An archaeological view of the history and variation of ironworking in southwestern Tanzania

MISSING IMAGE

Material Information

Title:
An archaeological view of the history and variation of ironworking in southwestern Tanzania
Physical Description:
393 leaves, photographs : ; 29 cm.
Language:
English
Creator:
Mapunda, Bertram Baltasar, 1957-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1995.
Bibliography:
Includes bibliographical references (leaves 363-392).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Bertram Baltasar Mapunda.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002054867
notis - AKP2848
oclc - 33497507
System ID:
AA00003202:00001

Full Text



ARCHAEOLOGICAL


VIEW


THE


HISTORY


AND


VARIATION


IRONWORKING


BERTRAM


SOUTHWESTERN


BALTASAR


TANZANIA


MAPUNDA


DISSERTATION


THE


PRESENTED


UNIVERSITY


THE


FLORIDA IN


GRADUATE
PARTIAL


SCHOOL


FULFILLMENT


THE REQUIREMENTS


FOR


THE


DEGREE


DOCTOR


PHILO


UNIVERSITY


SOPHY


FLORIDA


1995






Dedicated


father


, Baltasar Mapunda,


whose


commitment


my education has
Mungu ampumzishe


led me
kwa


to where


amani!


am today


Amina!










ACKNOWLEDGEMENTS


This


work


would


have


been


possible


without


combined


Peter


efforts


express
Schmidt


of many
v sincere


individual


organizations.


appreciation


advice


, guidance,


senior


would


advisor


encouragement,


constructive


qua


criticism


field


ghts


research


that


helped


, interpretation,


and


improve
i writing


dissertation.


would


also


thank


other


members


super


committee


Drs.


Steve n


Feierman


Michae


Moseley


teven


always


Brandt
ready


John


when


Ambrose
needed 1


and


Terry


them


Childs.


guidance,


They
reviews


were
and


comments.


Their time


concern


encouragement


greatly


appreciated


Specia


thanks


owing


ndividua


their


instance


es Salaam


material


, helped


analyses:
me with


. S.


geologica


Kapilima


analysis;


, University


Susan de


France


, University


Florida


, helped


confirm


fauna


spec


and


. Priya


Bendale


and


Richard


Crockett


, University


Florida


, provided


valuable


guidance


during


chemical


analysis.


Appreciation


le University


is also e


of Dar


tended to


es Salaam


colleagues


id the University


friends


if Flnridia










suggestions,


comments,


and


positive


criticisms.


Cordial


gratitude goes to the officials in


research area,


informants,


and crew members for their cooperation and keen interest in the


research


project.


would


also


like to


thank


villagers


their


friendliness,


generosity


and


cooperation


during


research


time.


My gratitude goes beyond measure and words to


my wife,


Victoria


frustration


Mapunda


during


this


sharing
period,


with


pain


especially


physical


and


and,


often


mental


absence


from the


"world"


vowed to share with


her.


remained mentally and physically healthy throughout my


school


because of the


endless tender


love,


care,


support,


and encouragement from my dear parents, Baltasar Mapundat and


Mercyana Mahundi, my sisters,


Kanasia,


Matia


, Regina,


Filomena,


Maryanna


and Aponsiana, and my brother Daudit.


"Musengwili sana!"


also


ndebted


various


institutions


their


financial


support.


Ph.D.


studies


were


supported


scholarship


from


Ford


Foundation


and


this


research


was


made possible


by grants from the


Ford Foundation,


the NORAD,


the Sigma Xi, and the University of Dar


es Salaam.










TABLE


CONTENTS


Dage
ACKNOW LEDGEMENTS ..........................a......................................................


LIST OF TABLES

LIST OF FIGURES


*.a..***aaa****SSI*SS U5*4#1tt1aS**Oalas**ssles...alseelasls...a VI**
aI)))IIIII1)IIIII IILt a a**a*a*5t*Stet*5**** Uta O 35*aSa* e t


ABSTRACT.


CHAPTERS


INTRODUCTION


history


Historiography


Age


Research


in East and Central Africa
Research Justifications..........


I'' 0"I m U'le moDQIOIIIII lllll~ll OqQl mi lO l
...... ... 5 ... a a a a a. ... mi ..... .. .
* .*.S .. ...................... as**** E** .e . a. ....


PHYSICAL


AND


SOCIO-CULTURAL


BACKGROUND


Physical Environment ................ .. ............ ................ ........
Socio-cultural Background ..................... ......... ......... ...... ...


LITERATURE


REVIEW


RON


METALLURGY


EAST


AND CENTRAL AFRICA


Origin


Iron


Meta


urgy


in sub-Saharan


Africa:


Schoolsof Thought..........


ronworking
ronworking


Previous


in East and Centra
Techniques Used


Knowledge


Africa


by the F


ronworking


ipa Neighbors


Techniques


Used by the Fipa ..................,...................... .............. 107










THE PRESENT
STRATEGIES ...


FIELD-RESEARCH:


METHODS


AND


............................. 1 1


Ethnographic Inquiries
Site Survey ..................
Excavations .................


Cataloguing and Quantitative Analyses
Exhibition and Research Assessment .....


S* eta. ..... S
.* .. ..a s a .... te.. .. .


Laboratory Analyses ......


113
116
129
131
132
135


EVIDENCE


FROM


ETHNOGRAPHY


AND


SITE


SURVEY


139


Ironworking Sites ................................
Ore Sources ...........................................
Habitation Sites ...................................


140
173
175
190
190


Source of Potting Clay
Ritual Sites ..................


Microlithic


industry


6 EVIDENCE FROM EXCAVATION


Katukutu Sites ....................................... ...... ............... .................
MalunguSites .........................................................................
Barongo-type Sites ..............................................................
Habitation Sites ............................................................... .......


7 LABORATORY EVIDENCE ......................................................


8 SUMMARY AND CONCLUSIONS....


The
The


Archaeologica


History


Age Culture
Ind Variation


Potential


History


Nkans
Nkans


Ironworking


i District
i District


195

196
231
244
247


265


317
317


Technology


in Nkansi District


Future Research ........ .............. ............................................


347


S ..........










APPENDICES


AN ANNOTATED LIST OF FORMAL


NFORMANTS


350


SASES


GRID


MAP


TANZANIA


SHOWING


LOCATIONS


OF THE RESEARCH LOCALITIES ...........


358


METALLOGRAPHIC AND ELEMENTAL ANALYSES
MATERIALS FROM SOUTHWESTERN TANZANIA


359


REFEREN CES ......................... ............................................................ .......... 363

BIOGRAPHICAL SKIs**ssaaETs ........................................ ....aaass sa......... 393










LIST


TABLES


Page


Table
2.1 Population density of Rukwa Region by District


projected


population


density


of Rukwa


Region


by District by 1995


Some materials collected for intensive analyses


136


Distribution


ironworking


sites


geographical


regions and types .....................


Geographic


distribution


ronworking


furnace


type


Measurements of complete tuyeres


142
1 50


Materials


excavated


from


outside


furnace


Unit 1


Materials
site Hvlk-1


site Hvlk-1


excavated


inside


furnace


unit 2,


208


6.3a


Material


unit 1


excavated outside


furnace


site Hvlk-17


6.3b


Materials


excavated


inside


furnace


unit


ite Hvlk-1 7


............... 2


6.4a


Materials


excavated


outside


furnace


unit 2


6.4b


site Hvlk-17


Materials


unit


6.5a


excavated


site Hvlk-17


Materials


unit 1


6.5b


Material


unit 1


6.6a


excavated


outside


site Hvlk-25


excavated


site Hvlk-25


Materials


excavated


outside


inside


inside


furnace


214


furnace


217


furnace


218


furnace


unit 2


6.6b


Material


site Hvlk-25


excavated


u11e.....*11 ......**1111111()1111111 S.*1SS** 5 5 3 **S** US 6 5 5 5 S


inside


furnace


unit 2


site Hvlk-25










6.7b


Materials


excavated


inside


furnace


unit 1


6.9a


Katukutu
Material


site Hvlk-32


materials excavated from site


excavated


outside


223
226


laim-


furnace


unit 1


6.9b


site Hvlk-39


Materials


excavated


. ..a".. ." ".me..... i"e...... ..... .me...... ... m ...... me. ........


inside


235


furnace


unit 1


6.10
6.11


site Hvlk-39


Materials excavated from site lalm-4


Material


excavated


inside


235
238


furnace


unit 1


6.13
6.14
6.15
6.16
6.17
6.18
6.19
6.20
6.21


site Hxlo-2


Materials excavated from unit
Materials excavated from unit


Material
Material


excavated from unit 3
excavated from unit 1


Materials excavated from unit 2


Material


te Hvlk-35


site


Hvik-3 5


site Hvlk-35
site Hvlk-11
site Hvik-11


.m..... .m.. .
.............


excavated from site Hvlk-19


Materials excavated from unit


Material


excavated from unit 2


Materials excavated from unit 3


Distribution of pottery
stratigraphy ......................


6.22 C14 Dates from excavated


types


sites


site Hvlk-58
site Hvlk-58
site Hvik-58


242
245
245
246
250
251
254
258
258
259


according to


259
264


7.1a
7.1b


Sample
Sample


Dominant


excavated outside the furnaces
excavated inside the furnaces .


microconstituents


in the


slag


286
286


from


katukutu technology....


Dominant


microconstituents


289


slag


from


malungu technology .....


Dominant


microconstituents


290


in the


Barongo-type technology


slag


from


290


Comparative


table of


selected


variables from


three technologies found in Nkansi District


338










LIST


FIGURES


Page


Figure


.1 Rukwa Region.................. ..


Localities


S........... ................. 22


intensively


investigated


during


current


field research ...............


2.1 Kirando eco-zones .....................
2.2 Kala and King'ombe eco-zones
2.3 Trough east of Kalundi ..............
2.4 Kalundi eco-zones .....................


. .. ...... .. i .. . m .... ....
.. .. .. .. ... ... ...... . .... . ... ... **. s.
. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
. .. .. .. .. .. .. .. .. .. .. .. .. .. ..


A geologic map of Rukwa Region


Venn diagram of


site types


Site locations at Kirando


Site locations at Kala and King'ombe
Site locations at Kalundi .....................


119
121
124


Katukutu


furnace


(F1)


from site


Hvlk-


as seen


on the surface...........


Decorated katukutu furnace (F10),


Vitrified piece of furnace (F5),


site Hvlk


site Hvlk-17


143
144
145


a primary


I Pile of tuyeres in
site Hvlk-20, Kirando


Pile of tuyere and furnace


context, site Hvlk-53,


context,


146


rubble


n a secondary


Kirando


Complete tuyeres from the katukutu technology


147
149


unau


(F1)


site


Hxlo-2


Kalund


construction nngs ..................................


