UNIVERSITY OF RWANDA

COLLEGE OF ARTS AND SOCIAL SCIENCES

DEPARTMENT OF HISTORY AND HERITAGE STUDIES
ACADEMIC YEAR: 2021-2022

Course: ARCHAEOLOGY OF RWANDA

Lecturer: Mr. André NTAGWABIRA

January 2022
 i

TABLE OF CONTENTS

Table of Contents ......................................................................................... i

PART I: GENERALITIES ............................................................................... 1

CHAPTer 1. TECHNIQUES OF ARCHAEOLOGY ........................................... 1

1.1. In the field ......................................................................................... 1
1.1.1. Finding archaeological sites ......................................................... 1
1.1.2. Excavation .................................................................................. 5
1.2. After the field .................................................................................. 10
1.2.1. Sorting and conservation of objects ........................................... 10
1.2.2. Classification ............................................................................ 11
1.2.3. Detailed typological analysis and technological studies .............. 11
1.2.4. Ordering and comparison .......................................................... 12
1.2.5. Dating methods ......................................................................... 12
1.2.6. Interpretation and publication ................................................... 17
CHAPTer 2. SCIENCES RELATED TO ARCHAEOLOGY .............................. 17
2.1. Ethnography ................................................................................... 17
2.2. Social anthropology ......................................................................... 18
2.3. Palynology ....................................................................................... 18
2.4. Sedimentology ................................................................................. 18
2.5. Stratigraphy .................................................................................... 19
2.6. Geology ........................................................................................... 19
2.7. Linguistics ...................................................................................... 19
2.8. Zoology............................................................................................ 19
2.9. Physics and chemistry ..................................................................... 19
2.10. Statistics ....................................................................................... 19
CHAPTer 3. PREHISTORIC PERIODS ......................................................... 20
3.1. Classical subdivisions (criterion: writing) ...................................... 20
3.2. Technological subdivisions ........................................................... 20

PART II. TECHNOLOGICAL EVOLUTION AND MODES OF LIFE IN

PREHISTORIC RWANDA ............................................................................ 22

CHAPTer 4. STONE AGE INDUSTRIES ...................................................... 22

 ii

4.1. Early Stone Age ............................................................................... 22

4.1.1. Oldowan.................................................................................... 22
4.1.2. Acheulean ................................................................................. 23
4.2. Middle Stone Age ............................................................................. 23
4.2.1. In the forest .............................................................................. 24
4.2.2. In savanna regions .................................................................... 24
4.3. Late Stone Age ................................................................................ 25
4.3.1. In the forest .............................................................................. 25
4.3.2. In savanna areas ....................................................................... 25
4.4. Prehistoric periods in Rwanda ......................................................... 25
4.4.1. Early Stone Age ......................................................................... 25
4.4.2. Middle Stone Age ....................................................................... 25
4.4.3. Late Stone Age .......................................................................... 26
4.4.4. Neolithic .................................................................................... 27
CHAPTer 5. IRON AGE .............................................................................. 27
5.1. Early Iron Age ................................................................................. 28
Early Iron Age Agriculture ................................................................... 29
5.2. Middle Iron Age? ............................................................................. 31
5.3. Late Iron Age ................................................................................... 32
Environmental consequences of iron production ................................. 32
CHAPTer 6. EMERGENCE OF GREAT LAKES KINGDOMS ......................... 35
6. 1. Consolidation and expansion of Great Lakes kingdoms ................... 38
6.1.1. Consolidation of the kingship institution ................................... 38
6.1.2. Wars of expansion: Case of Rwanda .......................................... 40
6.2. Subsistence economy ...................................................................... 41
References ................................................................................................. 43
 1

PART I: GENERALITIES

CHAPTER 1. TECHNIQUES OF ARCHAEOLOGY

1.1. In the field

How do archaeologists know where to dig? How do they discover and assess
archaeological sites? How do they dig them? These are main questions which
must be answered for better understanding archaeological fieldwork.

1.1.1. Finding archaeological sites

Finding archaeological sites concerns the process of locating sites, which is

either by accident or the result of deliberate archaeological survey.

1.1.1.1. Accidental discovery

Many archaeological sites come to light (are known) by accident as a result of

human activity, or as a result of natural phenomena. In the first case, they
are discovered for example during construction works (roads, building, dams,
etc.) and mining. On the other hand, natural agents that can expose
archaeological sites for us include earthquakes, erosion, wind erosion,
burrowing animals, etc.

1.1.1.2. Archaeological survey

Archaeological survey means walking within specific areas for locating and
recording archaeological sites. There are various types of survey depending on
the methods used.
- Field walking
Field walking is a survey carried out on foot, during which archaeologists
locate archaeological sites. Actually, there are numerous
inconspicuous/conspicuous signs that can indicate the site. These are for
example the gray organic soil eroding from a long-abandoned settlement,
subtle colors of rich vegetation that reveal long-buried houses, humanly
fractured stones indicating a Stone Age camp, potsherds, slags or other iron
smelting furnace remains that indicate iron working places, etc. in terms of
 2

time, field walking survey can vary from spending a couple of hours walking
around a small place for searching traces of historical structures to doing a
large-scale survey over several weeks, months or years of an entire river basin.
During survey, information on possible sites can even be provided by
knowledgeable local inhabitants, especially landowners (Fagan, 2006: 130).
“Today’s archaeologists relay on technology and elaborate instrumentation to
help them discover the past, especially on CRM projects. This is
nondestructive archaeology – the analysis of archaeological phenomena
without excavations or collecting of artifacts, both of which destroy the
archaeological record. The major methods in this approach are generally
labeled remote sensing. These techniques include aerial photography,
various magnetic prospecting methods, and sidescan radar” (Fagan, 2006:
134).
- Aerial photography
Actually, aerial photography gives an overhead view of the past. Sites can be
photographed from many directions, at different times of day, and at various
seasons. There are numerous sites that left almost no surface traces but which
can be discovered by analyzing aerial photographs. For example, even though
many earthworks and other complex structures can be exposed by plows or
erosion, their original layout shows up clearly from the air. The rising or
setting sun can make large shadows, emphasizing the relief of almost vanished
banks or ditches; the features of the site stand out in oblique light. Such
phenomena are sometimes called shadow sites (Fagan, 2006: 134).
In some areas, it is possible to detect differences in soil color and in the
richness of crop growth on a particular soil. Such marks are hard to detect on
the surface but often show up clearly from the air. The growth and color of a
crop are greatly determined by the amount of moisture the plant can derive
from the soil and subsoil. If the soil depth has been increased by digging
features such as pits and ditches, later filled in, or because additional earth
has been heaped up to form artificial banks or mounds, the crops growing
over such abandoned structures are high and well nourished. The opposite is
also true, where soil has been removed and the infertile soil is near the surface,
or where impenetrable surfaces such as paved streets are below ground level
 3

and the crops grow less thickly. Thus a dark crop mark can be taken for a
ditch or pit, and a lighter line may define a more substantial structure (Fagan,
2006: 135).
- Toponymy
Toponymy concerns the study of place names. It can guide archaeologists
when they are looking for archaeological site. This is because the names of
places can sometimes connote activities that took place there in the past. For
example, the name “Butare” in Kinyarwanda means “the place of iron ore” and
often indicates the regions where iron smelting activities took place or where
iron ore was found. Another example is “Mucuzi” which means the area
inhabited by blacksmiths.
- Geophysical methods of site survey
Geophysical methods of site survey are modern methods used for detecting
subterranean features. In fact, all archaeologists dream of a new and
revolutionary method that will enable them to find out what is underground
without the labor of excavation. This is why, even though only few researchers
have access to modern tools, the use of nonintrusive detection methods is
increasing in Archaeology. Nonintrusive methods include the use of metal
detectors, which have a bad name in the hands of treasure hunters but are
an effective way of finding artifacts to depths of about 20 centimeters (8 inches)
in properly supervised hands. Some of the Geophysical methods of site survey
are:
 Resistivity survey

Resistivity survey is particularly useful for locating structural remains

associated with building footings. In addition, this technique can locate former
ditches and pits. Surveys are possible over grass, crops and open soil. The
electrical resistance or resistivity of the soil depends upon the moisture
content and distribution within the soil. Buried features such as walls can
affect the moisture distribution and are usually more moisture resistant than
other features such as the infill of a ditch. A stone wall will generally give a
high resistance response and the moisture retentive content of a ditch can
give a low resistance response. See http://www.archaeological-
surveys.co.uk/subpage2/index.html for the image of resistivity survey.
 4

 Electromagnetic survey
Electromagnetic survey is the opposite of resistivity, measuring the
conductivity of features like walls, which is low, while the resistivity of the
same structure is high. This method is faster, does not require electronic
probes1, and can be used to locate structures over large areas.
 Subsurface (ground-penetrating) radar

Ground penetrating radar survey can prove effective at locating archaeological

features underneath a wide range of surfaces that cannot be surveyed by any
other geophysical technique. Good results can be obtained underneath
concrete, stone, tarmac and through grass and soil. In addition, an indication
of depth can be provided which may prove highly useful where intrusive work
is required. Gpr survey records reflected radio waves that are pulsed into the
ground as the radar antenna is dragged along the ground surface. Changes in
the ground make-up, such as from soil to stone, provide conditions where
some of the transmitted energy is returned to the surface. In this way,
subsurface features can be mapped.

Careful timing of the returned radio waves along with calculations of the
ground's dielectric constant can allow an estimate of depth to be made.
However, it is worth noting that rarely can Ground penetrating radar be used
to locate graves. Exceptions exist, particularly if disturbance is recent or if
there are buried voids or associated structural features. Visit
http://www.archaeological-surveys.co.uk/ground-penetrating-
radar/index.html to see with images how Ground penetrating radar survey is
done.
 Magnetometer surveys
Detailed magnetic survey known as magnetometry is an effective and efficient
geophysical survey technique used to define areas of past human activity by
mapping spatial variations and contrast in the magnetic properties of soil,
subsoil and bedrock. A Magnetometer survey uses the gradiometer2 which
provides magnetic readings as the surveyor walks along the grid lines across

1
. probe: a device that is put inside something to test or record information
2
. gradiometer: an instrument for measuring the gradient of a magnetic field
 5

a site. The readings are downloaded at the end of the session and provide a
contour map of subsurface anomalies that may represent the results of
human activity. The technique can be carried out over areas of grass, crop and
open soil. Visit http://www.archaeological-
surveys.co.uk/subpage1/index.html to see an example of magnetic survey.

Very often, surveyors combine several geophysical methods which actually is

an effective and economical way to locate subsurface features, and then
archaeological sites (Fagan, 2006: 140-142).

1.1.2. Excavation

In Archaeology, excavation is a meticulous process of digging and recording

both finds and their context in time and space (Fagan, 2006: 168). There are
some basic principles of excavation that archaeologists generally apply in the
field. However, every site has its own particularities, can present distinctive
challenges and lead to the modification of basic principles.

