, A. B. 1997. Khoikhoi and Bushman Pottery
in the Cape Colony: Ethno-history and Later Stone Age Ceramics of the South
African Interior. Journal of Anthropological Archaeology, 16(3): 269-299
Boeyens, J. C. A. 2003. The Late Iron Age sequence in the Marico and early Tswana
history. Southern African Archaeological Bulletin, 58 (178): 63-78
Bousman, C. B 1998. The Chronological Evidence for the Introduction of Domestic
Stock into Southern Africa. African Archaeological Review, 15 (2):
134-150
Chambanne, A. M & Monaka, K. C. 2008. Mapping Sekgalagadi in Southern Africa:
A Socio-historical and Linguistic Study. History in Africa, 35: 133-
143
Diskin, S & Ashley, S. in prep Characterisation of Archaeological Ceramics from
The Khwebe Hills of Northern Botswana.
Druc, I. C. 2001. Archaeology and Clay. Hadrian Books Ltd. John and Erica Hedges
LTD. England
Ehret, C. 2008. The early livestock-raiser of southern Africa. Southern African
Humanities, 20: 7-35
Emberling, G. 1997. Ethnicity in Complex Societies: Archaeological Perspectives.
Journal of Archaeological Research, 5(4): 295-344
Gosden, C. 2001. Postcolonial archaeology: issues of culture, identity and
knowledge. In Archaeological Theory Today (ed. I. R. Hodder).
Oxford: Blackwell, pp. 241–61.
73
Grauer, V. A. 2007. New perspectives on the Kalahari debate: a tale of two genomes.
Before farming, (2) 4: 14
Hall, M. 1984. Frontiers in Southern African Archaeology. In Hall, M., Avery, G.
Avery, D.M, Wilson, M. L & Humphrey, A. J. B. (eds), Frontiers:
southern African archaeology today. Great Britain: Bar international
series
Hall, S. 1998. A consideration of gender relation in the Late Iron Age “Sotho”
sequence of the western Highveld, South Africa in S Kent (e.d),
Gender in African Prehistory, AltaMira Press, CA, pp. 235- 258
Hall, S. 2012, Identity and Political Centralisation in the Western Regions of
Highveld, c.1770–c.1830: An Archaeological Perspective. Journal of
Southern African Studies 38(2): 301-318
Headman, T. N., Reid, L, A, Bicchieri, M. G, Bishop, C. A, Schroeder, F & Seitz, S.
1989. Hunter-gatherers and their neighbours from prehistory to the
present. Current Anthropology, 30(1): 43-66
Huffman, T. M. 1994. Toteng Pottery and the Origin of Bambata. Southern African
Archaeology, 3: 3-9
Huffman, T. M. 2002. Regionality in the Iron Age: the case of the Sotho-Tswana.
Southern African Humanities 14: 1-22
Huffman, T. N. 2007. Handbook to the Iron Age: The archaeology of pre-colonial
farming societies in southern Africa. University of KwaZulu-
Natal Press: South Africa
74
Humphrey, A. J. B. 1988 A Prehistoric Frontiers in the Northern Cape and Western
Orange Free State: Archaeological Evidence in Interaction and
Ideological Change. Kronos South African History 13: 3-13
Kent, S. 1992. The Current Forager Controversy: Real versus Ideal Views of Hunter-
Gatherer Man, New Series 27 (1): 45-70
Kopytoff, I. 1987. The African frontiers: the reproduction of traditional African
societies. USA: Indiana University Press
Lightfoot, K. G. & Martinez, A. 1995. Frontiers and Boundaries in Archaeological
Perspective. Annual Review of Anthropology, 24: 471-492
Marks, S. 1972. Khoisan resistance to the Dutch in the seventeenth and eighteenth
centuries. Journal of African History, 13: 55-80.
Mason, R. J. 1986. Origin of Black people of Johannesburg and the southern western
Transvaal, AD350-1880. Johannesburg: occasional paper.
University of Witwatersrand.