Kintenacwe, site lalm-1


, King'ombe


showing


157
158


Lilungu (F1)


tempered with katukutu slag and


tuyeres fragments,


5.10


Mzee


site lalm-1


Xavery Mwanakatwe


, King'ombe


former


digging for iron ore at site Hxlo-4,


5.11
S17


Cave shelter


site Hvlk-


Kalamhn mttrrv


170


smelter)


Kalundi


Kirando


as as as.et s. eta Bess.. S.C. SC.. S S S*S*


175
177
180










5.1 5 Katukutu pottery ............. ................................................. ........
5.1 6 Kirandr'afl1 ~~cYtte ........ ............. .................. ......... .........
5.1 7 Tabwa pottery ..............................................................................


187
188
189


Plan of site Hvlk-1


Kirando


198


Profile


of the


northern


wall


block


unit


site Hvlk-1 ...
Excavation


200


site


Hvlk-1


showing


ordinary


tuyere


ports and hammer stones .............................................
Profile of a katukutu furnace ............................... .


Palinyina,
'Central


furnace F1


pot'


from furnace


202
203
204


site Hvlk-1


vertical


half)


site Hvlk-1 7
'Central pot'


(pot base),


Plan of sites Hvlk-25


furnace F2


site Hvlk-17


Kirando


210
213
215


'Centra


', (hemispherica


bowl)


from


furnace


F6. site Hvlk-2


5 .......................


219


6.10


Furnace


site


Hvlk-25


showing


two


wall


laye


6.11
6.12


Plan of site lalm-1


'Central


220
225


, King'ombe


a hemispherica


(perforated


bowl


6.13
6.14
6.15
6.16
6.17


at the center),
Profile of a lil


unit 5


unoi


site lalm-1


U.......................................................


Excavated kintenawe


A new furnace
Plan of site Hvlk-


(refining furnace)


t around an old one


site


aim-1


site lalm-4


Kirando


Pile of rocks and daub exposed


n unit


228
234
236
239
248


site


Hvlk-11


Kirando ..............................


6.18
6.19


Plan of site Hvlk-58


Some


materials


Kirando


excavated


257


from


site


Hvlk-58


Kirando.


"mkungu"


from unit 2;


nail from unit 3


(c) copper bead from unit 1


Elemental


157-2


from


tk n mmrn~rn\


analysis


site
I f\\lIf


a katukutu bloom


Hvlk-2 5


unit


block


, sample
2 (inside


rS ............................:...........


~ncn knlnlll rl Fl~llm *rA:*IC


371










Elementa


analysis of a


katukutu


slag,


sample


160-3


type
Block


i(dripping)
(outside th,


slag


from site


furnace),


leve


Hvlk-25


Unit


00-20 cm


below datum point.......................
7.3 The Iron-Carbon diagram ............


275


Photomicrograph of a


bloom


, sample


154-1


, showing


coarse polygonal grains of ferrite


280


Photomicrograph of a


slag sample,


ab #


159


showing


magnetite


(bright


angular


particles)


as the dominant phase.
Photomicrograph of a


showing


large


grain


slag sample,


wustite


lab #


(bright,


160-2,
round


particles) as the dominant phase


282


Photomicrograph of


a slag


sample,


ab #


157


showing
Elementa


small


dendrites of wustite


analyst


a Barongo-type


in faya


slag,


283


sample


132


Unit 2


, type
. Block


(flow,


Leve


dense)


slag


from


site


Hvlk-35,


00-20 cm below datum point


298


(a) A piece of a
Kirando. (b) A l


hoe excavated from site Hvlk-35


excavated from


site


Hvlk-58


Kirando.


(c) Remnant of a Fipa "male hoe"


(ise) made


from local iron .....................


303


7.10


Photomicrograph of an


site


Hvlk-35


Kirando


iron hoe


, showing


lab #
ferrite


129-1,
(white


from
grain


and pearlite (dark grains)


304


7.11


Photomicrograph


site
and


Hvlk-


slag


Kirando


nclusions


an iron nai


, showing


dominated


lab #
ferrite


with


284-1
, fine


wustite


from


pearlite,


and


glass


306


7.12


Photomicrograph of an ethnographic iron


ab #


309-


almost


7.13


from
pure


Kalund
ferrite.


, Fipa
Slag


Photomicrograph of an


plateau,
inclusions


iron hoe


consi


lab #


sting


also
129-


found
. from


308


site Hvlk-35


, Kirando, showing weld line


7.14


Photomicrograph of an ethnographic


hoe


lab #


309-


from


Kalund


, Fipa


plateau,


showing


weld lines .......


315


5 ........









Abstract


Dissertation


Presented


the Graduate


Schoo


University


Requirements


Florida


for the


Parti a


Degree


Fulfi


of Doctor


ment


of Philosophy


ARCHAEOLOGICAL


VIEW


THE


HISTORY


AND


VARIATION


IRONWORKING


Bertram


SOUTHWESTERN


Baltasar


TANZANIA


Mapunda


May,


1995


Chairman:


Peter


Schmidt


Major Department:


Anthropology


This


dissertation


presents


archaeological


perspective


socioeconomic


and


cultural


history


southwe


stern


Tanzania
research


during th
discussed


two


this


study


ennia.
was


Archaeologica


conducted


field


Nkansi


District


Rukwa


Region,


southwestern


Tanzania


during


199


1993.


This


research


had


three


major


goals:


establ


archaeological


economic


and


potential
cultural


area;


history


during


reconstruct


Age;


SOCIO-


and


establish


archaeological


history


field


indigenous


each


ronworking


yielded


sites


technology
, including


The


Later


tone


Age


microlithic


industries


Iron


Age


settlement


tes,


ritua


locations


and


iron


-smelting


and


ron-refining


sites.


Much


archaeological


nquiry


focuses


upon


ronworking.


-


-










maluncqu,


and the Bar ngo-type.


These technologies


differ


metallurgical


tuyeres,


attributes


charcoa


such


species,


furnace


chemistry,


morphology,


ritua


slag,


presentation,


and


spatial


and


tempora


distributions.


Radiocarbon


dates


indicate that the katukutu technology is the oldest of the three;


dates


between


1550 and


1800 A.D.


The


malunqu


Barongo-type technologies


date


later than


mid-nineteenth


century.
Triangular


Some settlement
Incised pottery,


sites
daub,


indicated
and fauna


Kalambo


remains


and


however


date to the tenth century A.D.


multivariate


approach is used


in this study to examine


both inter- and


intra-technological


variations


through


attribute


analysis


materials


such


furnace


forms


and


decorative


applications,


tuyeres,


ore, and


ritua


medicines as wel


chemical


and


metallographic


analyses of


blooms,


metal artifacts.


These analyses help


to establish development


ironworking


technology


during


Later


Age


southwestern


Tanzania.


This


ssertation


demonstrates


that


variations


iron


technology

systematic


smelters


southwestern


and


adapt


constant


Tanzania


struggle


to different natural


were


different


and


caused


groups


influences


cultural


II a a "


A, a I a L a -


I


II


A L A Y A A II I


*A AAI IrA A


IAL~


I


I










growth


and


migrations;


and


beliefs


(medicinal


and


cosmological).
pre-industrial
technological


This


African


variations


disproves


conservative


metallurgy


resulted


from


was


static


uncertainty


beliefs


that


that


what the


metallurgists


were


doing.









CHAPTER


INTRODUCTION


This


chapter


is divided


two


sections.


section


outline


and


discuss


history


and


historiography


Age


research


conducted


East


and


Central


Africa


during


100 years.


By doing so,


historiography


used


this


work


into


a hi


torical


context


and


justify


selection.


second


research


section


objectives,


present


and


research


account


universe


selection


and


theoretical


orientation


dissertation


summer


dissertation


coverage.


History


Historioaraphy


Central


Age


Research


East


Africa


ever


scholars


1987


1990


Shaw


1989


Musonda


1990


1990


Maret


1990


Robertshaw


1990b;


Chami


1994)


have


attempted


write


storica


review


The term


"East and Centra


present day Kenya,
.1* -'a *


Uganda,


Africa"


Tanzania


as referred to in this work includes the


Rwanda


* a a *


, Burundi, Eastern Zaire, Zambia,


-C-. P t S -


*~ ~ ~ aI,, ~ Uru j *U- t*a r





paradigms


and


factors


which


influenced


research


and


writing


archaeology


Africa


ongoing


century.


However


most


critical


them


have


review


bias


their


towards


approach
external


indicates


factors.


that


More


emphas


background


placed


theoretical


archaeological


and


researchers.


philosophical
Discussions


often center on


prejudicia


beliefs


such


social


darwinism,


the hamitic myth,


and diffusionism


well


liberal


theories


including


structuralism


post-modernism.


southern


Africa


, processualism,


With


where


political


exception
conflicts


post-processualism


terature


have


and


from


been addressed


archaeology


198


1990;


Kiyaga-Mu


ndwa


199


internal


factors,


including national


politics,


influences


other


disciplines,


and


archaeological


"visibility"


which


view,


have


equally


affected


choice


subject


matter,


research


areas,


interpretation


findings,


and


publication


most


regions
deserve.


Africa


have


not received the


kind of


emphasis they


historiographic


review


berately


bias


discussion


towards


internal


factors.


This


meant


undermine


the role of theories and philosophies of the


western


scholars who have dominated the study of archaeology


n Africa


(in fact those factors are also brought up in the discussion).


objective


is to complement the current popular factors


for the


development of archaeology


East and Centra


Africa


and the


U A S





Three


categories


internal


factors


examined:


national


polit


archaeological


sibi


and


nterdiscip


nary


influence.


National


politics


refers


influences


prioritizing


colonial


research


and


topics,


post-colonial


funding


and


governments in
interpretations.


Archaeologica
archaeologists


ability


sometimes


refers


have


the


towards


materials


preference


that


easily


especially


features


such


architectural


monuments, rock art, and furnaces as well as


availability


written


information.


The


influence


from


other


disciplines


refers


mainly


to history and,


to some extent, ethnography.


understand


needs


bear


role


mind


history
the fact


influencing


that


archaeology


Anglophone


countries


(where most of East and Central Africa


belong)


archaeology was


(and


large


extent


still


considered


handmaiden


history (Hol


1990; Robertshaw


1990b).


critical


analysis


archaeological


research


conducted


East and


Central Africa


based on


criteria


such


research


topics,
reveal


aerna


four


coverage,


broad


interpretation


patterns


referred


and


modes


writing


archaeological


historiographies.