1.1.2.1. Types of excavation

Excavation occurs in order to preserve and record endangered sites or to

improve the knowledge of certain periods of human history. There are different
methods of digging and the way a site is dug depends on what the
archaeologists want to learn from the site and the nature of the site itself.
- Sampling
Sampling is “the science of controlling and measuring the reliability of
information through the theory of probability” (Fagan, 2006: 148). Sampling
techniques allow archaeologists to ensure a statistically reliable basis of
archaeological data from which they can make generalizations about their
research data. For example, after survey, one can proceed by sampling
excavation if the site contains thousands of artifacts among which only a small
sample can be dug and analyzed.
To ensure validity of the statistic samples, some form of unbiased sampling
are used to choose which part of a site is to be dug. Sampling methods include:
 6

1° Probabilistic sampling: a means of relating small samples of data in

mathematic ways to much larger populations. The classic example of this
technique, commonly used in the disciplines of statistics and statistical
theory, is the political opinion poll3, testing national feelings from tiny
samples. In archaeology, probabilistic sampling shows that the conclusions
reached from a survey or excavation on the basis of the samples are relatively
reliable.
2° Random sampling: is used when nothing is known ahead of excavation
and when the archaeologist wishes to obtain an unbiased sample of artifacts
from a site. One can arrive at this result by laying out a rectangular grid of
squares on a site and then randomly select the squares to be dug.
3°Stratified sampling: is used when the investigator uses previous knowledge
of an area, such as its topographic variation, to structure further research. In
this case, one can sample some selected units intensively, and others less
carefully (Fagan, 2006: 148-149).
- Site testing
Site testing is a stand-alone4 way of deciding whether the site is important for
further investigation or to establish its date, function, or type of occupation.
In these days of subsurface radar technology and other sophisticated
geomorphological studies, site testing has become more sophisticated than it
was in the past. However, alongside those modern geophysical methods, there
are classic ways for testing sites including:
1° Test pit: it remains the most useful way of obtaining preliminary
information on stratigraphy and culture history in advance of large-scale
excavation. Some test trenches are small control pits, dug carefully as a way
of anticipating subsurface stratigraphy and occupation layers. Such
excavations are reference points for planning an entire dig. More often, test
pits are laid out in lines and over considerable distances to establish the extent
of a site and the basic stratigraphy in different areas.
2° Shovel pits: they are a variation on the test pit theme, usually used in
surface survey to trace occupation deposits. They are little more than small

3
. Poll: a study in which people are asked for their opinions about a subject or person
4
. Stand-alone: capable of operating independently of any other device or system.
 7

holes dug with a shovel a few centimeters below the surface and are much
used to establish the boundaries of shallow settlement sites and features
(Fagan, 2006: 149).
- Vertical excavation
Vertical excavation concerns digging limited areas for specific information
such as on dating and stratigraphy. In fact, the layout of small digs is
determined not only by surface features, density of surface finds, or sampling
techniques, but by available funds as well. Most excavations are run on a
shoestring5, so small-scale operations have to be used to solve complex
problems with minimal expenditure of time and money. For example, vertical
trenches can be used to obtain artifact samples, to establish sequences of
ancient building construction or histories of complex earthworks6, and to
salvage sites threatened with destruction. The small trenches are often dug in
areas where the deposits are likely to be of maximal thickness or where
important structures may be found. Much vertical excavation consists of long
cross sections cut across mounds, buildings, or other structures designed to
establish their chronology and architectural sequence. Vertical excavation is
very successful in small sites such as caves and rock-shelters, where space is
limited and the excavators have to deal with various obstructions (Fagan,
2006: 150).
- Area / or horizontal excavation
Area or horizontal excavations are large-scale excavations normally used to
uncover wider areas of a site. In other words, area excavations expose large,
open areas of ground to a depth of several centimeters. For example, a complex
network of walls, houses, or abandoned storage pits may lie within the site.
Each of these features relates to other structures, a relationship that must be
carefully recorded for better interpreting the site. Since horizontal excavation
exposes large parts of a site, the excavator is confronted with the problem of
maintaining stratigraphic control from one side of the trench to the other. That
is why many area excavations are organized on grid systems, which allow walls
(often called balks) to be left between adjoining squares.

5
. shoestring: If you do something on a shoestring, you do it with a very small amount of money
6
. Earthwork: a bank of earth made, especially in the past, for defense against enemy attack
 8

Large open area excavations require meticulous recording over considerable

distances, made much easier when the position of features and finds can be
recorded with a total data station7 which records data that can be downloaded
into laptop computers at the end of the day’s work. Horizontal excavations are
expensive even if earthmoving machinery (using bulldozers, etc.) is used to
remove sterile layers, but they provide a unique horizontal view of human
occupation or of entire human settlements obtainable in no other way.
In these days where CRM excavations are most important, a combination of
excavation methods is used to acquire and record the maximum amount of
information in the most cost-effective manner possible while maintaining high
scientific standards (Fagan, 2006: 152-153).

1.1.2.2. Digging, tools, people

Digging: The way archaeologists dig depends on the type of site they are
excavating. For example, a huge burial mound on the Ohio River may be more
than 6 meters deep; but another burial mound at Bjorvollen terrace (Fana-
Bergen –Hordaland/Norway) may be 1.5 meter deep. Or the sterile layer at a
fugitive8 slave community habitation site of Kenya Coastal Province may be
reached at 25 centimeters; whilst the depth of an Early Iron Age furnace pit
at Nyaruhengeri (Gisagara District) may be 75 centimeters. Excavation is in
part a recording process, and accuracy is essential. Therefore, the dig must
be kept tidy at all times in order to produce precise records. The trench walls
must be always straight, so one can record the layers he/she is digging and
follow them across the site. When necessary, soil is sieved in order to find tiny
objects such as beads, fish bones, etc. On CRM excavations where
earthmoving machinery is now widely used to save time and money, its use
must be careful to minimize destruction. As soon as the archaeologists reach
layers in which finds are expected, they dig with meticulous care, removing
each layer in turn, recording the exact position of their finds upon discovery.
In few words, “the excavation is a laboratory and should be treated as such”
(Fagan, 2006: 155).

7
. Total data station: an electronic distance-measuring device with recording computer
8
. Fugitive: a person who is running away or hiding from the police or a dangerous situation
 9

Tools: archaeologist at work uses many tools depending on the specificity of

the site, layers and finds. These tools include trowels, shovels, buckets,
brushes, compass, twisted nylon line, kneeling mats, pencils, pens, ruler,
notebook, graph papers, hand and long tapes, waterproof clipboard, portable
GPS unit, dumping level and graduated stick, etc.
People: an archaeological excavation is headed by a director, who is
responsible both for organizing the excavation and for overseeing the
specialists (if any) and diggers. In fact, it is better if a team of experts (such as
palynologist, soil scientist, timber expert, experts on ancient metallurgy and
mammal bones) is available for working alongside the excavators and interpret
different phenomena. At their turn, diggers must be well trained for their job.

1.1.2.3. Recording

No dig is more important than its records. The excavation notebooks provide
a day-to-day record of each trench, of new layers and significant finds. Before
any trench is measured out, the entire site must be laid out on a grid of
squares. Important finds, or details of features such as a house or a storage
pit, are measured on the site plan by simple three-dimensional measuring
techniques or with an electronic recording instrument. It is the information
from the records, as well as the artifacts from the dig, that form the priceless
archive of the excavation. Therefore, if the records are incomplete, the
excavation is little better that a treasure hunt (Fagan, 2006: 156).

1.1.2.4. Stratigraphic observation

Careful stratigraphic observation is the cornerstone of the excavation, for it

enables archaeologists to see the changes occurring at the site through time
by comparing strata. In fact, archaeology lays on the laws of superposition
and association which help to know the context of archaeological finds in time
and space. The layers of archaeological sites, be they natural or humanly
created, form much quicker than geological levels, but they are subject to the
same law of superposition. This law of superposition assumes that the deeper
the layer is, the older the time period it relates to. Thus the excavated
stratigraphic profiles at an archaeological site represent a sequence of layers
 10

that have formed through time. Stratigraphic observation is the process of

recording, studying and evaluating stratified layers in archaeological sites,
layers that were deposited horizontally, but are studied in the vertical (time)
dimension.
However, stratigraphic observation involves not only recording the layers, but
confirming that they represent a sequence in time. This is because several
factors can disturb stratified layers. For example, animals (rabbit, ants, etc.)
can burrow through soft earth, or latter occupants of a house may dig into
underlying layers to construct storage pits, build foundations, or even bury
the dead. This is where the law of association comes in, for the artifacts
associated with stratified, undisturbed archaeological layers can then be
placed in a relative chronology, and if radiocarbon samples are dated from one
or more layers, perhaps in an absolute one as well (Fagan, 2006: 156).

1.2. After the field

When they are back from the field, archaeologists proceed with analysis and
classification of their finds, although these works basically start in the field.

1.2.1. Sorting and conservation of objects

The elaborate process of analysis and classification of finds of all kinds starts
in the field alongside excavation – processing and organizing the finds so that
they can be analyzed. These first stages in processing newly excavated
archaeological finds are entirely routine. Most large excavations maintain
some form of field laboratory. It is here that the major site records are
maintained, stratigraphic profile drawings are kept up to date, and
radiocarbon samples and other special finds are packed form examination by
specialists. A small team must be in charge of field laboratory to ensure that
all finds are cleaned, processed on time, packed carefully, labeled and
recorded precisely. A successful laboratory operation allows the director of the
excavation to evaluate the available data daily, even hourly. It is here, too, that
basic conservation work is carried out: e.g.: reassembling fragmented pots,
hardening bones with chemicals, or stabilizing fragile objects. Computers play
 11

an important role in the field laboratory, for they are used to code vast
quantities of information for later use.
The analysis continues back in the home laboratory, where one needs a good
eye for detail, an orderly mind, and, above all, infinite patience for it takes
weeks to sort and classify even a relatively small artifact collection (Fagan,
2006: 170-171).

1.2.2. Classification

In archaeology, classification is a process of dividing artifacts and other data

into discrete types (Fagan 2006: 172). Classification is done on the basis of
the artifact attributes. Such operation is not an end in itself but a technique
by which an archaeologist can attain specified objectives (Rouse 1960: 313).
Fagan (2006: 172-173) believes that by classifying artifacts, the purpose is to
organize data into manageable units; describing the types; identifying the
relationships between those types; and finally to study the assemblage
variability in the archaeological record. In other words, the classification of
artifacts is important for comparative purposes.

1.2.3. Detailed typological analysis and technological studies

The types of artifacts are established on the basis of attributes defined as

logically irreducible characters of two or more states, acting as independent
variables within a specific artifact system (Clarke 1968:187). These attributes
are believed to be the manifestations of ideas held in common by makers and
users of the artifacts (Reid 2007:23). They reveal information on the behaviors
of the people who lived at an archaeological site.
However, the identification of attributes distinctive of types challenges
because, depending on his objectives, archaeologist can himself choose the
criteria which are to be considered as attributes of his classes (Haaland 1972,
Rouse 1960). But many scholars argue that there are some main aspects that
may be such as technology, function or the [symbolic] use of the artifact that
the attributes should reflect (Haaland 1972:96). These criteria are especially
distinctive when the materials to be classified are related to iron working. In
this way, for example, specific techniques adopted by smelters can be used as
 12

criteria for differentiating various types of iron furnaces and understanding

the technology used at each single site.