Mautle, G. 1986. Bakgalagadi-Bakwena relationship: a case study of slavery, c1840-
c1920. Botswana Notes and Records, 18: 19-31
Naum, M. 2010. Re-emerging Frontiers: Postcolonial Theory and Historical
Archaeology of the Borderlands. Journal of Archaeological Method
and Theory, 17:101–131
Phillipson, W. D. 2005. African archaeology. Third edition. United Kingdom
University Press: Cambridge
75
Potten, D. H. 1976. Aspects of the recent history of Ngamiland. Botswana Notes and
Records 8: 63-68
Reid, A., Sadr, K & Hanson-James, N. 1998. Herding traditions. In Lane. P, Reid, A
& Segobye, A. (eds) Ditswa Mmung: the archaeology of Botswana. Pula
press and the Botswana Society: Botswana
Reid, A. Lane, P. Segobye, A., Borjeson, L., Mathibidi, N & Sekgarametso, P. 1997
. Tswana Architecture and Responses to Colonialism. World
Archaeology, 28(3): 370-392
Rockman, M. 2003. Knowledge and Learning in the Archaeology of colonisation. In
Rockman, M & Steele, J (eds): Colonization of Unfamiliar
Landscape: The Archaeology of Adaptation. Routledge:
London
Rudner, J. 1979. The Use of Stone Artefacts and Pottery among the Khoisan Peoples
in Historic and Protohistoric. The South African Archaeological
Bulletin, 34 (129): 3-17
Sadr, K. 1997. Kalahari Archaeology and the Bushman Debate. Current
Anthropology, 38(1): 104-112
Sadr, K. 2002. Encapsulated Bushmen in the archaeology of Thamaga. In: S. Kent
(e.d), Ethnicity, hunter gatherers, and the “Other”. Washington &
London: Smithsonian Institution Press, pp. 29–47
Sadr, K. 2008. An ageless view of the first millennium AD southern African
ceramics. Journal of African Archaeology, 6(1): 103-129
76
Sadr, K. 2003. The Neolithic of southern Africa. The Journal of African History,
44(2): 195-209
Sadr, K. 2008. An ageless view of the first millennium AD southern African
ceramics. Journal of African Archaeology, 6(1): 103-129
Sadr, K. 2008. Invisible herders? The archaeology of Khoekhoe pastoralists. Southern
African Humanities, 20: 179-203
Sadr, K & Sampson, C. G. 1999. Khoekhoe Ceramics of the Upper Seacow River
Valley. The South African Archaeological Bulletin, 54(169):3-15
Sadr, K & Smith, B. 1991. On Ceramic Variation in the South-Western Cape, South
Africa. The South African Archaeological Bulletin, 46(154): 107-114
Stein, J. G. From Passive Periphery to Active Agents: Emerging Perspectives in the
Archaeology of Interregional Interaction: Archaeology Division
Distinguished Lecture AAA Annual Meeting. American
Anthropologist, New Series, 104 (3) pp. 903-916
Schrire, C. 1980. An Inquiry into the Evolutionary Status and Apparent Identity of
San Hunter-Gatherers. Human Ecology, Vol. 8, No. 1 (Mar., 1980),
pp. 9-32
Smith, A. B. 1998. Keeping people on the periphery: the ideology of social
hierarchies between hunter-gatherers and herders. Journal of
Anthropological Archaeology 17: 201-215
Smith, A. B. 2001. Ethno-history and the Archaeology of the Ju/’Hoansi Bushman.
African study of monographs, 26: 15-25
77
Thebe, C. P. 1996. An Archaeological Study of the Pottery from Ntsweng, Molepolole.
Unpublished: Honours dissertation
Tlou, T and Campbell, C. A. 1997. History of Botswana. Macmillan Botswana
publishing company, Gaborone, 1997.
Tlou, T. 1980. A history of Ngamiland~1750 to 1906: The formation of an African
state. Macmillan Botswana publishing Company: Botswana.
Van Dommelen, P. 2002. Ambiguous Matters: Colonialism and the Local Identity in
Punic Sardinia. In Lyons, C. L & Papadopoulos, J. K (eds):
The Archaeology of Colonialism. Getty Research Institute:
USA
Van Dommelen, P. 2005. Colonial Interaction and hybrid Practices. In Stein, P (ed):
The Archaeology of Colonial Encounters. School of American
Research Press: Santa Fe
Van Dommelen, P. 2006. Colonial Matters: Material Culture and the Post-colonial
Theory in Colonial Structures. In Tilley, C., Keane. W., Kuchler, S.
Rowlands, M. Spyer, P. (eds): Handbook of material culture. SAGE
Publication Ltd: London
Van Waarden, C. 1998. The Later Iron Age. In Lane. P, Reid, A & Segobye, A. (eds)
Ditswa Mmung: the archaeology of Botswana. Pula press and
the Botswana Society: Botswana
Wilmsen, E. N. 2003. Further Lessons in Kalahari Ethnography and History. History
in Africa, 30: 327-420