These include


colonia


historiography,


dating


between the


1880s to the


1960s


neo-colonial


historiography,


dating


between


1960s


and


mid-1970s;


nationalist


historiography,


dating


between


mid-19


70s


1980s;


and


post-nationalist


historiography


1990s2


AI f *


4-karen~ k; n ear~ nnr-nd


frnllce nn tho


Irnn Ana


I~~rn





iron


metallurgy,


areas most relevant for this work.


This


helps


identify


and


justify


theoretical


and


methodological


orientation


current


work


discussed


subsequent


sections.


Colonial Historiography (ca.


1880-1960)


The history of archaeological


research


East and


Central


Africa,


and,


ndeed


whole


Africa


began


with


commencement


Throughout


colonial


colonial


period


occupation
archaeological


(Gowlett


1990).


research


East


and


Central


Africa,


with


exception


coast,


concentrated


primarily


Stone


Age


Age3


was


treated as a marginal topic.


Iron Age research along the coast


started


relatively


early


(e.g.,


Kirkman


1952


1954,


1959;


Freeman-Grenville


1958) because researchers were attracted by


architectural


monuments


which,


according


the


evolutionary


scheme


(Morgan


1877


Childe


1951,


195


were


cultural


traits


therefore im

archaeologists


"civilized


portant


know


and


superior
colonial


identify


races."


was


historians


with


their


and


fellow


Caucasians


monuments


who


and


came


Swahi


before


culture


them.


insisting


genera


were


that


products


The two terms "Stone Age" and "Iron Age"


both time and technology.


as used in this work take into account


"Stone Age" refers to the time period in the history when


stone was used to manufacture tools and weapons.


time period ranges from


In East and Central Africa this


million years ago to the middle of this millennium.


The "Iron Age" refers to the time period in the history when iron was used for the


l -r t I I ,. ....... r' i I


"'





foreigners


these


scholars


were


obliging


themselves


their


respective colonial governments whose goal was to demonstrate


nferiority


Africans


thus


bringing


forth


"mora


colonia


justification
1990:103).


n addition to monuments,


which


, again,


the colonial


settlement


the coast had


archaeologists


and


Africa"


written


historians


(Ha


history,
valued


most.


Commenting on the uneven research coverage between the


coast and the interior


, Andrew Roberts, an outstanding


historian


of this time,


writes:


Given the nature of the evidence,


it is easy to understand


preoccupation


historians


[and


archaeologists]


, up


the late


1950s,


with the history of the coast and islands


of Tanzania and other parts of eastern Africa.


beginning


of the


countries visited


written
literate


Christian


era,


[East African]


records of many such


Muslim


civilization


travellers from


coast,


journeys.


linked


and w


From the


literate
ie have


Furthermore,


through


trade


and


migration
on the co;
buildings,


to other parts of the Indian Ocean,


3st.


This left both written


ruins of


which


marked


developed


records and stone


obvious


sites


excavation
history of
formidable


archaeologist.


peoples of


challenge.


This


interior


was


partly


comparison,


presented


more


because


evidence--whether


archaeological,


less easy to locate and exploit (Roberts


or literary--was
1968:v).


Field


research


pertaining


iron


technology was virtually


absent


East


and


Central Africa


during


this


time


period


(and


did not


begin


until


late


1970s


(Schmidt


Avery


1978;


\ /-in 4rJr LAnnAC -mnrl Aunt r *


II falO\)


C' -


A ..,*. rr ~ rj CA~ In; rl~


"113*7Q


wrr trio





there


was


large


corpus


ethnographic


information


subject


collected


European


explorers,


traveller


and


miss


onarles


nineteenth


century


(Wa


1874


Burton


1860)


well


those


ected


both


amateur


and


profe
1894


ona


Wyckaert


ethnographers


1914


Greig


193


subsequent


de Rosemond


decade


1943


(Last


would


therefore


have


been


easy


archaeology


Lists


continue


from


where


Central


previous
I Africa


technology


and


reporters


agged


meta


surprisingly


behind


urgy


research


genera


East


pertaining


compared


and


iron


northern,


western


, and


southern


Africa


where


works


such


as Garland


Bannister


amy


Dart


(1904)
Grande


, Desplagnes


(19


(1907)


, Stanley


name


few


appeared


much early


Why?


understand this


paradox


we need


bear


mind


that


absence


interest


metallurgy


was


confined


this


subject


rather


Age a


whole.


Robert


amented,


Stone


Age,


course


, archaeologists


already


unearthed
elsewhere.


much


important


Age


evidence


received


Olduvai4


title


and


attention


from


archaeology


(Roberts


1968:v).


It seem


meta


that


East


urgica


Central Africa


research


other


backed


regions


kind


had.


timuli


Accounting


should


noted


that


discovery


archaeological


nalapnanthrnnnlnniral citpq at Ifllrviwi ('nrna wac 2rrirdint2


(I 021(01 l7fl nr4


(I p~~p\l


I





for the interest in


northern Africa,


Kense,


example


writes,


"the


focus


investigation


inevitably


concentrated


upon


northern


Africa


since


this


area


was


considered


probable


point


contact


between


the iron-using


Middle


East


and


Mediterranean world and Africa"


to the Middle East,


world's


(Kense
earliest


1983:16).


ironworking


Being close


center


demonstrated


chapter


northern


Africa,


was


hoped,
Under


could provide evidence for the diffusion of iron


this


assumption


West


Africa


was


believed


Africa.


have


bridged the technological diffusion between


northern Africa and


sub-Equatoria


Africa (Shinnie


1967


, 1971).


The early rise of research interest in the


ron Age and


technology


southern


Africa


accounted


political


reasons


(Hall


1984,


1987


, 1990;


Kiyaga-Mulindwa


1993).


The


Iron


Age


research


there was


needed


to demonstrate


that


presence
southern


Black


Africa


was


Africans


recent


(especially


and,


Bantu


therefore,


-speakers)


both


White


Bantu-speakers had an equal right


in occupying the


and


since


both were newcomers.


Kiyaga-Mulindwa (1993


) laments that,


The notion that southern Africa was once an empty land


(Marks


1980),


claim


agents and sympathizers


most
of .


strongly


south


encouraged


Africa


regimes


order to justify white domination in southern Africa,


created


interpretation
literature has


general


problem


past.


. Much


created an impression


presentation


existing


that there were


iron
arriv


using farmers


al


or pastoralists


Setswana-speaking


area


groups


until


across


.I Sr 'a II


- ~ ~ ~ L~ -L -lA a aIA. n A t n


r,, ,,.





place


recently


middle


seventeenth


century


earliest


(Kiyaga-Mu


ndwa


199


386)


Iron


mbabwe


Age
ruin


each


there


was


(considered


also


used


colonial


prove


anthropology


that


sts/


archaeologists


as a key


index


"civi


nation")


were


built


Africans


a "superior


race


" (Caucasians).


fraudulent


interpretation


had


two


objectives:


first


demon


state


inferiority


ived


Africans


southern Africa


or before


ixteenth


second


before


century


prove


that


coming of the


official


Caucasians


Bantu


speakers


date


first


Dutch settlement in the Cape


Coast


of South Africa


198


hort


selection


research


locations


throughout


colonia


period


was


based


archaeological


opposed


documentation


stematic


(referring


survey)


coast


wel


written


Interpretation


and


writing


was


influenced


colonia


deology


philosophy.


terms


inves


spatial


tigated


than


subject
interior


coverage


and


very


coast


research


was


was


better


done


on the


Age


and virtually


none on


ron technology


Neo


-colonia


Historioaqraphy


(1960s


mid-


1970s)


time


research


period
interior


was


marked


Paradigmatica


expansion


this


Iron


was


Age


time


trans


ition


from


colonia


ibera


(neo-colonial)


thoughts.


Outstanding


Age


research


included


Robinson


(1961


1968





Noten (1964), Chapman (1967),


Fagan and Yellen (1968),


Sutton


and Roberts


1968


and Soper (1967a, b,


1971a, b), to name but a


few


This change can be attributed to the emergence of the C14


dating
wanted


technique


expand


well


African


pressure


history


from


historians


beyond--what


who


Posnansky


called,


"the bounds of the colonial


period"


(Posnansky


1965:1).


There


was a


genera


plea during


this time


that


"the


trowe


archaeologist


urgently


needed


begin


where


historian


must perforce


leave off"


(Freeman-Grenville


196


short,


write


there


diachronic


was a


growth


interest


histories--histories


that


among
could


historians


revea


what


had existed before the coming of White people.
We should also bear in mind that this was the decade when


nationalism (struggle


for decolonization)


reached climax in


East


and Centra


Africa.


Therefore historians


, especially


those


who


disagreed


with


colonial


paradigm


(Coupland


1938),


were


free


after


independence


investigate


beyond


colonial


period
political


disprove


this


claim,


sentiment of the time.


least


comply


with


As Neale notes,


was


historian
political
history,
rejected.


this


had


picture
to deal


position
which I


failure


with


Africans


then,


when


was


rationalised


The new nations needed


reversed


their


that


sixt
and


subjection,


African


colonial


was


a new history,


must refute the old, because the old was both wrong and


damaging to African pride (Neale


1985:9).