1.2.4. Ordering and comparison

After establishing different types of artifacts, it is easy to describe (them) and

identify the relationship between them. In this way, one will compare artifacts
from different contexts at the same site, or compare artifacts from two different
sites.

1.2.5. Dating methods

Dating archaeological material can be made by a direct study of an artifact, or

may be deduced by association with materials found in the context the item
is drawn from. Dating is normally a post-excavation work. However, for good
practice, some preliminary dating work starts in the field alongside excavation.
Dating is very important in archaeology for it helps constructing models of the
past. There are two types of dating methods: relative and absolute methods.

1.2.5.1. Relative methods

The first, and in some ways the most important, step in much archaeological
research involves ordering things into sequences. The things to be put into
sequence can be archaeological deposits in a stratigraphic excavation, or they
can be artifacts as in a typological sequence. All these sequences can be used
for relative dating.
Stratigraphy: Stratigraphy, as seen in chapter 2, is the study of stratification
– the laying down or depositing of strata or layers (also called deposits) one
above the other. From the point of view of relative dating, the important
principle is that the underlying layer was deposited first and therefore earlier
than the overlaying layer. Thus a succession of layers should provide a relative
chronological sequence, from earliest (bottom) to last (top).
Good stratigraphic excavation at an archaeological site is designed to obtain
such a sequence. Part of this work involves detecting whether there has been
any human or natural disturbance of the layers since they were originally
deposited. The disturbances can be for example rubbish pits dug down by
later occupants of a site into earlier layers, animal burrowing holes, and floods
 13

washing layers away and re-depositing them elsewhere in a secondary context.

Armed with carefully observed stratigraphic information, the archaeologist
can hope to construct a reliable relative chronological sequence for the
deposition of the different layers.
But of course what archaeologists mostly want to date are not so much the
layers or deposits themselves as the humanly generated materials within them
– artifacts, structures, organic remains – which ultimately reveal past human
activities at the site. Here the idea of association is important.
When it is said that two objects were found in the association within the same
archaeological deposit, it generally means that they became buried at the same
time. Provided that the deposit has never been disturbed and does not contain
objects from another deposit, the associated objects can be said to be no later
(no more recent) than the deposit itself. A sequence of sealed deposits thus
gives a sequence – and relative chronology – for the time of burial of the objects
found associated in those deposits. If one of those objects can later be given
an absolute date – say a piece of charcoal that can be dated by radiocarbon in
laboratory – then it is possible to assign that absolute date not only to the
charcoal but to the sealed deposit and the other objects associated with it as
well. A series of such dates from different deposits will give an absolute
chronology for the whole sequence. It is this interconnecting of stratigraphic
sequences with absolute dating methods that provides the most reliable basis
for dating archaeological sites and their contents (Renfrew and Bahn,
2000:118-119).

Typological sequences: when we look at artifacts, buildings and indeed any

of the human creations around us, most of us can mentally arrange some of
them into a rough chronological sequence. One kind of aircraft looks older
than another, one set of clothes looks more old-fashioned than the next. How
do archaeologists exploit this ability for relative dating?
The form of an artifact such as a pot can be defined by its specific attributes
of material, shape, and decoration. Several pots with the same attribute
constitute a pot type, and typology groups artifacts into such types.
Underlying the notion of relative dating through typology are two other ideas:
 14

The first is that the products of a given period and place have a recognizable
style: through their distinctive shape, and decoration they are in some sense
characteristic of the society that produced them. The archaeologist can often
recognize and classify individual artifacts by their style, and hence assign
them to a particular place in a typological sequence.
The second idea is that the change in style (shape and decoration) of artifacts
is often gradual, or evolutionary. It means that particular artifacts produced
at about the same time are often alike; whereas those produced several
centuries apart will be different as a result of centuries of change.
Consequently, it appears that when artifacts are properly grouped, it is
possible to assign a relative date to each of them by matching it with an artifact
already recognized within a well-established typological system (Renfrew and
Bahn, 2000:120-121).

1.2.5.2. Absolute methods

Despite the great utility of methods of relative dating, archaeologists finally

want to know how old sequences, sites and artifacts are in calendar years. To
achieve this, they need to use the methods of absolute dating some of which
are described in the following sections.
a. Radiocarbon dating
It is a radiometric dating 9 method that uses the naturally occurring
radioisotope carbon-14 (C14) to determine the age of carbonaceous materials
actually up to about 60 000 years. One of the most frequent uses of
radiocarbon dating is to estimate the age of organic10 remains from
archaeological sites. When plants fix atmospheric carbon dioxide (CO2) into
organic material during photosynthesis they incorporate a quantity of C14
that approximately matches the level of this isotope in the atmosphere (a small
difference occurs because of isotope fractionation, but this is corrected after
laboratory analysis). After plants die or they are consumed by other organisms
(for example, by humans or other animals), the C14 fraction of this organic
material declines at a fixed exponential rate due to the radioactive decay of

9
. Radiometric dating: any method of dating material based on the decay of its constituent radioactive atoms, such
as potassium-argon dating or rubidium-strontium dating Also called radioactive dating.
10
. Organic: being or coming from living plants and animals
 15

C14. Comparing the remaining C14 fraction of a sample to that expected from
atmospheric C14 allows the age of the sample to be estimated.
b. Dendrochronology or tree-ring dating
Dendrochronology or tree-ring dating is the scientific method of dating based
on the analysis of patterns of tree-rings. Dendrochronology can date the time
at which tree-rings were formed, in many types of wood, to the exact calendar
year. In some areas of the world, it is possible to date wood back a few
thousand years, or even many thousands. In most areas, however, wood can
only be dated back several hundred years, if at all.
Many trees in temperate zones make one growth ring each year, with the
newest adjacent to the back. For the entire period of a tree’s life, a year-by-
year record or ring pattern is formed that reflects the climate conditions in
which the tree grew. Adequate moisture11 and long growing season result in a
wide ring. A drought year may result in a very narrow one. Alternating poor
and favorable conditions, such as mid-summer droughts, can result in several
rings forming in a given year. For example, missing ring are rare in oak12 and
elm13 trees – the only recorded instance of a missing ring in oak trees occurred
in the year 1816, also known as the Year Without a Summer. Trees from the
same region will tend to develop the same patterns in ring widths for a given
period. These patterns can be compared and matched ring for ring with trees
growing in the same geographical zone and under similar climatic conditions.
Following these tree-ring patterns from living trees back through time,
chronologies can be built up, both for entire regions, and for sub-regions of
the world. Thus wood from ancient structures can be matched to known
chronologies (a technique called cross-dating) and the age of the wood
determined precisely. Cross-dating was originally done by visual inspection,
until computers were used to do statistical matching.
c. Thermo-luminescence dating
Thermoluminescence (TL) dating is the determination by means of measuring
the accumulated radiation dose of the time elapsed since material containing

11
. Moisture: a liquid such as water in the form of very small drops, either in the air, in a substance, or on a surface
12
. Oak: a large tree that is common especially in northern countries, or the hard wood of this tree
13
. Elm: a large tree which loses its leaves in winter, or the wood from this tree
 16

crystalline minerals was either heated (lava, ceramics) or exposed to sunlight

(sediments). As the material is heated during measurements, a weak light
signal, the thermoluminescence, proportional to the radiation dose is
produced.
Thermoluminescence dating is used for material where radiocarbon dating is
not available, like sediments14. Its use is now common in the authentication
of old ceramic wares, for which it gives the approximate date of the last firing.
Optical dating is a related measurement method which replaces heating with
exposure to intense light. The sample material is illuminated with a very bright
source of infrared light (for feldspars15) or green or blue light (for quartz).
Ultraviolet light emitted by the sample is detected for measurement.
d. Potassium-argon dating
Potassium-argon dating or K-Ar dating is a radiometric dating method used
in geochronology and archaeology. It is based on measurement of the product
of the radioactive decay of an isotope of potassium (K), which is a common
element found in many materials, such as micas16, clay minerals, tephra17,
and evaporates, into argon. In these materials, the decay product 40Ar is not
trapped by the rock while it is liquid (molten), but starts to accumulate when
the rock solidifies (recrystallizes). Time since recrystallization is calculated by
measuring the ratio of the amount of 40Ar accumulated to the amount of 40K
remaining. The long half-life of 40K allows the method to be used to calculate
the absolute age of samples older than a few thousand years.

To conclude this point, the answer to the question “when” in archaeology has
two main components:
- Relative dating methods which allows determining that something is
relatively older or younger than something else.
- Absolute methods which allows giving a date in years.

14
. Sediment: a soft substance that is like a wet powder and consists of very small pieces of a solid material which
have fallen to the bottom of a liquid; or material that has been deposed from water, ice or wind.
15
. Feldspar: A type of hard rock.
16
. Mica: a natural glass-like substance that breaks easily into thin layers and is not damaged by heat, often used
in electric equipment.
17
. Tephra: solid matter ejected during a volcanic eruption.
 17

However, archaeological dating is at its most reliable when the two methods
are used together. For example, when the relative order assigned to layers in
an excavation can be confirmed by absolute dates for each layer. Wherever
possible, results from one absolute method should be cross-checked by those
from another. For example, radiocarbon by tree-ring dating, etc. (Renfrew and
Bahn, 2000:170).

1.2.6. Interpretation and publication

The last stage of an archaeological research is to interpret the data (finds),

draw the conclusions and disseminate them. Depending on the type of data
recorded, it is even possible to use ethnographic material in the interpretation
of archaeological finds, but this must be done carefully.

CHAPTER 2. SCIENCES RELATED TO ARCHAEOLOGY

2.1. Ethnography

Ethnography is the branch of anthropology that deals with the scientific

description of individual human societies. Ethnographic research relevant to
archaeology concerns research focused on material culture. It consists of
observing people working with traditional objects or practicing traditional
crafts, as well as interviewing elders in the society. Members of the community
can provide information on changes in the manufacture, use and disposal of
some material culture items during their lifetime. Sometimes, they even
recount early information learned from deceased relatives (Atherton 1983: 77).
Additionally, it is believed that through ethnography much information can be
gleaned regarding technological aspects and this can give insights on human
behaviors in the past (Humphris 2010: 76). When an archaeologist observes
carefully how local craftsmen organize and perform their activities, and
combines their observations with critical analysis of information provided by
elders, it is possible to interpret the archaeological record.
 18

2.2. Social anthropology

Social anthropology is one of social sciences which studies “primitive”

societies. Social anthropologists are interested in ecology, economy, political
and judicial institutions, familial structures, parental systems, religions,
techniques, arts, etc. as an integrating part of global social systems. Hence
social anthropology can help to better understand some phenomena screened
in the archaeological record.

2.3. Palynology

Palynology is the study of living and fossil pollen18 grains and plant spores19.
The fossil pollens in which archaeologists are interested are especially those
found in peat bogs (Chrétien, 2003: 44) In fact, analysis of pollens found in
sedimentary rocks, stomachs of animals or bodies of humans can throw light
on diet, ancient vegetation and change of landscapes.

2.4. Sedimentology

Sedimentology concerns the study of modern sediments such as sand, mud

(silt), and clay, and the processes that result in their deposition.
Sedimentologists apply their understanding of modern processes to interpret
geologic history through observations of sedimentary rocks and sedimentary
structures. Sedimentary rocks cover most of the Earth’s surface, record much
of the Earth’s History, and shelter the fossil record. Sedimentology is closely
linked to stratigraphy.
The assumption that the processes affecting the earth today are the same as
in the past is the basis for determining how sedimentary features in the rock
record were formed. By comparing similar features today to features in the
rock record – for example, by comparing modern sand dunes20 to dunes
preserved in ancient aeolian21 sandstones – geologists reconstruct past
environments.

18
. Pollen: a powder produced by the male part of a flower, which is carried by insects or the wind and causes the
female part of the same type of flower to produce seeds.
19
. Spore: a reproductive cell produced by some plants and simple organisms such as ferns and mushrooms.
20
. Dune: a small hill of sand beside a beach or in a desert
21
. Aeolian: relating to the wind; produced or carried by the wind.
 19

2.5. Stratigraphy

Stratigraphy is the study of the physical and temporal relationships between

rock layers or strata. In other words, it shows chronological succession of rock
layers of earth.

2.6. Geology

Geology is the scientific study of the origin, history, structure and composition
of the earth. It helps to understand the structure and evolution of the earth.

2.7. Linguistics

The language is an expression of Culture and there is a direct link between

the language, culture and society. Resemblances and differences between
languages indicate the level of relationship existing between cultures. A
comparative study of languages reveals their common origin and diversity, as
well as their similarities. Through the classification of languages, linguists can
know origin and history of the communities who speak those languages. Here
are three ways languages can be classified:
- Genetical classification: it reveals the original community that spoke the
language
- Typological classification: grouping languages with sound resemblance
on the level of their structures and systems
- Geographical classification: comparing and grouping coexisting
languages.

2.8. Zoology

Zoology is the study of animals, including their classification, structure,

physiology, and history. It helps to understand evolution of animals and their
adaptation to various conditions.

2.9. Physics and chemistry

These sciences are necessary for laboratory works to reveal information about
the properties of material culture items during their lifetime or before they
entered the archaeological record.
2.10. Statistics
 20

In Archaeology, statistics are basically used during survey or excavation to

examine the reliability of archaeological data from which generalizations are
made.

CHAPTER 3. PREHISTORIC PERIODS

Archaeological history stretching from 2 or 3 million years ago has been

subdivided into different periods on the basis of criteria such as writing and
technology.

3.1. Classical subdivisions (criterion: writing)

Traditionally, the emergence of writing was taken as reference for dividing

prehistoric periods as follow:
- Prehistory: the period for which there is no written documents. The
study of this period relies entirely on archaeological finds.
- Protohistory: the period immediately prior to the emergence of writing,
and during which there are some documents written by foreigners on
local people.
- History: The period for which local people produce written documents
themselves.
This way of dividing periods is no longer acceptable because of the confusion
between prehistory and protohistory and the substantial meaning of the term
“writing”. Over time the Prehistory was used to refer to the period covering
Iron Age and Bronze Age, an assimilation which confuses when applied to
certain regions. For example, in most of sub-Saharan Africa stone tools were
used even after the advent of ironworking technology and this persisted at
least until the first millennium A.D. But protohistory is dated from the last
millennium B.C., whilst the existence of ancient documents dates from the
15th century or dating back to the 10th century A.D. in coastal regions.

3.2. Technological subdivisions

These subdivisions are based on the evolution of techniques used in the

manufacture of tools; and well documented archaeological periods are:
 21

- Palaeolithic: it is a long prehistoric period characterized by the

emergence of primitive man and the manufacture of unpolished
chipped22 stone tools about 2.5 million to 3 million years ago until about
12 000 B.C. The Palaeolithic is subdivided into three periods:
 Lower Palaeolothic: is the earliest of the three periods of the
palaeolithic, beginning about 3 million years ago and ending
about 70 000 B.C. with the emergence of Neanderthal man23.
 Middle Palaeolithic: is the period between Lower palaeolithic
and Upper palaeolithic. It stretches approximately between
70 000 – 32 000 B.C. Middle palaeolithic is usually taken as
equivalent to the Mousterian period24.
 Upper Palaeolithic: is the latest of the three periods of the
Palaeolithic, beginning about 40 000 B.C. and ending, in
Europe, about 12 000 B.C. Upper Palaeolithic is characterized
by the emergence of modern man, Homo sapiens sapiens.
- Mesolithic: is the period between the Palaeolithic and the Neolithic, in
Europe from about 12 000 to 3 000 B.C. The Mesolithic is characterized
by the appearance of microliths25.
- Neolithic: is the cultural period emerging after the Mesolithic. In South
Western Asia, the Neolithic lasted from about 9000 to 6000 B.C. and in
Europe from about 4 000 to 2 400 B.C. Technologically, the Neolithic
is characterized by the use of polished stone and flint tools and
weapons; whilst in social and economic domain, it is characterized by
primitive crop growing, stock rearing, domesticating animals and
sedentary life.
- Bronze Age: is a technological stage between the Stone Age and Iron
Age, beginning in the Middle East about 4 500 B.C. and lasting in

22
. To chip: to accidentally break a small piece off something
23
. Neanderthal man: is a type of primitive man, Homo Neanderthalensis or Homo sapiens Neanderthalensis,
occurring throughout much of Europe in late palaeolithic times. It is thought that Neanderthal men did not
interbreed with other early humans and are not the ancestors of modern humans. The fossils of Neanderthal man
were first found in 1857 in the Valley of Neanderthal, near Düsseldorf, in German.
24
. Mousterian: is a culture characterized by flint flake tools and associated with Neanderthal man, found
throughout Europe, North Africa, and the Near East, dating from before 70 000 – 32 000 B.C. Mousterian culture
was named from archaeological finds dating from the same period and found in the cave of Le Moustier, Dordogne,
in France.
25
. Microlith: a small Mesolithic flint tool which was made from a blade and formed part of hafted tools.
 22

Britain from about 2000 to 500 B.C. During this period, weapons and
tools were made of bronze and there was intensive trading.
- Iron Age: is the period following the Bronze Age characterized by the
extremely rapid spread of iron tools and weapons. In the Middle East,
Iron Age period began about 1 100 B.C. and in the second half of the
last millennium B.C. in Great Lakes Africa, and in the first centuries of
the first millennium A.D. in the rest of sub-Saharan Africa.
Prehistoric periods subdivided on the basis of technological evolution vary
from a region to another, thus the periods recognized in the archaeological
record across Great Lakes Africa will be displayed in chapter V and VI.

PART II. TECHNOLOGICAL EVOLUTION AND MODES OF LIFE IN

PREHISTORIC RWANDA

Great Lakes Africa in general has a long record of human occupation shown
by ancient industries that represent almost all the periods of the history of
humanity.

CHAPTER 4. STONE AGE INDUSTRIES

4.1. Early Stone Age

In sub-Saharan Africa, Early Stone Age is an ancient prehistoric period

corresponding to the Lower Palaeolithic. This period is represented by two
lithic industries: Oldowan and Acheulean.

4.1.1. Oldowan

Oldown industry encompasses the oldest stone tools in Africa and adjacent
regions of Eurasia. It is characterized by deliberate flaking of stone to produce
sharp cutting, chopping, and scraping edges, as well as tool use in the form
of unmodified hammerstones and stone anvils used for pounding and grinding
(Barham and Mitchell, 2008: 60). Oldowan tools are dated to the period
between 2.4 – 1.5 million years ago. They have been found in Olduvai Gorge
(Tanzania), on the shores of Semiliki River, along the shores of the Lake
Turkana (Kenya), and in the Valley of Omo (Ethiopia).
 23

In Eastern Africa, Oldowan tools are associated with hominin fossils, probably
Homo habilis. Homo habilis (2.3 – 1.6 million years ago) were hunter-gatherers
that used stone tools and possibly tools made in perishable material such as
wood, bones, etc. Oldowan tools are divided into two phases: Classic Oldowan
and Developed Oldowan. The latter is the last technological phase of oldowan
industry and is characterized by the appearance of bifacial flaked tools.

4.1.2. Acheulean

Acheulean industry appears to have been developed in Eastern Africa (in Bed
II of Olduvai Gorge) from about 1 million years ago. It is mainly characterized
by hand axe development. Outside Africa, Acheulean tools have been found in
Asia and Europe.

In Africa, Acheulean industry is characterized by the regularity of techniques,

and making bifacial tools especially large flakes as blanks for hand axes and
cleavers26. Acheulean tools are associated with Homo Erectus, hunter-
gatherers who lived in savanna and woodlands. They moved periodically
following the fluctuation of animal and plant resources.
There are three phases of Acheulean: ancient, middle and upper Acheulean.
The late phase is characterized by tools miniaturization, noticeable evolution
of mode of life and hunting methods. However, neither specialization nor
regional adaptation can be attributed to Acheulean tools makers.

4.2. Middle Stone Age

In sub-Saharan Africa, the Middle Stone Age is a prehistoric period starting

after the Acheulean and corresponding to the Middle Palaeolithic as well as a
part of the Upper Palaeolithic. It is characterized by a diversification of
industries adapted to environments and the persistence of microlithic tools
production. The Middle Stone Age tools belong to Homo Sapiens. It is
characterized by significant technological and cognitive transitions, including
the abandonment of bifaces, the widespread adoption of hafted tools, and the

26
. Cleaver: a heavy knife with a large square blade
 24

appearance of symbolic behaviors (Barham and Mitchell, 2008: 257). The use
of fire is documented in the Middle Stone Age.
The second aspect of the Middle Stone Age is a regional specialization that
would have given the population adaptive flexibility in a variety of
environments (Barham and Mitchell, 2008: 237).
Middle Stone Age adapted industries are:

4.2.1. In the forest

4.2.1.1. The Sangoan

The Sangoan industry was named after surface finds made near Sango Bay,
Lake Victoria, Uganda. It groups lithic industries developing after the upper
Acheulean across Central Africa. The Sangoan is the first evidence of human
occupation in a large part of equatorial forest. It testifies an adaptation to the
dry period during which equatorial forests have been considerably reduced.
The Sangoan is characterized by microlithics tools including few or no
handaxes and cleavers, but by heavy-duty picks that show continuity with the
Acheulean and the innovation core-axes. Small flake tools are common, with
Levallois core and the first bifacial points also present (Barham and Mitchell,
2008: 234-235). The exact date of Sangoan industry is well not known, but it
can be situated about 100 000 – 80 000 B.P.

4.2.1.2. The Lupemban

The Lupemban was named after the Lupemba river terraces in Kasai province,
Democratic Republic of Congo, where mining uncovered deposits containing
blades and elongated bifacially retouched points (Barham and Mitchell, 2008:
235). The oldest lupemban layers are dated between 42 000 – 38 000 B.P. In
central Africa, Lupemban industry is characterized by a systematic blade-
making, backed tools and bifacially retouched points (Barham and Mitchell,
2008: 223).