Given the


fact that very little


Iron


Age


archaeology


had


- -a


I-


* ~~~~ ~ I I.I *L 1 tIr -1 .A... -n lA a.a -- a -Uic *-a n -


"





the historic period and the Stone Age period).


Archaeologists


intensified research on the Iron Age to remedy this problem and,


by the end of the


decade


, books


dealing


specifically


with


Iron Age appeared (e.g., Fagan


1967


Fagan et al.


1969


Robinson


1970,


73).


A critical assessment of the Iron Age


research during


this


period


They


indicates


consisted


monumental


that


almost


architecture


research


topics


were


exclusively


, and


pottery


highly
Bantu


(see


restricted.


migration,
example


Posnansky


1961


Inskeep


196


Fagan


and


van


Noten


1964;


Chapman


1967


Sutton and Roberts


1968; Soper 1967a,


71a,


b, c).


The emphasis on


Bantu


migration was


influenced


by both


historians and ethnographers.


It should


realized that much


historical


and


especially


ethnographic


research h


conducted


East


and


Central


Africa


during


colonial


period,


such


Richards (1939)


, Tew (1950),


Wilson (1958) and many others,


were directly sponsored by colonial


governments.


The purpose


was to understand better the people who were governed.


Bantu-


speakers,


focus


majority


study


this


of the


goal.


indigenous
Emphasis


people,


became


Bantu


research


continued through the


post-colonia


period.


1965


Astor


Foundation founded and sponsored


project


called


"The


Bantu


Stud


Research


Project"


which


dealt


specifically


with


"the


origins and early migrations of the Bantu" (Soper


1971a:1).


This


S r .., .Y I -L a a a:!, -


nl





to pursue


Iron Age research and


publication


Eastern


Africa


since the early


1960s.


Monumental


architecture,


such


as Bigo


Bweyorere


Uganda
(Oliver


attracted


1959;


Shinnie


interest


1960;


milar


Posnansky


that


1968,


along
1969)


coast


, however,


with


different


goals.


Archaeologists


and


historians


became


aware


research


imbalance


between


coast


and


interior


and


they


wanted


rectify


problem


they


provided
colonial


evidence


period.


impediments:


Africa


Unfortunately,


archaeological


achievement


these


invisibility


during


efforts


pre-


met


hamitic


two
myth.


Commenting on


the invisibility,


Connah writes:


The basic


problem


... For instance,


[was]


one


the capital


city


archaeological visibility.


of Buganda,


a state


near Lake


one


constructed


Victoria,


great
totally


was described


capitals


grass,


Ashe]


1889


Africa


wood


was


other


organic


materials


and


moved


frequently,


particularly


death of the Kabaka, the ruler of Buganda.


The sites of


these


large


settlements


[were]


, therefore,


unlikely


have much depth of deposit or structural remains and no


archaeological


investigation of them seems ever to have


been attempted (Connah 1987:214).


Additionally


few


that


were


investigated such


Bigo


were


interpreted (Oliver


1959


Shinnie


1960) as having been built by


Hamites


(Caucasian


pastoralists


from


north)--an


interpretation which


paralleled that


used by their


counterparts


along


coast


explain the


rise


Swahili


culture.


The


rnntrihitinn nf RI~ rk Afrirmnc wc nnrP anain nh~Cijrp.d





Pottery
ubiquity


became


and


important


preservability.


Age


was


research


(and


because


used


answer


questions


that


backed


material


evidence.


Such


questions


included


ethnicity


language,


and


population


distribution


space and time


(Soper


1971b, c).


The


preoccupation


with


three


research


topics


(Bantu


migrations,


monument


architecture


and


pottery)


during


period


perhaps


explains


why


there


was


concern


among


archaeologists


about


what


entailed


Age"


The


term


"Iron


Age


", though


denotes


technology


was


defined


only


reference


time.


other word


, when


talking


about


Age"


little


metallurgy.


concern


The


terms


was


"Iron


given
Age"


presence


and


"ironworking"


were


used


as s


ynonyms.


This


irony


nicely


demonstrated


Grahame


Clark


discu


ssion


Age


East


and


Central


Africa.


writ


East


of the Congo
inhabitants of


identified


ron-working


Kenya,


Ruanda


archaeological


was


being


, Uganda
record


practised


and


Tan


bow


zania


and


globular
the first


pottery
century


vesse


A.D.


south


with


About


Zambez


dimple


same


where


time
was


as early as
it extended
associated


with


working


mbabwe


300, was
the fifth
where it


site


analogou


was


whose


s pottery.
practised


settlement


found stratified


century


was


carried


Broederstroom


Beyond


first


, dating


under the


spread


had


on by


(Clark


Zambez


i iron-


inhabitant


from


stone
beyond


tock-keeping


197


before


A.D.


necropolis. tby
the Limpopo,
farmers at the


-245).


I I' S


II


*


..


1.,


I





desired


free


themselves


from


colonial


paradigm


and


demonstrate


Africans'


achievements


prior


coming


White people,
(diffusionist


but at the same time they used the same colonial


and


racial--hamitic)


paradigms


explain


change


in Africa.


Such contradictions are clear evidence


a colonial


legacy.


Iron Age research,


however, expanded


n the


interior but


was


restricted


Bantu


migration,


pottery


and


architectural


monuments.


technology


was


less


interesting


archaeologists


n both East and Central Africa.


Nationalist Historioaraphy (mid-19


70s to


1980s)


This


time


period


was


characterized


the emergence of


systematic


research on


iron technology.


Examples


include


work


Peter


Schmidt


and


colleagues


hmidt


78b;


1980


1981


hmidt and Avery


78),


van Noten (19


79),


van Grunderbeek and her colleagues (van Grunderbeek


1981


van


Grunderbeek et al.


1983)


interlacustrine


region,


well


as van der Merwe and his colleagues


Merwe and Avery


n central Malawi (Van der


1982).


The research in the interlacustrine


region


concentrated


the early


ronworking.


Chronometric dates from excavated sites


showed that


ron technology in the


interlacustrine


region


began


more


than


2000


years


ago


hmidt


and


Childs


1985:


Grunderbeek


1981


van Grunderbeek et al.


1983).


This


finding


significantly


affected


how


subsequent


researchers


viewed


- S S e. S S





simplistic

spread of


diffusionist


theories


knowledge


used


previously


ironworking


into


explain


region


were


called


knowledge


question.


iron


One


such


metallurgy


theory
Eastern


postulated


and


Centra


that


Africa


diffused


from


Meroe


Shinnie


196


. The


research


interlacustrine


region


challenged


this


egation.


The


archaeological


evidence


showed


that


ironworking


earlier than at Meroe, and that the smelting


there


techniques


began
applied


in the two regions differed significantly.


For example, smelters


Meroe


used


domed,


slag


tapping


furnaces


(Shinnie


1985),


whereas


Northwestern


Tanzania


they


used


shaft,


non-


slag tapping furnaces (Schmidt and Childs 1

The emergence of iron technology as


985)


a research theme


the archaeology of East and Centra


Africa came with a


package


field


methods


techniques


which


included


experimental


observations of smelting and forging


reliance


oral


traditions


processes


(Wembah-Ra


and


1969


significant
Schmidt


1981


Merwe


and


Avery


1987).


Along


with


these


experiments,


several


films


and


video


tapes


were


produced.


Examples include Ironworkincq


Ufipa--a


film documentation of


Fipa


iron-smelting


process


conducted


experimental


setting in Dar es Salaam and organized by the National Museum


Tanzania


1967


(National


Museum


Tanzania


1967


(Wembah-Rashid 1969));


The Tree of Iron a film and video tape


documenting iron smelting


among


Bahaya


northwestern


r~~ -I -


2 a AAA H


r_


r


-- m


I
"


* I






smelting


using domed furnaces and goat-skin


bellows


(Dewey,


., quoted by Herbert and Pole 1988).


Archaeometallurgy


also


became


important


methodolo
the area.


gica


aspect in the archaeology


of iron technology


The works of Peter Schmidt and his colleagues


n the


northwestern


Tanzania


the late


70s and


early


1980s


(Avery and Schmidt 197
should be observed that


mineralogica


Childs


few


petrographic


1986) are good examples.


archaeologists
information


used


include


metallurgica


materials


in their monographs or books in the past as wel


, but


this


information


was


appended


(Clark


1974)


rather


than


integrated into the text or discussion.
The difference between this period and the
was that now archaeologists were bold enough to


previous


nterpret


one
and


write


about


what


achievements.


they
They


considered


argued


that


independent


Africans


too


had


African


been


innovative and that they not only had a longer


history


iron


production


than


previously


thought,


also


their


iron


technology was in


many ways more sophisticated than


that of


contemporary


Europe


(Schmidt


1981).


questioned the traditional diffusionist


iron


technology


ub-Saharan


Additionally,


explanation


Africa


eved


they
origin
have


diffused from the Middle East)


and strongly advocated a


local-


invention


hypothesis (Schmidt and Avery


1978)


(see details in


chapter


--





need to prove that


indigenous


African


iron


metallurgy


longer
colonia


history


than


historians


that


and


claimed


archaeologists


colonial


resulted


and


paying


neo-


more


attention to the Early


ron Age and


neglecting the


later period.


Researchers


started


racing


"the earliest


evidence"


instead


of explaining


processes


involved


historical


development


of the iron technology.


Second, most researchers


one


Other


or two


material


variables,


such


namely


as tuyeres,


tended


smelting


slag,


to concentrate


furnaces


charcoal,


and


quarried


on only
bellows.


ores,


quarries,


metal


objects,


bloom


, bellows,


anvil


, hammers,


etc.


were not given enough emphasis.


It is true that some


these


materials


(e.g.,


bloom


meta


objects,


and


wooden


bellow


find


weather more


quickly


archaeological


than
record


furnaces


and


also


difficult


true


that


some materials


that were


not sufficiently


investigated


such


slag and tuyeres have a comparable


a better


chance


preservation


than


furnaces.


concentrating


just


few


variables


recognize
resulting


Consequently


few


samples


intra-regional


failure


they


have


and


investigators


often


intra-technological


appreciate
sometimes


change


labeled


variations


through
African


time.


iron


technology and the societies as static when


in fact it was their


methodology that was static.


case


mind


following


interpretation


from southern Africa:





site near Broederstroom and that of a number of samples


from six later Iron Age


Transvaal


Swaziland,


sites of the central and western


and