4.2.2. In savanna regions

- Fauresmith industry: It was named after the finds from near the
Fauresmith town of Orange Free State in South Africa. It is contemporaneous
 25

with the upper acheulean. It is characterized by large blades, prepared cores,

Levallois unifacial points, and smaller bifaces.

4.3. Late Stone Age

In sub-Saharan Africa, Late Stone Age is a prehistoric period corresponding

with the upper palaeolithic. The diversification of lithic industries recorded in
the Middle Stone Age persists in the Late Stone Age.

4.3.1. In the forest

- The Tshitolian: it was named after the finds from the site of Tshitolo on the
Plateau of western Kasai (DRC). It is characterized by the persistence of
geometric microlithics, bifacially retouched points and core tools.

4.3.2. In savanna areas

- The Wilton: Wilton industry was named after the finds from a rockshelter
in the farm of Wilton, Cap, South Africa. Appearing from 18 000 B.P., the
Wilton industry is characterized by increasing use of finer-grained rocks,
systematic production of backed microliths (segments, backed points, and
bladelets) and small, standardized, often thumbnail-shaped scrapers, and a
greater variety of wood, bone, and shell items (Barham and Mitchell, 2008:
318).

4.4. Prehistoric periods in Rwanda

4.4.1. Early Stone Age

In Rwanda, lithic tools have been found both at open sites and rock-shelters.
However, no Early Stone Age site has been recorded with precision yet.

4.4.2. Middle Stone Age

This period is characterized by flaked tools belonging to industries adapted to

different environments.

4.4.2.1. In the forest

Lithic industry adapted to forest areas documented in Rwanda is the

Lupemban recorded at the Plage des Hippopotammes in the eastern region.
 26

4.4.2.2. In savanna areas

Lithic industry adapted to savanna areas in Rwanda has been found at

Rutonde by J. Nenquin. He excavated four layers, and the first layer provided
Magosian27 tools including cores, bifaced and unifaced points, scrapers,
backed blades or flakes and other knives, etc.

4.4.3. Late Stone Age

The Late Stone Age industries are not well dated in Rwanda, and we cannot
be sure whether all recorded material can be attributed to this period. This is
because in sub-Saharan Africa there are some microlithic tools that are
completely different from those dated to the Middle Stone Age, but also do not
belong to the Late Stone Age. Such artifacts are attributed to a transitional
industry covering the period approximately between 40 000 – 20 000 B.P.
These lithics include blades and hafted tools.

4.4.3.1. In the forest

The Tshitolian is the main assemblage recorded in Rwanda.

4.4.3.2. In savanna

The Wilton: artifacts attributed to Wilton Industry have been identified in the
rock-shelter located in South-western Rwanda. This rock-shelter named
Mukinanira was excavated by J. Hiernaux. Artifacts found at the Plage des
Hippopotammes also include Late Stone Age material.
Late Stone Age industries are exclusively characterized by geometrically
shaped microlithic material and possibly that is the period during which bow
and allows were used for hunting. To these innovations can be added the use
of poisonous allows and nets when hunting. Bones were used for preparing
cloths, beads and points. Microlithic tools may have been associated with
wood and/or bones for making hafted tools although we do not find them
because organic material is not preserved for a long time in the archaeological
record.

27
. Magosi: an archaeological site found in Uganda.
 27

4.4.4. Neolithic

The existence of a "Neolithic period" remains an open question in the

prehistory of Great Lakes Africa in general (Chrétien, 2003: 45), thus it is early
to confirm the existence of that period in Rwanda. There is only little
information on the transition between the Late Stone Age and the Early Iron
Age. Excavations at open sites have revealed an association of LSA artifacts
with EIA potsherds in the same contexts, a disturbance that has been
interpreted as follow:
- Firstly, there may have been a coexistence of two populations, one using
lithic tools, the second having already acquired agricultural and metallurgical
technologies.
- Alternatively, local population-users of lithic tools may have acquired pottery
brought by newcomers (iron users).

CHAPTER 5. IRON AGE

Iron Age period in Great Lakes Africa stretches approximately from 500 B.C.
to 1900 A.D. (Giblin 2010). It has been traditionally divided into two
components: the Early Iron Age (EIA) and the Late Iron Age (LIA) (e.g.: see
Kanimba and Van Pee 2008). The transition period between the EIA and the
LIA is poorly understood because of the lack of archaeological evidence,
although it is believed to have been characterized by the transformation of iron
production alongside the societal change across the region (Humphris
2010:33). However, recent works curried out in the region (e.g. Robertshaw
1991, Giblin 2010, Humphris 2004, 2010, Iles 2011) have brought new
insights and seem to recognize some kind of transformational period, from a
technological stance at least. Humphris has applied the term “Middle Iron Age”
to this period, although such a classification has not been yet accepted within
the wider archaeology of the region28.

28
.
 28

In fact, ceramics from Gogo Falls (Robertshaw 1991) and archaeometallurgical

materials from Rwanda (Humphris 2010: 102) have revealed that “society and
technology were changing together in varieties of ways across the region”
between the 9th century A.D. and the 16th century A.D. Particularly in
Rwanda, this was an era of great change where emergence of complex society
is evident in technology, politics, social and economic systems. Broader and
dramatic impact of these changes has shaped new lifestyles of the society
which is why Humphris (2010) terms this period as the Middle Iron Age (MIA)
in recognition of the period as one of transition.

5.1. Early Iron Age

In Rwanda, C14 dates for the earliest iron smelting sites lie in the last
centuries of the first millennium B.C. (Van Grunderbeek et al. 2001:276).
Geographically, these sites are concentrated in the southern region of the
country formerly called Butare, a Kinyarwanda word that means a “place of
iron ore” (Giblin 2010:112). Early iron working communities’ technology has
been associated with “Urewe culture” (Craddock et al. 2007:2). The latter takes
its name from pottery and a site in Kenya at Kavirondo Bay located northeast
of Lake Victoria dating from the first millennium B.C. The main characteristics
of Urewe culture are remarkable uniform ceramics with incised patterns, and
sophisticated iron smelting furnaces, sometimes with decorated bricks
(Grunderbeek and Roche, 1989). A developed Urewe culture flourished in
Great Lakes Africa from the first millennium B.C. to the 7th century A.D. and
lasting up to the second millennium in some places.

The settlement map of Urewe users in Rwanda and Burundi (see Craddock et
al.2007:2) suggests that the EIA population lived in places suited for mixed
economy including iron production. Archaeological evidence from the
southern part of Rwanda supports such a notion and suggests that the Butare
area was ideal due to the availability of iron ore and wood for charcoal (Van
Grunderbeek et al. 2001:273).
 29

Urewe pots and decorated bricks used to build up furnace wall (Exhibited at the
Ethnographic Museum of Huye)
Photos: Kanimba and Van Pee 2008

Early Iron Age Agriculture

In Great Lakes Region, different traditions of food production stretch back at

least 2500 years and possibly another millennium before, but significant
agricultural change is noticeable since the first millennium B.C. That change
has been associated with the settlement of different iron-using communities
who raised domestic animals, sowed grains and planted root crops in the west
and central parts of the Great Lakes Region beginning as earlier as 500 B.C
or possibly as early as 900 B.C. (Schoenbrun 1983: 271).

Since the arrival of the makers of Urewe ceramics in the first millennium B.C.,
archaeological records show that they have occupied the plateau or hill regions
where the subsoil provided iron ore in small, easy to exploit deposits and the
natural tree savannah environment could be cleared to produce charcoal for
firing (Van Grunderbeek and Roche, 1989). This settlement pattern - where
populations left “poor land” and colonized “best land”, mainly because they
were looking for suitable land both for agricultural practices and iron
production -, is believed to result from the efficiency of iron technology which
enabled them to cut forests and produce foods. All over Great Lakes Region
(i.e. Rwanda, Burundi, the north and west of Victoria Nyanza in Kenya,
Uganda, Tanzania, and in the west of Lake Kivu in DRC), archaeological
researches have documented the settlements of the early iron-using
communities in the woodland savanna of the hills (generally with elevation of
 30

1500-1700 above sea level) from the first millennium B.C. until the 7th century
A.D. (Van Grunderbeek and Roche, 1991: 304).
However, as Philipson (2005: 251) emphasizes, although effective research has
not yet been undertaken to illustrate the subsistence economy practiced by
the EIA communities; there are some indications that they were probably
herders of domestic cattle and cultivators of both finger millet and sorghum.

Particularly in Rwanda, there is a general lack of archaeological evidence and

materials available for the EIA can only give superficial information on
subsistence economy. Pollen analysis from Butare indicates that mixed
farming including agriculture and herding was into subsistence strategies
throughout that period (Humphris 2010: 92-93). For example, materials
excavated on the hill of Kabuye suggest the cultivation of finger millet, eleusine
and sorghum between 100 and 300 A.D. These cereals were cultivated in small
plots, and are likely to have been brought by iron smelting communities from
outside the region (Van Grunderbeek and Roche (1991: 307).
Another site that provides information on the subsistence of EIA population is
Kabusanze where palaeobotanical analyses suggest the cultivation of sorghum
and pearl millet in the region (Giblin 2008). Herding activities during this
period are documented by the remains (teeth) of a domestic animal found
within an iron furnace at Remera (Van Noten 1983: 65, Kanimba 1997).
The shifting agriculture combined with the heavy demand for firewood for
smelting has contributed to the rapid expansion of iron technology in Great
Lakes Region (Stahl, 2005: 403). Thereafter, this expansion has accelerated
forest clearance both for agriculture and iron production favoring the
pastoralism of cattle by creating areas, free of thick bush, tsetse flies, and
ticks29, suitable for cows. This is one of the factors that favored the beginning
of the specialization of cultivation and cattle herding in Great Lakes Region
around 800 A.D. The mixed agro-pastoral economy in the region is evidenced
by abundant cattle remains, grindstones and sorghum often found at sites
dating from 800 A.D. (Stahl, 2005: 403). Southwestern Uganda and Victoria

29
. Tick: a very small creature like an insect which lives on and sucks the blood of other animals.
 31

Nyanza are examples of regions where economy was dominated by specialist

cattle, small-stock husbandry and banana cultivation (Ashley, 2010; Connah,
1997: 29).

5.2. Middle Iron Age?

Around 800 A.D. technological transformation in the archaeological record

began to be seen. Ceramic production and use moves away from urewe pottery
to roulette-decorated ceramics, and iron smelting practices “change from
decorated-brick furnaces producing smaller flows of slag, to apparently
undecorated furnaces which can produce large blocks of slag” (Humphris 2010:
99, Soper 1985:48). The period is also characterized by the rarity of iron
furnaces in the archaeological record until the 11th century, suggesting that
there was a regression of smelting activities. This regression is believed to be
one of the consequences of environmental degradation occurring around 600
A.D. However, few furnaces dated between 800 and 1100 A.D - such as those
excavated at Cyamukuza (Van Grounderbeek at el. 2001: 295) - differ from
furnaces used during the EIA, and are similar with the ones dating after the
11th century A.D. This means that the new technology occurring around the
8th century A.D. has continued even after the 11th century in spite of
regression. It was the beginning of the Middle Iron Age.

MIA technological change appears to be characterized by a simplification of

techniques used by artisans, both in iron production and ceramics. Iron
furnaces built with decorated bricks have disappeared; potters have reduced
a number of vessel forms and ornate decoration, and adopted roulette -
decorated ceramics (Giblin 2010, Humpris 2010).

Alongside technological transformation, is social complexity and political

organization occurring after 800 A.D. In Rwanda, the new socio-economic
system is believed to have led to the centralization of power and emergence of
kingdom, dominated by Nyiginya dynasty since around the 17th century A.D.
(Kohtamaki 2010:300) although other researchers suggest that the kingship
institution may be for an earlier date probably around 1500 A.D. (Kanimba
 32

2008; Guddal 2010:30). Archaeometallurgical materials suggest that towards

the end of the MIA, iron production intensified alongside the socio-economic
and political shift towards kingship, to eventually become a cultural pillar in
the iconography of the kingdom throughout the LIA (Humphris 2010: 103).

5.3. Late Iron Age

From the 17th century A.D., new and developing socio-political configurations
emerged in Rwanda, formulating the Nyiginya kingdom. Oral histories,
archaeological and ethnographic sources indicate that this kingdom had
“social structure based around a king and a top-down level of organization and
power” (Humphris 2010:104). The court was organized around symbolic
legitimization of the king; and apart from dynastic drums, it seems that iron
played an important ritualistic role associated with the power of the King. This
assumption is supported by iron objects exhibited at the Ethnographic
Museum of Huye including hammers; spears and anvil, found in the grave of
the monarch Cyilima Rujugira who died around 1700 A.D.

Environmental consequences of iron production

Although the advent of ironworking technology undoubtedly changed the life

style of individuals, it also had environmental impacts which can be seen in
the archaeological record across the region.
The impacts of ironworking technology gradually became acute as the
technology spread, because there was a growing dependence on the
environment for wood for fuel used in both smelting and forging. Thereafter,
iron production increased, forest exploitation exceeded its regeneration; and
over time, the regions around iron production centers became entirely
deforested. Consequently, such gradual deforestation has favored wind and
water erosion of the exposed soils.
 Deforestation
Pollen analyses demonstrated that between 3000 and 2500 years ago, the
hillsides and valley-bottoms in the region which is currently Uganda were
covered by secondary forest (Schoenbrun, 1994: 275). Charcoal and pollen
analyses from the EIA sites in Buhaya (Tanzania), Kabuye (Rwanda) and
 33

Ndurumu (Burundi) produced similar information (Schmidt, 1997; Van

Grunderbeek and Roche, 1991). More specifically, the investigations carried
out in Kemondo Bay revealed that the charcoal used at the site KM3 had been
produced from at least twelve tree species, while the charcoal used at KM2
had been obtained from sixteen tree species. To interpret this, Schmidt (1997:
406) claims that such a variety of wood indicates an environment of many
mature trees; but also it shows that “by the early first century B.C., iron
technologists were exploiting predominantly first-growth, mature, wet-forest
trees” from both valleys and ridge tops. But other archaeological evidence from
the same sites with a later date has revealed that by the 4th and 5th century,
there was disappearance of many tree-species. For example, at KM3, only 2
moist-forest species were identified and 3 species at KM2. Therefore Schmidt
(1997: 407) concludes that by around the 5th century A.D. “clearly the variety
of forest species formerly found in the environment was no longer available.
Large remnant trees were used, but the iron smelters by then consistently
ventured into the swamps to obtain supplementary fuel”.

Another case of forest clearance during the EIA was identified in Kigezi forests
(Uganda) where the mode of clearance was by burning and cutting, for grazing
as well as farming purposes. Pollen analysis from Ahakagyezi (Schoenbrun,
1994:280) has revealed that the degradation of that forest started around 200
B.C. when iron-using communities settled in that region. According to this
analysis, it appears that after clearing surrounding areas, these communities
have moved gradually up the slopes of Kigezi's up-lands in search of more
resources (land for cultivation and wood for charcoal). By moving up the slope,
they continually cut mature trees and this process reduced forest cover more
generally than in the previous thousand years. The clearance of Kigezi forest
has led to the nearly complete removal of even ridge top forests by about AD
1000. Briefly, archaeological evidence from that region let us believe that as
charcoal requirements for iron production consume vast quantities of specific
types of wood, this was likely a major motivation in the clearing of Kigezi
forests (Schoenbrun, 1994:279-280).
 34

In Rwanda and Burundi, the distribution of natural vegetation depended on

factors such as elevation, temperature and rainfall before the advent of iron-
using communities. But when Urewe users settled in the region, they cleared
surrounding areas for both agriculture and iron production. This is probably
why human impact on the environment is noticeable since the first
millennium A.D., becoming devastating from 500 A.D. Later on, the effects of
human activities were exacerbated by a dry climatic phase (Van Grunderbeek
and Roche, 1991: 302).
 Erosion
The consequences of iron production and agricultural activities that occurred
in the middle of the first millennium A.D. include erosion alongside
deforestation and land clearance. Several cases of erosion have been reported
all over the region, either by archaeological excavations or by pollen analyses.
At some sites, experienced archaeologists can even notice these erosion effects
immediately as they are generally evidenced by disturbances encountered
during excavations. In Rwanda, Van Noten (1983: 3-4) has reported that at
the Stone Age sites, artifacts of different periods are often mixed in one
stratigraphic layer, and this is likely to be the results of hill creep movement,
suggesting an erosive sequence occurring after the Stone Age. Van Noten
(1983) reached that conclusion after noticing that the disturbances were only
encountered at open sites, whilst the distribution of finds from sites with good
preservation conditions like caves and rock-shelters respected the order of
stratigraphic layers, for example in Ruhimandyarya and the rock-shelter of
Mukinanira.

On the central plateau nearby Butare, pollen analyses of valley deposits have
also confirmed that from the 2nd century A.D., human activities caused soil
degradation which increased as climatic conditions worsened at the end of the
4th and the beginning of the 5th century A.D. In their interpretations, Van
Grunderbeek and Roche (1989) said that the materials from that region show
the erosion following not only deforestation for iron smelting on the hills, but
also coincided with the first agricultural activities, mainly for the growing of
cereal crops by slash and burn. Archaeological materials from Kabuye
 35

(Schoenbrun, 1994:280 citing Van Grunderbeek and Roche 1985) suggest

that the heavy soil erosion and a general degradation of soil fertility caused by
a prolonged period of settlement and exploitation, possibly led to a reduction
in human activity in the south of Rwanda between about A.D 600 and 1000.
Another case of erosion has been reported by pollen analyses from Lake Ikimba
(Buhaya) where a phase of erosion occurred during the 3rd century B.C. and
became very strong, approximately A.D. 400 to 500. These analyses suggest
that this erosion has resulted from a process of forest clearance starting in the
valley bottoms, then progressing to the slopes and finally reaching the ridge
tops (Schmidt 1997: 408-409). A similar sequence of erosion was identified in
south-western Uganda.
To conclude this section, it is clear that iron production had a significant
impact on the populations and the environment within which it was practiced.

CHAPTER 6. EMERGENCE OF GREAT LAKES KINGDOMS

The Great Lakes Africa kingdoms emerged in the LIA resulting from increasing
hierarchical social systems occurring in the first half of the second millennium
A.D. across the region. The hierarchical social system was stimulated by
competition for prime resources as population pressures increased in some
areas. Elite individuals were able to manipulate large groups of people,
legitimizing their position by assumption of sacred status and basing their
power on economic control (Connah 1989; 2001:290).

In fact, a mixed farming tradition relying on the cultivation of cereal and other
crops as well as on the herding of domestic animals, probably supplemented
by fishing and hunting, has been documented in Rwanda and neighboring
countries since the EIA (Van Gunderbeek 2001, Giblin 2008, Giblin et al.
2010, Connah 1989, 2001), although this domestication may be of an earlier
date in the “Neolithic”.
The region offers advantages based on a generally good rainfall, equatorial
temperatures moderated by altitude, and relatively fertile soils that permitted
the growing of a range of crops including finger millet, sorghum, and above
 36

all, bananas in many varieties. Lakes and rivers yielded important supplies of
fish which, until recent times, was also substantial resource of wild game
(Connah 2001:268, Kusimba and Kusimba 2005). Forest clearance – resulting
from EIA iron working technology and agricultural activities – increased the
availability of best grasslands, often free of tsetse, and permitted specialized
pastoralism and intensive banana farming that developed between 800-1300
A.D. (Stahl 2004:157).

For Rwanda, Burundi, Buganda and north-west Tanzania, another important

economic resource was iron; whilst in Upemba Depression it was large copper
deposits. In Bunyoro, the salt was also an important factor for accumulating
wealth.
Such a sound subsistence economy is likely to have promoted the population
growth that inevitably led to competition for control of the best of those
resources. Families who succeeded in gaining that control could then
accumulate a surplus in one form or another and use this as means of
establishing authority over larger numbers of the less fortunate members of
the population (Connah 2001:287).

However, it seems that there were more enviable resources than others which
became limited as the population increased in number. Researchers (e.g.
Connah 1989; 2001:290; Reid and MacLean 1995) suggest that, apart from
land for cultivation, enviable and limited resources are likely to be cattle (self-
reproductive resource in a way land is not), iron (the very basis of the
technology on which society depended), salt (so keenly sought for both human
and livestock use) and copper (valued symbol of status and means of exchange
in Upemba Depression).

Furthermore, there were more suitable lands for these limited resources than
other lands, and groups who had access on these lands accumulated more
wealthy than the rest of the population (Connah 1989, Guddal 2010). This
tends to be the case of cattle-keeping groups who, thereafter, established a
cattle protection and distribution system, a process which led to the creation
 37

of wealthy and powerful elites. In Rwanda, the distribution system of cattle

consisted in social contract (clientage/ubuhake) between pastoralists
(patrons) and ones who desired to have cattle (clients). The client had to pay
service and give military support to his patron in times of unrest or merely if
he wanted to expand his fief (Guddal 2010). Although there is a lack of
archaeological materials that can show when the clientage started in Rwanda,
it is believed to have favored pastoralist families and led to the emergence of
kingdom dominated by Nyinyinya dynasty since around the 17th century A.D
(Kohtamaki 2010:300, Humphris 2010, Guddal 2010, Giblin 2010).