strengthens the concept of


that


remained


basically


period of the south African


AD) (Friede et al.


The same


Botswana.


This


fact


iron-smelting technology


unchanged


during


ron Age (4th to


whole


19th century


1982:47).


authors in another publication argue along the same


line of thought:


In any society,


or needs,
Limpopo,


those


but,
few


required


innovations


most


radical


regions of


technological


answer to


Africa


challenges


south


changes


high-temperature


process seem to have occurred during the


3rd


19th


century


bloomery-smelting


AD).


process


The
was


such


steel-making


Iron


tradition
adapted


Age (ca
African


the


requirements of apparently fairly static societies


relatively


been


imple


sufficient


melting
the


3 technology
needs of th


seems
South


have


African


subsistence farmer (Friede et al. 1984:296).


Post-nationalist


Historioqraphy


(1990s)


Metallurgica
carried forward all


researchers


'good'


elements of the


1990s
1980s.


seem
For


have


example,


investigations


more


focused


and


intensive


1990;


Barndon


1992


Kusimba


1993).


That


to say,


ron technology


does not appear as a side subject as was


generally


past


primary


research


goal


Ethnographic


nquine


former


smelters


continues


emphasized


(Barndon


199


within


framework


"rescue ethnography"


This


is because


"these


technologies


now


extinct


and


few


surviving





former


ironworkers


elderly"


(Childs


and


Killick


1993:318).


After a decade or so all will disappear.


also


new


directions:


'earliest


evidence


syndrome


progressively


' seems to have


investigated


lost power.


(Killick


The


1990


Later


Barndon


Age
199


Kusimba


1993).


scrutinized


Multiple
through


technological


attribute


and symbolic variab


analyst


wel


microscopic


stud


(Killick


1990;


1990,


1991a,


Kusimba


1993


Finally,


appea


researchers


dealing


with


indigenous


African


metallurgy


direct


efforts


towards


exposing


and


understanding


ntra-regional


and


intra-technological


variations.


This is


because


these


variations are


key towards a


better


understanding


and


appreciation


innovative


African


metallurgists


their


constant


struggle


meet


the ever-growing demands of iron


both symbolic


and


utilitarian.


Therefore


central


thesis


this


dissertation


that


technological


and


symbolic variations


such


those


presented


this


work


metallurgists


indications


were doing,


as some


uncertainty


scholars (Cline


what


1937) put it,


but


rather


clear


proofs


constant


experimentation


conducted


by the


metallurgists


efforts to


improve


both the


quantity


and


quality


iron


which


was


highly


valued


walks


of life:


economically,


politically


religiously.





Research


Justifications


Research Scope


The


field


research


upon


which


this


ssertation


based


involved


ethnographic


investigation


archaeological


survey, and 3) excavation.


Ethnographic studies were conducted


collect


information


pertaining


history


and


socioeconomic aspects of the contemporary inhabitants


research


area.


This


information was expected to contribute


the reconstruction of the later history of the region as we


offer


night


that


could


assist


nterpreting


the


archaeological
through the u


findings


from


controlled


both


site


analogy.


survey


and


excavations


Archaeological


survey


was aimed


locating occurrences


of archaeological


materials,


studying


their


archaeological


spatia


land


distribution


patterns.


Sites


and


located


examining


during


survey were


scrutinized


for their


potential


contribution


goals


project.


Excavations


were


conducted


expose


subsurface


materials in order to understand their use patterns,


deposition


well


sequence,


obtain


cultural


radiometrically


and


geological


datable


stratigraphy


material


specially


charcoal).


Research


Universe


The


research


was


conducted


Nkansi


District,


Rukwa


n anaSaI. aaan". 1 -nS- ,r:,II--tn n*r


n* a r'


r,,,,,:, /tr,


TC\A []IIIllrrA


I I t




research


universe


because


region


was


found


have


rich


ethnographic
ironworking


information


continued


pertaining


region


ironworking.


unti


relatively


indigenous
recently:


1930


in most


parts


and


1950s


Katumba-A


mio


south


Sumbawanga


(Wright


198


1985


Wembah


-Rashid


1969).


The


presence


relics


ronworking


(e.g.,


furnaces,


tuyeres,


and


slag)


on the


landscape


was


also


known.


Furthermore,


ingu


stic


(Ehret


1991)


archaeological


a site


evidence


located


, especla


from


south


Kalambo


current


(Clark


research


area,


indicated


was


very


that


ikely


region


inhabited


along


shore


throughout


Lake


Age.


Tanganyika
Ecologically,


area


have


seemed


been


similar


found


other


East


ocalities


Central


where


Africa


Early
such


Age


as along


Lake


Victoria


(Leakey


1948;


Chapman


1967


hmidt


78b


, 1980),


Kivu (Hiernaux and Maquet


1956)


, Nyasa


(Robinson


Mapunda


1991b


Mapunda


and


Burg


1991)


and


Tanganyika


(Clark


197


The


field-


research


was


focu


sed


four


localit


Kirandos


and


Kala


along


hore,


King'ombe


Fipa


escarpment,


and


Kalund


on the


Fipa


plateau


(Fig.


Kirando


and


Kala


were


selected


ecologica


and


ogistica


grounds.


Both


were


protected


natural


harbors


with


relatively


wide


drained


hinterland


dissected


perennia


rivers.


Kirando is a collective term for a


arge shore plain which encompasses several


*5* .


..






The


fact


that


neighboring


areas


lacked


shore


plains


suggested


attracting


Furthermore


both


that


human


both


Kirando


occupation


places


Kala


early


were


easi


high


history


ikelihood


as they


accessed


road


today.
from


Namanyere
Headquarters


and


Sumbawanga


ship


strict


and


and


boat


Regiona


Kalund


and


King'ombe
informants


were


once


selected


was


field


erved


that


with


shore


help


oca


ties


(Kirando


and


Kala)


were


biased


toward


katukutu


and


"Barongo" type


malunqu


technologies


technology6


with


Thus


almost


selection


representation


localitie


Fipa


escarpment


(King'ombe)


Fipa


plateau


(Kalundi)


were


meant


bring


a balance


in the


research


coverage


well


provide


an opportunity


to compare


three


technologies.


b Both katukutu and malungu terms have been borrowed from Fipa language.
has its root in amalunqu (singular icilunqu), meaning (tall) iron furnaces.


Malunqu
Katukutu


roots from katukutu a word for short or dwarf (due to their


size


vis-a-vis


malunqu)


-- ~ -










.1 -r


6 -


* a S*


.0 *I 0
S.


Nkansi


ZA


- Ba


TABORA


REGION


8I.


te~


Mpanda


Mpanda


District "


Namanyere
ai*-


Symbawanga


ZAMB


Sumbawanga


0,


District


r-


MBEYA


REGION


District


km 100


100 km
I a


I L _


'


r,


v









To M panda


asolo


rl S


TANZANIA


NAMANYERE


a~I


Chala


To Lake Rukwa


Nkundi


Kalundi


mpembe


SUMBAWANGA


King'ombe


* r"\


KEY


i Regional HQ.
* District HQ


. Village


Localities Intensively
Investigated
All Weather Roads
Dry Weather Roads


40 km


Chapota


Kasa


Iu \


Kalambo Fafis


ZAMBIA


* p a


-'S


a *


A


Ir'


'l


I I


..







Research


Obiectives


This


research


had


three


major


goals:


establish


archaeological


potential


research


area;


reconstruct


diachronic


socio-economic


and


cultural


history;


and


establ


the history and development of


ronworking technology in Nkansi


District in


particular, and southwestern


Tanzania


n general.


Archaeological


Potentia


relatively rich in ethnography (Greig


Nkans


193


District.


Wise


1958:


Although
Wembah-


Rashid


1969; Willis


1966


, 1968,


1981


; Wright


1982


, 1985; and


Barndon


prior


1992),


this


Nkansi


research


District


was


project.


archaeologically


few


archaeological


unexplored


research


projects,


however,


had been conducted in the neighboring area.


These


include Brian Fagan'


and John


Yellen' s


1966


excavations


at the


Ivuna


salt-processing


site,


southwest


Lake


Rukwa (Fagan and Yellen


the early 19(
Zambia (Clark


test


1968)


50s at Kalambo


1969


excavations


Desmond Clark'


excavations


, along the border of Tanzania


, 1974), and Pamela Willoughby'


along


Lake


Rukwa


trough


and


survey and
(Willoughby


1990).


With


exception of Willoughby'


surveys,


others


were


shore


spatia


Lake


restrictive


Tanganyika


, leaving


north


Fipa


Kalambo


plateau


Falls


and


completely


unexplored.


Diachronic


Culture


History.


Although


southeastern


shore


Lake


t h nn nr nhi


Tanganyika
I m m n I i1


and


?tn rmr '


hinterland


Sn r n rmni- in


rich


I n o.


"nn


*





1992),


questions related to the earlier


history


modern


and


recent


Fipa


socioculture


expressed,


example,


ronworking


sociocultural


technology,


variable


settlement


have


style
only


farming


been


and


other


superficially


investigated by these studies.


Additionally,


there is a fifteen-


century
around


information gap between the


beginning


first


linguistic
millennium


evidence


A.D


from


and


ethnographic and historical evidence beginning from


1700 A.D.


This work


provides


socio-economic


culture


history


Nkansi


District


particular


and


southwestern


Tanzania


genera


from


archaeological


Early


, oral,


and


Age


literary


present.


information


uses


explain


peopling of the area, settlement,


socio-economic


contribution


economy,


iron technology and


society


placing the culture and traditions of southwestern


region


Tanzania


(East and Central Africa) and continental perspective.


History


and


Development


ronworkinq.


This


study


examines the


historical development of ironworking


technology


using


diachronic and multivariate


and interprets intra-regional


and


approach.


intra-technological


examines
variations


and looks at ironworking within


research


Nkansi


District


cultural


revealed


three


settings.
different


The


field


types


bloomery


The


first


technologies:
(katukutu)


katukutu, malungu and "Barongo" type.


involved


single


process


bloom


7 Rlnnmprv tprhnnlnnv refers to "a variety of iron smeltina process in which ore





production in short


globular,


natural-draft


furnaces


with


multiple


tuyere


ports;


the second


(malungu)


involved


the production of bloom in two stage


each of which used different


smelting


types of furnaces.


and


refining,


melting


was


conducted


ports and a


-3.5


natural-draft


shaft


mode


furnaces


combustion,


with


while


tuyere
refining


was carried out in small


(0.3-4


furnaces with


three


tuyere


ports


worked


bellows.


The


last


("Barongo")


technology


employed
mounds.


shaft


furnaces


made


slabs


Similar to the malungu technology,


from


Barongo


termite


type


technology


involved


smelting


and


refining