In neighboring countries, archaeological evidence suggests the existence of

chiefdoms based on cattle pastoralism and cereal cultivation or copper
exploitation since around the 14th century A.D (Connah 1989, 2001,
Robertshaw 1994, Reid & MacLean 1995). Archaeological remains from Ntusi
in Uganda show that the site was a chiefdom based on cattle pastoralism and
sorghum cultivation, rather than a centralized state, by the 14th and 15th
century A.D. This means that a social hierarchy based on cattle had perhaps
not yet emerged. In the 15th century the inhabitants of Ntusi abandoned the
site and probably occupied Bigo in the north of Ankole, a site bordered on one
side by the Katonga River, since it was more viable centre for cattle than the
dry grassland area of Ntusi (Connah 2001:282). The evidence from Bigo is a
royal enclosure associated with important cattle remains that indicate the
importance of cattle to its occupants (Posnasky 1968). The settlement is also
associated with large earthworks found there and for researchers (Posnasky
1969, cited by Connah 1989: 225), the construction of these earthworks with
all that it implies in terms of control of resources, planning and direction of
labour suggests the existence of powerful elites whose wealth was
accumulated in the form of cattle.
Concerning Bunyoro, Robertshaw (1994) claims that a hierarchy of site sizes
existed from about the 14th century A.D, some of the large being protected by
earthworks. The period of competing polities between about the 14th and 16th
century may eventually have led to collapse, and the Babito dynasty who ruled
Bunyoro when the first Europeans arrived were able to exploit it. Thus,
 38

Bunyoro may have emerged as a kingdom in the 17th century A.D, although
the socio-political changes that gave rise to it could have started earlier
(Connah 2001:283).
In summary, the main factor of the emergence of kingdoms around the 17th
century A.D in Great Lakes Africa is likely to be the population pressures
resulting in competition for some particularly important, but limited
resources. In Rwanda, Burundi, Karagwe, Ankole and Buganda, cattle were
undoubtedly of considerable importance, but also the smelting of iron must
have been of great significance. The social and economic role of iron has
definitely valued to it a political and religious meaning. That’s why iron anvils
have been documented both archaeologically and ethnographically among
items of royal insignia across the region. The important factors underlying the
accumulation of the wealth and power were access and control of grazing
rights, fertile cultivable soils in higher rainfall areas, water supplies for
livestock, iron ore and charcoal. Particularly in Bunyoro, research shows that
access to the sources of salt and control of its production played also
important role in the rise of the kingdom (Robertshaw 1994, Connah 1989,
2001).

For Upemba Depression, it could have been accessing to prime fishing areas,
or to fertile alluvial soils in an environment where dense population has to
compete for the few stretches of land not inundated during the annual flood.
Furthermore, the eventual use of copper as a form of currency indicates that
a political structure has gradually developed prior to the kingdom. It seems
that a society of cultivators, herdsmen and fishermen was ruled by small
privileged groups – likely groups who had access and control of copper – that
had attained at least chieftain status, and one of which was able to found the
Luba State by the 18th century A.D (Connah 1989, 2001:286).

6. 1. Consolidation and expansion of Great Lakes kingdoms

6.1.1. Consolidation of the kingship institution

It has been mentioned that pastoralism, agriculture and iron working are
salient elements of the LIA economy, and the position of the rulers depended
 39

upon the control of conflicting interest groups involved within these activities.
It seems that the king maintained his control over these groups through the
use of ritual and symbolism which gave him the power to mediate between the
natural and cultural world, between ecology, technology and society (Reid and
MacLean 1995:153). Traditionally, these powers of mediation were the powers
of smelter who was able to transform ore into iron (Herbert 1993, Haaland
2004). That is why, across Great Lakes Africa, kings legitimated and
consolidated their power by identifying themselves with ironworking lineages
as means of creating identity and a way of opening new paths for central
taxation and direct management of iron production (Schmidt 1997:404).
The symbolic association between leadership and iron working has highly
contributed to the consolidation of the kingship institution, as it has been
documented both archaeologically and ethnographically (Schmidt 1997, Van
Noten 1983). This relationship was manifested in royal regalia containing
various iron items. For instance, ethnographic materials from Rwanda
(Kanimba and Van Pee 2008:45) show that, alongside the royal drum, the royal
regalia included:
- Two iron blades (nyarushara), each furnished with two hooks: the king
kept them under his bed each night even when traveling;
- An adze (incyamuro): a ritual protection for the emblem drum Karinga;
- A smith’s hammer probably illustrating the idea of “the smith king”;
- Two burins (ishinjo): guardians of the dynastic drums that were stuck
each evening by the king while he delivered his ceremonial speeches;
- The axe of justice that was hanged at the entrance of the court during
royal trials.

Additionally, grave-goods including iron spears, anvil and iron hammers from
the tomb of a 17th century monarch, Cyilima Rujugira, demonstrate the
importance of iron within the symbolism of kingship and the power of the
kingdom (Van Noten 1983).
The use of iron as a symbol of power has also been documented in Karagwe
where royal insignia included four hammers/anvils and ten iron cows. In their
interpretations, Reid and MacLean (1995:146) claim that such items indicate
 40

the deliberate incorporation of iron working within a wide framework of

economy, symbolism and power.
Another state that has used iron to consolidate its kingship institution is the
Kingdom of Luba. Grave-goods excavated at Nsanga are quite informative and,
furthermore, suggest that iron was not only associated with power for its
economic importance, but also its transformative nature. This is because,
although wealth in Luba was based on copper and agricultural means, these
graves contained iron items including axes, anvils or bells that are thought to
be symbols of power. Connah (2001:286) claims that individuals with who
they were buried were invested with political or sacred authority.

6.1.2. Wars of expansion: Case of Rwanda

In Great Lakes Africa, emergent states shared the same cultural and
traditional features, and their politics have been shaped by competition and
warfare until colonial contact at the end of the 19th century (Kusimba and
Kusimba 2005). This is shown by LIA archaeometallurgical remains from
Rwanda (Humphris 2010), Buganda (Humphris 2004) and Bunyoro (Iles 2011)
dominated by large slag blocks and large furnace pits which imply that there
was a mass of iron production, not only for providing the kingdoms with
agricultural tools and royal regalia items; but also with weapons. Particularly
in Rwanda, archaeological evidence corroborates ethnographic materials
exhibited in museums and oral traditions which inform us that the Nyiginya
dynasty had extensive armies involving thousands of warriors that may have
demanded a considerable supply of iron weapons (Humphris 2010).

After the emergence of the kingdom of Rwanda, through conquests and raids,
more land and people have become incorporated within this kingdom under
the power of the king supported by ritual institutions (Newbury 1974; Vansina
2004). Throughout the reign of Rwabugiri in the 19th century A.D, the king
and his troops conducted several expeditions in order to annex all
Kinyarwanda speaking chiefdoms, and form one political unit under his
control. That’s how Gisaka in the east, Kinyaga and other regions in the west
were annexed to Rwanda (Newbury, 1974:182).
 41

Nonetheless, Rwandan troops attacked other chiefdoms on the western shore

of Lake Kivu including Bunyabungo, Buhavu, Buhunde, Ijwi, Butembo and
Bushi, not for incorporating them administratively, but for weakening them
by killing their chiefs and grabbing cattle (Newbury 1974:182). Royal accounts
reiterate that Rwandan famous warriors such as Rwanyonga, Nyamushanja
and Nyirimigabo were killed during some of these expeditions. Most successful
military expeditions reiterated in royal accounts are the expedition of
Bumpaka (in the east of the Lake Edward in Uganda) for having grabbed large
cattle, and the expedition against Burundi (Newbury 1974:185). The latter is
even documented ethnographically by a royal drum exhibited at the
Ethnographic museum of Huye captured from Burundi after it was defeated.
Although archaeology has contributed relatively little to our knowledge of
Great Lakes kingdoms’ hegemony, ethnographic materials and oral traditions
discussed above show that expansion was the target of their rulers. This may
explain why the states held control over iron production and blacksmiths
remained the kings’ favorites because they made deadly arrows and spears
that helped to conquer the enemies and accumulating wealth (Humphris
2010:105).

6.2. Subsistence economy

Due to the limitation of archaeological evidence to iron smelting and burial

sites, there is a lack of information about settlement sites across Great Lakes
Africa; thus it is difficult to assess the subsistence economy in detail. However,
a sound subsistence economy, dominated by cattle pastoralism, banana
farming and cereal cultivation, is believed to be a highly successful one that
supported LIA substantial populations across the region (Connah 2001:284).
In Rwanda, the subsistence of most of the population laid on agriculture and
cattle-raising. Plants of the rainy season, such as beans, peas, finger millet
and sorghum were planted on the sides and tops of hills; whilst sweet potatoes
and other dry season crops occupied the lowlands and wet areas. Banana
garden and beans fields often occupied the same plot around the home area
 42

(Kanimba and Van Pee 2008:133). Storage was done in granaries and large
grain baskets of various sizes depending on the quantity to be stored.
Cattle-raising was practiced at least since A.D. 300 and, alongside its principal
role in the political structure and social relationships, the use of secondary
cattle products, such as various forms of milk and blood, indicates how it was
integrated into food system not only in Rwanda, but across the region
(Schoenbrun 1993).
Although agriculture and pastoralism were the major occupation of most of
the Rwandese, there was also a hunter-gatherer community whose
subsistence economy depended upon food harvest and hunting. The harvest
included gathering crops, fruits, leaves, tubers, roots, and wild mushrooms
(Kanimba and Van Pee 2008:132).
It seems that the LIA subsistence economy had many similarities across Great
Lakes Africa. But the work of Reid (2002) suggests that Buganda is an
exception because its subsistence depended much more on banana farming
than other food sources.
For Luba kingdom, fertile, well-watered, alluvial soils and access to mountain
grasslands permitted a mixed agriculture economy, to which hunting of the
prolific local game made a significant contribution. This is evidenced by oil-
palm nuts, grain of finger millet, bones of goats and chicken, fish bones,
mollusk shells and bones from wild animals found in various archaeological
contexts (Connah 2001:284).
Briefly, it appears that the Great Lakes kingdoms had a high subsistence base.
Alongside a cattle-dominated livestock, banana farming, fishing and hunting,
families were capable of keeping some stock of grains. Thus, states were able
to support locally dense populations, a typical case for Rwanda.
 43

REFERENCES

1. ALPERN, S. B., 2005. Did They or Didn't They Invent It? Iron in Sub-
Saharan Africa. In History in Africa, Volume 32, pp. 41-94.
2. ATHERTON, H. J., 1983. Ethnoarchaeology in Africa. In The African
Archaeological Review, vol.1:75-104.
3. BARHAM, L. and MITCHELL, P., 2008. The first Africans: African
archaeology from the earliest tool makers to most recent foragers,
Cambridge.
4. BENTON, M. J. and HARPER, D. A. T., 2009. Introduction to Paleobiology
and the Fossil Record, Oxford, UK.
5. BENTON, M. J., 2005. Vertebrate palaeontology, 3rd ed., Oxford, UK.
6. CELIS, G., 1987. Introduction à la métallurgie traditionnelle au Rwanda.
Techniques et croyances. Institut Nationale de Recherche Scientifique.
Butare-Rwanda.
7. CHILDS, S. T. and KILLICK, D. 1993. Indigenous African Metallurgy:
Nature and Culture. Annual Review of Anthropology, Vol. 22. 317-337.
8. CHRETIEN, J. P., 2003. The Great Lakes of Africa: Two Thousand Years
of History, (Translated by Scott Straus), New York.
9. CLARKE, L. D., 1968. Analytical Archaeology. Columbia University
Press, New York.
10. COLLETT, D. P., 1993. Metaphors and representations associated
with precolonial iron-smelting in eastern and southern Africa. The
Archaeology of Africa, Food, Metals and Towns: 499-511.
11. CONNAH, G., 1989. African civilizations. Precolonial cities and
states in tropical Africa: An archaeological perspective. Cambridge.
12. CRADDOCK, P. et. al., 2007. Early Iron Age iron-smelting debris
from Rwanda and Burundi, East Africa. Histrorical Methallurgy, Vol. 41,
Part I. 1-14.
13. D’HERTEFELT, M. et COUPEZ, A., 1964. La royauté sacrée de
l’ancien Rwanda. Tervuren.
 44