separate


processes.


addition


technological


attributes


three


technologies


varied


spatial


and


temporal


distribution.


The


katukutu technology dated between


1500 and 1800 A.D.


and was


located


along


hore


Lake


Tanganyika


and


escarpment;


malungu technology dated between


1850


1930 A.D. and was located mainly on the Fipa plateau and on the


escarpment;


and


Barongo


type


technology


dated


nineteenth century and was located along the shore.


This


study


examines


each


three


technologies


separately


establish


factors


that


favored


their


historical


development,
technological
sociocultural,


sociocultura


expression.


and


chronological


contributions,


also


and


unravels


relationships


ritual


and


technological


found


between


these


trchnnlnoai s.





Theoretical


Orientation


have argued elsewhere (Mapunda


1991a) that the role


any


academic


discipline,


physical


science


behavioral


science,


won't


complete


without


taking


into


account


potential


contribution


towards


solving


development


faced by the local populations in the research location.


problems
Arguing


along the same line of thought,


Sinclair


et al. (1993) state that


"archaeology
transmission


one


means


historical


collection


information


interpretation


specific


social


contexts and not merely


way


accumulating


knowledge


past human behaviour"


(Sinclair et al.


1993


428; see also


Shanks


and Tilley


1987).


achieve


the social goals of archaeology


researcher


needs to answer several fundamental


her or his project.


questions in the course of


These include,


What is the object of study?


For whom?


What emphasis


should be placed on which set of meanings embedded in


material


-mass


teams?
towards


culture? How should the work be implemented-


involvement


Should
theoretical


or small
research


advances


, highly


trained


oriented


should


technical


academically


resources


redirected to convey an awareness of the existence


characteristics of the precolonial past to


workers?


(Sinclair


... peasants and


et al. 1993:428).


These


important


questions


because


they


guide


one


towards


information


determining


presentation


methodology


(writing).


interpretation,


However,


and


case


Africa,


also important to distinguish


between


foreign





above.


example,


an i


indigenous


African


researcher


influenced


historic


(e.g.,


need


correct


the


colonial


historiography)


resources


most


economic


countries


priority


have)


zation


more


(due


kely


meager
socially


oriented


toward


justifying


her/his


course


and


buying


acceptance


into


society


necessa ri


fundamental


concentration


which


belongs.


problem


istemology


foreign


(theory


and


These


researcher


method)


factory


itself.


This


tudy


was


conceived


conducted


and


presented


here


with


for the


whom


purpose
belong,


reconstructing


therefore


, my


tory


own


history


people


both


with


passage


past


soCIO


time


tical


and


through


goals.


purposefu
searching fc


distortion


suit


information


research


goals mentioned


above


were


used


as genera


guide


nes.


Both


formal


and


informa


opportunities


were


opened


loca


people


research


area


influence


methodology


interpretation


manner


which


we al


eved


represented


'right'


history


Emphasis


was


is given


question


most


important


history


of the


indigenous


people.


Economically


this


research


(and


other


rese


arch


projects


that deal


with


pre-industria


Iron


technology


Africa)


potential
especially
industrial


encourage


rura


sector


areas


Loca


development


, supplements


iron


smith


ron-s


unstable


produce


mithing


which


commercial


and


maintain





rural


income


residents


(see


also


Wright


(1985)


and


Kapinga (1990) for this view).


Dissertation


Coverage


The


dissertation


divided


into


eight


chapters


The


second chapter provides a description of the


physical and social


background of the


research area.


This


meant


provide


spatial context for the study, as we


as to set out a background


upon which a


wider,


trans-regiona


techno-cultura


comparison


made.


providing


spatial


and


socio-cultural


background,


the chapter justifies field


methods


and techniques


applied during the field project.


Chapter


development


presents a


iron


literature


technology


review


globa


historical


, continental,


region
studies


and


loca


conducted


levels.


within


Empha


close


however,


Ufipa


with


given
view


isolating


some


technological


socio-cultural,


and


environmental factors that might have


influenced some patterns


of adaptation in Ufipa,


such as the spread of knowledge


iron


metallurgy


population dynamics, and settlement.


The


field


methods


and


techniques


employed


during


ethnographic studies,


ite survey,


excavations,


and


microscopic


analy


are dealt with


n the fourth chapter.


This chapter also


accounts


selection


archaeological


methods


and


techniques.


Sampling


strategic


research


ocalities,


areas


a S


I





The


following


three


chapters


and


more


focused in their coverage.


ironworking


Nkans


They


dea


District.


principally
Chapter


with


evidence


focuses


ethnographic evidence and


surface occurrence of archaeological


materials.


Chapter


provides


evidence


from excavations,


chapter


7 presents evidence from metallographic and element


analy


of iron ore


, slag,


bloom, and metal artifacts.


The last


objective (
questions
clarification
include arc
production,


chapter


research


addressed


and


provides


summary


project.


previous


conclusions


:haeological


and


based


revisits


chapters


controversial


potential


interpretation


peopling


and


Issues.


important


provides


These


, history


technological


iron


variations.


Finally,


offers suggestions for future


research


based on


success and failures of this research


project.









CHAPTER


PHYSICAL


AND


SOCIO-CULTURAL


BACKGROUND


this


chapter


present


an overview


physiographic,


climatic


and


biological


environment


SOCIO-


cultural


background


research


area.


The


here


provide

discuss,


context


in the


needed


subsequent


better


understand


materials


chapters.


Physical


Environment


Location


The


field


southwestern


research


Tan


zania.


was


The


conducted


Rukwa


Region


Rukwa


ocated


Region,

between


latitudes


north


5 and


region


9 south


bound


and longitude


Kigoma


30 and


and Tabora


east.


Region


east by Mbeya Region,


to the south by


Zambia


west


hares


Lake


Tanganyika


with


Zaire


(Fig.


The


region


comprised


three


districts:


Sumbawanga


south


Nkansi


center


Mpanda


in the


north.


these


district


this


field


research


was


concentrated


mostly


Nkans


District.


- -


JI a -. a l


fl^ I n.. I


'n at ,-,


r n


mnrnr n n n


.4C,;


F_ LI-


'---- -


I


al. I., A






Physiography


Physiographically,


Nkansi


District


divided


into


three


major zones:


the Lake


Tanganyika shore,


the Fipa


plateau,


escarpment


between


lake


and


plateau.


Lake


Tanganyika,


670 km


ong and on the average 50 km wide,


belongs


to a


series of five


lakes


(Rukwa,


Kivu


Edward


Amin),


Albert (Mobutu)) formed in the western branch of the Great Rift


System.
1470 m,


meters


Tanganyika


above


level


the second deepest


and


with


lake


a depth


world--


exceeded only by Lake Baika


(Grove


1986; Ntakimazi 1992


The water level,


however,


been


changing


throughout


its six


million


year-history


Pleistocene and the Holocene,


(Michel


et al.


1992).


for example, Lake


During the
Tanganyika


late


well


other


Habeyan and


lakes


Hecky


East


1987


Africa


Davison


(Street
1991),


and


Grove


witnessed


least


seven high water levels and six


corresponding


water


levels.


High water levels have been recorded around


35,000 B.P. (with a


rise of


around


100


, 000 B.P.,


11,000 B.P.,


2500


1,000, 500 B.P.


and in the


1980s.


Low water levels have been


recorded for around


000 B.P


. (with a record of about 600 m


below the current level),


15,000 B.P., 5,000 B.P.


1.500 B.P.


750


and


0 B.P.


(Livingstone


1965


Street


and


Grove


1976;


Haberyan and Hecky


1987:


Scholz and Rosendah


1988;


Davison


1991


Ntakimazi


related to


1992)


changes


These


climate.


variations


although


were


tectonic


primarily


factors


uL *J V.J


B. P.


.





fluctuations


Lake


Tanganyika


were


low


actual


measurement


, ranging


between n


However


, they


great


impact


land-use


pattern


along


hore


(D.D.Y


. 1957:


detail


chapters


The


natural


shore


harbors.


genera
bound


straight


precipitou


and


carps


very


few


most


places


which


often


straight


from


water.


The


scarps


interrupted


formed


rivers


few


that


scarp


places,


originate


however


, by


from


Good


gently
plateau


examp


sloping


or by
such


plains


streams


plain


Nkans


nten


vely


trict


include


investigated


Kirando1


during


Kala


current


(Fig.
field


which


were


research.


Kirando


plain


incorporates


four


rivers:


from


north


south they are


Malembwe


, Kavunja,


Mkamba


Luafi


(Fig.


2.1).


The


Kavunja


and


Luaf


rivers


perennia


whereas


Malembwe


and


Mkamba


seasonal.


They


become


completely


during


two


-to-three


months


season


(August-October).


The


plain


extend


about


nland


before


reaches


900


m contour which


marks


base


series


beginning


escarpment.


These


include:


Chongo and Wangubo to


north


, Chabya


Mosi-wa-


Mpepo to the east, and Nkanga to the south (Fig.


Kirando plain comprises


kVmwanda


of several settlement clusters


Itete.Mabatini. Katete. MData.


including:


Mtakuja,


Katonaolo, Jiwenikamba, Masolo and










































= Lacustrine mash
= Seasonal Swamp


= Hills (900 masl)


=


Shore Plain


/// = Village


km 5 0 5 11
I I I I





lake


Nvuna


shore


, Lupita,


there


Nkondwe


seven


islands,


Kamanda


Kerenge,


Mwila.


They


range


from


less


than


km2


(e.g.,


Nkondwe


and


Kamanda) to about


2.5 km2


(e.g.


Ulwile).


The


islands command a


strategic


Only


position


larger


both


island


fishing


, Kerenge


and


and


defense


Ulw


past).


permanently


nhabited


because


they


have


deep,


cultivable


The


smaller


rocky


Kala


is a


natural


used


harbor


mainly


as season


, protected


two


fishing
island


camps.


Kala


Mikongolo (Fig.


. Unlike


Kirando


which


a wide


hore


plain


Kala


a narrow


shore


plain


about


one


wide


and


crossed


one


perennia


river


Mwiu.


The


escarpment


is generally


precipitous.


some


place


however


forms


relatively


wide


, up


five


terraces


that


accommodate


small


clusters


settlements.


One


example


King'ombe,
east of Kala


ocality


(Fig.


current


research