14. DESMEDT, C., 1991. Poteries anciennes décorées à la roulette

dans la Région des Grands Lacs. The African Archaeological Review, Vol.
9. 161-196.
15. FAGAN, M. B., 2006. Archaeology. A Brief Introduction. Ninth
Edition. New Jersey.
16. GIBLIN, J., 2008. New work on the Later Archaeology of Rwanda
2006 to 2007: A preliminary fieldwork report. In Nyame Akuma, No 69,
pp 45-55.
17. GIBLIN, J., 2010. Re-constructing the past in Post-Genocide
Rwanda: An Archaeological contribution, unpublished PhD thesis, UCL.
18. GIBLIN, J., HUMPHRIS, J., MUGABOWAGAHUNDE, M. and
NTAGWABIRA, A., 2011. Challenges for Pre-Colonial Archaeological Site
Management in Rwanda. In conservation and management of
archaeological sites, Vol. 13 Nos 2–3, 2011, 174–88.
19. GUDDAL, I.S., 2010. Archaeology and the construction of identities
in Past and Present Rwanda, unpublished Master thesis in Archaeology,
UiB.
20. HAALAND, G., HAALAND, R. and DEA, D., 2004. Smelting Iron:
caste and its symbolism in south-western Ethiopia.
21. HAALAND, R. and SHINNIE, P., 1985. African Iron working.
Ancient and traditional. Bergen.
22. HAALAND, R., 1972. Neolithic and A group sites.
Universitetsforlaget.
23. HAALAND, R., 2004. Technology, Transformation and Symbolism:
Ethnographic Perspectives on European Iron Working. Norwegian
Archaeological Review, Vol. 37, No. 1.
24. HERBERT, E. W., 1993. Iron, gender and power. Indiana
University Press.
25. HUMPHRIS, J., 2010. An Archaeometallurgical Investigation of
Iron Smelting Traditions in Southern Rwanda, unpublished PhD thesis,
UCL.
 45

26. ILES, E.L., 2011. Reconstructing the Iron Production Technologies

of Western Uganda: reconciling archaeometallurgical and
ethnoarchaeological approaches, unpublished PhD thesis, UCL.
27. JOLLY, D., et al., 1997. Vegetation Dynamics in Central Africa
Since 18,000 yr BP: Pollen Records from the Interlacustrine Highlands
of Burundi, Rwanda and Western Uganda. In Journal of Biogeography,
Vol. 24, No. 4, pp. 495-512.
28. KANIMBA, M. C. & MESAS, T., 2003. Les arts du feu au Rwanda.
Poterie et fer forgé, Editions SEPIA.
29. KANIMBA, M. C. and VAN PEE, L., 2008. Rwanda and its cultural
heritage. Past and Present. Institut des Musées Nationaux du Rwanda.
30. KANIMBA, M.C., 1997. Introduction de l’Agriculture et de
l’Elevage en Afrique Centrale. Cahier Lumière et Société, No. 6 :61-73.
31. KANIMBA, M.C., 1997. Le Peuplement du Territoire Rwandais à
la Lumière des Découvertes Archéologiques. Cahier Lumière et Société,
No. 5:68-99.
32. KOHTAMAKI, M., 2010. An ethnoarchaeological study of Twa
potters in southern Rwanda. In Azania: Archaeological Research in
Africa, Vol. 45, No. 3: 298 — 320.
33. MARET, P. de, 1980. Ceux qui jouent avec le feu: La Place du
forgeron en Afrique central. Africa: the Journal of the International
African Institute, Vol. 50, No. 3. 263-279.
34. MARET, P. de, 1985. Recent Archaeological Research and Dates
from Central Africa. The Journal of African History, Vol. 26, No. 2/3.
129-148.
35. MUGABOWAGAHUNDE, M., 2011. An Ethnoarchaeological
Investigation of Beer Consumption in Rwanda: The social, economic and
political context. Unpublished Master thesis in Archaeology, UiB.
36. NENQUIN, J., 1964. The Magosian Industry of Rutonde, Rwanda.
The South African Archaeological Bulletin, Vol. 19, No. 76: 83-90.
37. NENQUIN, J., 1967. Contribution to the study of prehistoric
cultures of Rwanda and Burundi, Musée Royal de l’Afrique Centrale,
Tervuren, Belgique.
 46

38. NEWBURY, D. S., 1980. Lake Kivu Regional Trade in the

nineteenth century. In Journal des Africanistes, tome 50, fascicule 2. Pp
7-30.
39. NEWBURY, D., 1974. Les campagnes de Rwabugiri: chronologie
et bibliographie (Rwabugiri's War-Campaigns: AChronology and
Bibliography). Cahiers d'Études Africaines, Vol. 14, Cahier 53, Le
Problème de la Domination Étatique au Rwanda Histoire et Économie.
181-191.
40. NTAGWABIRA, A., 2012. Classification of Artifacts Related to Iron
Smelting Activities in Rwanda: Furnaces and Tuyeres. The use of
Archaeological data and Ethnographic museum collections for
interpretation, unpublished Master thesis in Archaeology, University of
Bergen.
41. PHILLIPSON, D. W., 2005. African Archaeology, third edition,
Cambridge University Press.
42. PINGEL, F., 2008. Can Truth Be Negotiated? History Textbook
Revision as a Means to Reconciliation. In Annals of the American
Academy of Political and Social Science, Vol. 617, The Politics of History
in Comparative Perspective, pp. 181-198.
43. POSNANSKY, M., 1968. Bantu Genesis-Archaeological Reflexions.
In The Journal of African History, Vol. 9, No. 1, pp. 1-11.
44. READ, D. R. 2007. Artifact Classification: A Conceptual and
Methodological Approach. Left Coast Press.
45. REID, A. & MACLEAN, R., 1995. Symbolism and the Social
Contexts of Iron Production in Karagwe. In World Archaeology, Vol. 27,
No. 1, Symbolic Aspects of Early Technologies. 144-161.
46. REID, A., 2002. Bananas and the archaeology of Buganda.
47. RENFREW, C. and BAHN, P., 2000. Archaeology: Theories,
Methods and Practice, 3rd ed., Thames & Hudson, London.
48. ROBERTSHAW, P. 1994. Archaeological Survey, Ceramic
Analysis, and State Formation in Western Uganda. The African
Archaeological Review, Vol. 12. 105-131.
 47

49. ROUSE, I., 1960. The classification of Artifacts in Archaeology. In

American Antiquity, vol. 25, No. 3: 313 - 323.
50. SASSOON, H., 1983. Kings, Cattle and Blacksmiths: Royal
insignia and Religious symbolism in the Interlacustrine States. In
Azania, No. 18, p.93-106.
51. SCHMIDT, P. R. and CHILDS, S. T., 1985. Innovation and
Industry during the Early Iron Age in East Africa: The KM2 and KM3
Sites of Northwest Tanzania. The African Archaeological Review, Vol. 3.
53-94.
52. SCHMIDT, P. R. and CHILDS, S. T., 1995. Ancient African Iron
Production. In American Scientist, Vol. 83, No. 6, pp. 524-533.
53. SCHMIDT, P. R., 1997. Archaeological Views on a History of
Landscape Change in East Africa. The Journal of African History, Vol.
38, No. 3. 393-421.
54. SCHMIDT, P.R. 1975. A New Look at Interpretations of the Early
Iron Age in East Africa. History in Africa, Vol. 2. 127-136
55. SCHOENBRUN, D.L., 1993. Cattle Herds and Banana Gardens:
The Historical Geography of the Western Great Lakes Region, ca AD
800-1500. The African Archaeological Review, Vol. 11, Papers in Honour
of Merrick Posnansky.39-72
56. SCHOENBRUN, D.L., 1993. We Are What We Eat: Ancient
Agriculture between the Great Lakes. The Journal of African History, Vol.
34, No. 1. 1-31
57. SCHOENBRUN, D.L., 1994. The Contours of Vegetation Change
and Human Agency in Eastern Africa's Great Lakes Region: ca. 2000 BC
to ca. AD 1000. History in Africa, Vol. 21. 269-302.
58. SOPER, R., 1985. Roulette Decoration on African Pottery:
Technical Considerations, Dating and Distributions. The African
Archaeological Review, Vol. 3. 29-51.
59. STAHL, A. B., 2005. African archaeology, U.k.
60. STEWART, K. A., 1993. Iron Age Ceramic Studies in Great Lakes
Eastern Africa: A Critical and Historiographical Review. In The African
 48

Archaeological Review, Vol. 11, Papers in Honour of Merrick Posnansky,

pp. 21-37.
61. SUTTON, J. E. G., 1993. The Antecedents of the Interlacustrine
Kingdoms. The Journal of African History, Vol. 34, No. 1. 33-64.
62. VAN GRUNDERBEEK, M-C. and ROCHE, E., 1989. Environmental
degradation resulting from human activities. Essay to further recognition
through pollen and charcoal analysis.
63. VAN GRUNDERBEEK, M-C. and ROCHE, E., 2007.
Multidisciplinary Evidence of Mixed Farming during the early Iron Age
in Rwanda and Burundi. In: T. Denham, J. Iriate and L. Vrydaghs (eds.),
Rethinking Agriculture: archaeological and ethnoarchaeological
perspectives. Walnut Creek: Left Coast Press, 299-319.
64. VAN GRUNDERBEEK, M-C., ROCHE, E. & DOUTRELPONT, H.
1983. Le Premier Age du Fer au Rwanda et au Burundi, Archaeologie et
Environment. Brussels: Institut National de Recherché scientifique.
65. VAN GRUNDERBEEK, M-C., ROCHE, E. and DOUTRELEPONT,
H., 2001. Un Type de Font de Fer Associe à la Culture Urewe (Age de
Fer Ancien) au Rwanda et au Burundi. In Mediterranean Archaeology
14: 271-297.
66. VAN NOTEN, F., 1972. Les tombes du roi Cyirima Rujugira et de la
reine-mère Nyirayuhi Kanjogera: Description archéologique. Tervuren :
Musée royal de l'Afrique centrale. Série in 8ê. Sciences humaines 77.
67. VAN NOTEN, F., 1983. Histoire archéologique du Rwanda.
Tervuren.
68. VANSINA, J. 1985. L'homme, les forêts et le passé en Afrique.
Annales. Histoire, Sciences Sociales, 40e Année, No. 6, L'Afrique: Un
Autre Espace Historique. 1307-1334.
69. VANSINA, J., 1984. Western Bantu Expansion. In The Journal of
African History, Vol. 25, No. 2, pp. 129-145.
70. VANSINA, J., 2006. Linguistic Evidence for the Introduction of
Ironworking into Bantu-Speaking Africa. In History in Africa, Vol. 33.
321-361.