located


2.2).


In some literature the names Kerenge and Ulwile are prefixed by "Manda"


"Manda Kerenge" or "Manda Ulwile". This prefix
work because the word "manda" means "island"


of one of the islands--means


"a small


island").


{ has been omitted throughout this


in Kifipa (e.g., Kamanda


Thus saying


-- a name


"Manda Kerenge


island" is not only redundant but also does not make sense since the other islands























































*= Hills and Escarpment
(900 mast)
= Shore Plain


= Basin


= Village


= River plain


m 1000
1


7 km
I


La;' I I I r






















































= Plain land


= Basin

= Village


= Granitic Ridge


km 5
I


10km
I


I





The Fipa plateau, averaging 2000 m above see level (masi),


and 80 km wide, is concave in shape.


The center is composed of


open rolling country comprised of ridges and basins (Fig.


2.3 and


2.4),


most of which retain water throughout the year.


On the


eastern side


, the Fipa plateau falls to the Lake Rukwa valley at


900 meters above sea level.


Another


important


relation to this research


physiographic
is termitaries.


feature,


especia


They (averaging


m in


height and 9 m in basa


diameter) are found


n large numbers


three


physiographic


regions.


The


highest


concentration


recorded
hectare.


during
The cc


this


common


field


research


species


include


termitaries


Macrotermes


falciger


found mainly along the


hore and M. natalensis, prevalent on the


Fipa plateau.


In Ufipa (a


in other parts of


Africa


(Childs


1986;


Killick


1990


Roberts


199


Agorsah


1994)),


termitaries


associated


with


various


symbolic


roles


related


hea


magic,


and beliefs as will be demonstrated in chapters


Recent studies have shown that Macrotermes


falcicqer are


and 6.

major


agents


terrestrial


sedimentation


especially


African


savanna


and


rain-forest


areas,


both


within


African


Rift


System and

conducted i


n epeirogenic basins"


Karonga


(Crossley


section


1986)


Malaw


. In a study


Rift


Valley,


Crossley


has observed that a


vertica


transfer


of sediment


Macrotermes


falciger, at an estimated rate of 0.24 mm / yr,


resulted


n a "homogeneous red clayey sand


heets, averaging


r a a .~: 1 -an. a
Si .a
a a a a A. -


rAA(ILY: Ylf


I'


h I ~III


hhn


----


NI





have


been


observed


along


northeastern


hore


Lake


Tanganyika


that


(around


Macroterme


Kigoma)

falciqer


And


have


chapters


been


burying


5 and


report


archaeological


sites


along the


Kirando


plain at a


rate of 0.8


mm


Geomorpholoqv and Geology


Nkan

confluence


district


and


eastern


the

and


whole

western


Ufipa


branch


Great


Rift


stem


which


tota


runs


over


7000


from


lower


Zambez


i basin


Mozambique


Jordan


eastern


sometimes


Mediterranean


referred


. This


as the


area


confluence,


"Nyasa-Tanganyika


corridor"4


been


ubjected


great


tectonic


upheava


Consequently,


remarkable


including

volcanic


(Masoko

mountain


cones


and


(e.g


geomorphological


(e.g.,


(e.g.


Wentze

.. the Fi


Rungwe

Heckman


pa plateau).


features


Nyasa,


Mbeya


Mbeya


Geo


have


Rukwa


been


and


Region)


Region)


-historic


formed


Tanganyika)


crater


and


studies


block


indicate


term "Ufipa" is derived from ifipa which means "


esca


rpments" in the


local


language,


iciFipa.


1981).


Literally,


Ufipa means


"land of escarpments"


referring to the escarpments that surround the Fipa plateau.


In this work the word


"Ufipa"


is used synonymously with


"Fipaland"


to mean


not strictly the


inhabited by the Fipa, but rather the shore and the plateau in both Sumbawanga and


Nkansi Districts.


This, therefore, includes Ulungu.


"Corridor" because the area


ooks like a pass-way on maps between


akes Nyasa


to the south


countries,


S


, and Tanganyika


Tanzania


Malaw


I I


to the north.


Zambia


- a a S


Today this area fa


(Fig.


Some


* S


within three


historians


ar 1 ..q .n n a ~ a n t1 In-* i ii in. xLrn aM I 111 L 5 -


/





that


northern part


of the


system,


which


includes the


Eastern


Branch from
Cenozoic rift


north


faulting.


Mt.

The


Rungwe


relief


the

the


outcome


southernmost


part


including


Lakes


Nyasa,


Tanganyika


and


Rukwa


associa


with


much


older


dislocation


which


took


place


during


Karroo


geologic


formations


early


Mesozoic


, 270-160


years ago (Grove


1986).


The


floors


"corridor"


region,


well


as I


ake


Nyasa


, Tanganyika


and


Rukwa


made


Karroo


stratae


overlaid


with


sediments


derived


ater


during


Late


Meso


ZOIC


Cenozoic


eras


from


Karroo


rocks


upstanding


when


faulting


took place.


proto


Grove


-Zambez


notes


that


, proto-


during


Limpopo


Jurassic


and


ndian


period,


Ocean


when


were


formed


Malaw


"the


[Nyasa]


tributarie


Rukwa


proto


Tanganyika


ambez


valley


excavated


and


trough


Cretaceous


and


Cenozoic


sediments


accumulated


and


remain"


(Grove


1986:9).


Fault


dislocations


corresponding


some


degree


with


ancient


locations


have


continued


into


sent.


result


rifts


have


deepened,


floors


Lakes


Tanganyika


Nyasa


have


subdivided


into


distinctive


basins


--each


about


20 km


long


wide--


and


Lake


Rukwa


trough


been


rejuvenated


(Rosendah


et al.


1986


Grove


1986).


Seismic


movement


common


in the


area


The


western


side


"corridor"


, for example,


rank


second


seismic


inten


r, _.A -


Ar


S


- -I.


I__I


I I


=


1





Geologically,


Nkansi


District


and


Rukwa


Region


relatively


complex,


comprised of


at least six


orogenic


systems


(Fig.
both


2.5).


The escarpments and the shore of Lake


Nkansi


plutonic
Mobilized


and


rocks


Sumbawanga


(intrusive


Granite


Districts


granite),


2.5).


consist


sometimes


The


rocks


Tanganyika


mainly
known


probably


contemporaneous


with


Nyanzian


system which dates


over


500


million years.


Most


of the


plateau


and


area


north


Lake


Rukwa


belong to the Ubendian rock formation,


high-grade


, strongly-folded


metamorphic


consisting


rocks


and


complex,
intrusive


granites
volcanic


2.5).


in origin


(Temple


The


197


rocks


mostly


The metamorphic


pelitic


and


rocks


composed


mainly of non-porphyritic gneiss which,


according


Parkinson (1932),


can further be


subdivided


into


aplitic


gneiss


and "Rukwa gneiss"


percentage


percentage
plagioclase


The former (aplitic gneiss) is older and has


ferro-magnesian


microcline


feldspars


and


minerals


potassium


pink


garnet.


The


and


feldspar)


latter


high
acid


("Rukwa


Gneiss")


biotite,


high


percentage


phenocrysts


ferro-magnesian


feldspar,


rare


minerals,


occurrences


microcline,


and


accessory


mineral


which


include


hornblende,


epidote,


and


apatite.


The


rock


was


folded


and


largely


metamorphosed around


2100-1950 million years ago.


The


west-central


part


Mpanda


District


consists


thin


(fIpc


than


.I l. I


trm mnir~a\


M~nrl


V f~nllP nAln.A3n not n


mt-I,


E II





largely


argillaceous


and


been


mildly


metamorphosed


phyllites,
arenaceous


argillites,
formations


and


low-grade


have


been


sericitic


changed


schists,
quartzite.


while
The


system was folded and metamorphosed some


1300-1100


million


years ago though post-tectonic events may be as young as 850


million years (Temple


1972).


The Bukoban rock system (4,


in Fig.


2.5) occupies much of


northwestern


part of the


region through the center of Mpanda


District and the southwestern corner of the region.


characterized


conglomerates,


thick-bedded


The system
sandstones,


red shales,


quartzite, dolomitic limestones, and extensive flows


of basalt.
folded, cc


It is predominantly terrestrial and volcanic,


considerably faulted


ghtly-


and virtually unmetamorphosed.


The


Kirando plain


central-western


Nkansi


District,


and


north-eastern


part


of the


District


consist


Karroo


rocks (5,


principally


n Fig.


2.5), dated to


terrestrial


270-160 million years.


sediments,


consisting


These are


sandstones,


conglomerates,
limestone. Ge


tillites


I o-historically,


Karroo


and
rocks


gray r
were


mudstone,


laid


down


downfaulted or downwarped areas and have been preserved.


The


Karoo


series,


though


generally


easily


eroded


have


been


preserved because they occupy structural basins (e.g.


Kirando).


The


Karroo


formations


oldest


containing


uncontested


plant and animal remains (Survey Division


1956;


Temple


197





i
N 3C


.: :a e c mm
"\



'4..


\,%\
"4


\-s


^i


___ cr *


]31





















i4 Hi
: t

Si 1***5 ***
..a .
cm .cvea ...+me

a a a I m a e a a m a

;:iiiiiiii i::: 4"


" .-


?.?<






aa


m: S
a.? m..~














*** *







r"' **** *r
*** -
^**************
I***************

1~~* *** ******k^^




'* .:.**



.:..:... ...
A"**k****se******^












.a.. mu..t: ..t.....
** : a.:::: mc
S *****a m *a*s? Aem
f-^^^ .^ F r^ "* *- fc *-*_ ^* ** 1 S ~ *-***







eKa .S *m* e** a







lin. S aca m ma a m
C. C1 8 t1.
11!!:a:.:a1a. a









m :i


C..ea~.. .a..a
*xS0SB~i!*S ****M~g*S* *:* S'S.Ai:* 2. :* *! :8 !* t* Smm!!
sis^^siisi~si^Si^^^ iKifii








;::::;::::::::tm* m.% m m;:: ::!!:::



fe 5 :. a C.::
k* **t *^C **f~^ U^f^P^'^FpJr^ ^kw*** *^*- w

ijs~si~sii~~ii~i ||J| ^K/^l11iS^^


^*kA ^ MrSl^.^ ^ ^ ^ f B fj k fe-k J^ ^J k^ -f1


S.......e.........

:.?. m.e. ..a... -*


KEY


1 | Plutonic Rock


4 3 Bukoban


N. Fault lines


2 1 Ubendean

3 E Karaawe-Ankolean


5 lKarroo

6 0 Cretaceous and Ouatemarv Sediments


1!-*^
a: 9^ v
i"S*S"S a
!"S*S*Sc
e. oI .CC a. m CSam
Caj:::s*#::: ::!!1:!*jj.;'*..




mmmCf- .:mas ~~ -
p* :. 1..: mm*ma^m. C-f~ t? m~:m mmta*222 ? a
ms. c..m&a_....t:...:.
Ij*********m**
a rrrr u_ a rrrr.lrr b _* _- 70

c*-*m ftS* a** _
k^ k *** kB -in...*
**_*__*jfa**_*ma..
^* B- _* *_ _* m2
&*_* **** **** -*** *- ::
*****&***:.-- :. .:




_*_**&**_**_r*_a*.K

*ac .. ..a ...m^Bm S****** ^
cm a.. e .C a a me.. *** &* -F^
.m *m..m a a*.mAc..a.* IA^^k


- 80-

















90 -


J2


J~O~i~





The Lake Rukwa trough and some parts of the


plateau


west


Sumbawanga


terrestrial


depos


consist


of various


young
; kinds


(Cretaceous


2.5).


Quaternary)


It should


noted


that


alluvium


depo


sition


along


shore


Lake


Tanganyika


amount


750-


; relatively
1000 mm),


because


mited


catchment


area


rainfall


affluent


rivers


(60


maximum


ength)


and


hard


parent


rocks--


metamorphic


gneou


rocks


(Yuretich


198


Grove


1986


Haberyan and Hecky


198


The


trough


on the


plateau


formed


aterite,


some


which


rich


ron.


number


centuries


, they


have served


as sources


bog


monite)


indigenous


iron


smelters


in the


area.


A good


example


4X1


basin


west


Kalund


lage


ocality


current


research


ocated


northwest


Sumbawanga


(Fig.


2.4).


Sma


amounts


non-ferrou


ores


, namely


lead


, copper


silver


gold


well


as preciou


stone


e.g.


garnet,


reported


from


over the


commercial


Rukwa


quantity


Region.


The


only


None


commercial


them


present


mining


in any
region,


now


closed


due


exhaustion


involved


, copper


silver


gold


in the


1950s


Uruwira


Mpanda


strict


(Moffett


1958).


R -. 4.......-.. ,-a.r.I...nt-i rr.^n m- -^ frrm thm cri itho-rn nrart Af tht la^ia vi lrlnd





Climate


Nkansi


District


and


Ufipa


general,


receive


moderate


mean


annua


rainfall


between


750-


1000


mm


that


spread


over


months


from


December


May


The


heaviest


rain


usually


occurs


between


January


and


March.


The


escarpment


more


rain


than


shore


or the


plateau


due


both


and


vegetation


cover (as elaborated


under Flora


below)


The


plateau


is generally


coo


throughout


year


with


July


being the coldest month


November the


warmest


month


The mountain


tops


may


record


temperatures


ow as 40


n July.


The


hore


on the other hand


is hot


humid


year


during the
proximity i


rainy season


but for the


cooling


greater part


and


of the


stab


effect


daily


variation


temperature


rarely


more


than


five


degrees centigrade


(Moffett


1958).


During


strong
shore


easterly


dry
and


generally


season


, July-October


southeasterly


protected


winds.


from


the
The


strong


plateau
Lake T;


receives


anganyika


winds


plateau.


instead


gets


coo


winds


, especially


night.


Flora


The


hore


character


zed


and


nterspaced


with


Miombo (Brachy


stegia)


woodland.


Grass


formations occur


on the


wider


horses


and


lacustrine


swamps.


Such


localities


include


Kabwe


Kirando


, Wampembe,


and


Kala.


few


areas


, especially





precipitous


or "sacred"6


have


retained


cluster


wood.


The


dominant


species


such


stands


include


Brachv


steqla


spiciformis


and


antune


siana


with


community


dominance


ndex


62.4%


(Kikula


1979)


. Other


species


in the


area


include


Pterocarpus


anaolens


, Diplorynchu


condvlocampon,


Trich


emetica


Brachy


tecia


Julbernardia


globiflora


and


Brachv


steaia


busse


, wh


DiDlorynchus


condvlocarpon


forms


main


hrubs


specie


underground


layer


(Kikula


1979).


Both


escarpments


hills


east


shore


most wooded


obstructs


through


eco-zones


easy


farming


in the


accessib


or timber


district


and


sawing.


because


check


These


steep


human


two


relief


effects


three


most


destructive


means


deforestation


area


with


other


being


bush-


fires.


Some


woodland


under


Government


reserve.


The


dominant


tree


pecies


escarpment8


include


Brachvstenia


piciform


and


Acacia


spp.


with


community


dominance


ndex


70.3%.


Other


species


include


Brachvstecaia


microphy


, Brachvstecqia


ssei


, Cu


ssonia


arborea


, Bride


carthartica,


Sterculia


auinau


oba,


Grewia sp.,


Pericopsis


anqolensis,


Diospyros


mespiliform


Dracaena


spp.


Some hills


n the past were highly respected because


that "Miao" messengers of God (Leza) lived on them.


them


, including burning them, hunting, cutting trees,


even passing (walking) through them.


e the local people believed
Several taboos surrounded


cultivating,


and sometimes


Only selected members of the society were


allowed go there to offer sacrifices and communicate with the "Miao" on behalf of


their communities.


An example


of such a hill is Nswa, north of Kirando.


rnco


cttirl, frnr Vir ndrn and l inil


1(V; W.1-'


979'


SI 1\ I IV k11


nf





and


lerocarva


caffra.


Most trees


, however,


short with


sma


mean


basal


area


that


stand


resemb


regenerating woodland (Kikula


1979)


The plateau is generally treeless,


dominated by grass.


The


main grasses of the plateau are Hyparrhenia spp.


Setaria and some


Themeda spp.


(Kikula


1979).


, Laudetia spp.,
The margins of


large


rivers


tips


precipitous


"sacred"


characterized by mosaic evergreen forest stands which,


in most


cases,


lack


transition


vegetation


with


neighboring


vegetation types (grasslands).


Common species of the


mosaic


stands


include Svzicium


auineense, Salacia


Stuhlamaniana, and


Parinari


curatellifolia.


The


presence of mosaic forests on the


plateau


suggests that the


area


was wel


forested


past.


Aside


from


this


indication,


presence of


large


tree


stumps


and


tands


ferns


(dominant


pecies


being


Preridium


acau


num)--typical


undergrowth


forested


areas


(Kikula


79)--also


plateau was


lead to


forested.


assertion


that


Additionally


long


sources9


ago


testify


Fipa
that


most of the region was forested even as late as the beginning of
this century.


There is little doubt that the disappearance of the


forests


has been caused by human


influence (Nyange


1993).


The


most


probable direct causes


include


shifting


cultivation


, especially


eleusine


year,


(finger)


bush-fire,


and


millet
fuel


which

such


demands


fresh


charcoa


and


field


every


wood


rr~~~~~ ~ ~ ~ nC( I~l fl rm l n .-f n r -ns -nr 1- a t


L ft. A a*


Al -a- -


n~r in rll I nn i I* n A* r~ rr rA *A


I


rm





long


include


timber


sawing


peeling


tree


bark for making


beehives.


Both activities


specie


selective.


mber


sawing


affect


mainly


Brachv


stegla


spiciformis


(Mtundu)


Pterocarpus


anqolensis


(Mninga)


vesicolor


(Mtanqa) and


Sterculia


quinquiloba (Mbalamwezi


, whereas


bark


pee


affects


Brachysteqia


manca (Myombo).


Fauna


The


Fipa


derness


hosts


a variety


anima


most


common


which


baboons


wild


-pigs


hares


-dik,


antelope


, variety


nakes


, bird


many


members


other


animal


families.


The


lacustrine


swamps


and


river


mouths


support


crocodiles


hippopotami.


Game


anima


such


roan


antelope,


hartebeest


kudu


eland


bushbuck


, reedbuck


duiker


buffalo


, elephant


and


were


commonplace


on the


plateau


around


northern


shore


area


until


1930


(Popplewe


1937).


These


anima


migrated


Katav


i Game


reserve


ocated


northeast,


mostly


expulsion
factors.


Mpanda


game


The


most


District


animals


important


owing


area


an i


decades.


attributed


increased


human


The


several


population


which


past.
which


Also


resulted


occupying


important


in addition


that


increase


tends


scare


was


rifles


vacant


hunting


animals


more


than


other


(local)


hunting


technique





woodland


Glossina


morsitans.


areas


infested


which


past,


with


made


tsetse


sleeping


, mostly
sickness


endemic


1958).


veterinary


Today


region
cattle


medicine


in considerable


bush


and


and


prohibited


raising


population


clearance


cattle


common


growth
sidence


raising


thanks


which


(Moffett
modern


resulted


cultivation.


70s the


government


launched


a campaign


clear


bush


some


place


, for example


Kirando


, in order to create


a bush/tree


free


belt at the


bottom of the escarpment to separate


tsetse


breeding
required


habitat


from


continuou


unfortunately


residential


rechecking


that


areas.


method


growing


ow-up


died


which


bushes


natural


death.


The


cyclop
specie


1992


rich


pecie


and


Most of


fauna


gastropod


around


these


200


spec


species


specie


resources


including


bivalve species,


(Evert


endemic.


197


about


7 crab


let al.
(1992)


Miche


Miche


note


that


only


four


species


cichlid


fish


endemic;


seven


molluscs


species


are unique


crab
> this


endemic


lake.


This


and


unique


percent


fauna


probably
liability (
periods


result


ong


colonizing


hypersa


organ


period


isms


nity


isolation


and


history


, phylogenitic


absence


(Ham


1982


Miche


199


Although


annua


fisherie


potential


estimated


-L A A .a


nr nnrr


3nn nnh i,,,


,, I,


L





shore


poor


infrastructure


, especially


road


motor


sels,


and


fishing


gear


(Roest


199


Lack


modern


shing


equipment


explains


why


fishing


industry


is concentrated


pelagic


pecie


such


Stolothrissa


tanqanicae


Limnothr


issa


miodon


(Lumpu)


Laciolates


staopersi


(Mikebuka/M'volo),


four


centropomid


predators


genu


Lates


(Sanqara)


namely


staperss


microlep


auqustifrons,


and


Mariae.


Other


specie


found


Nkansi


district


ude


Hydrocyon


ineatus


(Kibebe)


Alestes


Marophthalmus


(Man


r I


apia


tanqanicae


Boulencaerochrom


Microlep


(Kuhe)


Chrysichthv


brachvnema


(Kibonde)


; Synodontis


(Ncogqo)


, Dinotopterus


cunninqtoni


(Sing a),


and Clarias mossambicua (Kambale)


(Evert


197


Roest


1992


Teoph


pers.


comm. 11).


Socio-cultural


Background


Population


Tables


present,


respectively,


summary


population


census


figure
1988


Rukwa


Region


projected


based


population


last


figures


national


1995


based


nation


on a growth
(Tanzania) ;


rate of


also


12 for


provided


region.


comparative


Figures


for the


purposes.


M. Teophi


is a fisheries officer and during the


time of this research (1992-


3) he was stationed at Kala.
distribution of species in N


I am grateful to him for information on the
kansi District.


(D alqaa),


(Nae~e),







TABLE


Population


Density


Rukwa


Region


District


(Population figures based on the


Statistics


1988 census:


135-


Bureau of


140)


Location


Population


Area
mass


land


Density
(per sq.km)


Tanzania 23,174,336 881,289 26
Rukwa Region 694,974 68,635 10.
Sumbawanga 328,312 13,417 25-
Mpanda 256,487 45,843 6
Nkansi 110,175 9,385 112

TABLE 2.2 A Projected Population Density of Rukwa
Region by District by 1995

Location Population Area of land Density
mass (er sq.km)
Tanzania 28,010,676 881,289 32
Rukwa Reg ion 933,166 68,635 114
Sumbawan a 440,835 13,417 33
Mpanda 344,394 45,843 8
Nkansi 147,936 9,385 1 6


With


a density


people


km2


, 2.6


times


smaller


than


38%


comparatively


national


sparsely


populated.


density,
Anselm


Rukwa


Tambila


region


notes


that


population


Rukwa


remained


sparse


throughout


one


century


a half


"The


official


figure


year


1912/13"


notes


"gave


region,


then


known


Bismarckburg


population


density


people


km2"


(Tambila


1981


15).





The


unequa


population


tribution


natural


uniformly


resources


distributed
, especially


due


arable


the
land


and water


In


Nkans


District,


example,


about


percent


population


and


percent


nhabit


plateau,


terraces


percent


escarpments


shore;


(Bureau


stat


stics


1988)


This


research


demonstrates


that


limited


arable


along


the shore


lake


fluctuation


were


chief


factors


population


movements


last


500


years


(and


perhaps


longer)


between


the lake


shore


on the


one


hand


Fipa


escarpment


plateau


on the other


People


migrated


the escarpment and the plateau during high


returned


during


waters


(more


water


information


periods
chapter


These


population


hore


and


movements


uplands


created


which


eudo


-vacancies


, together


with


along
trade


opportunities


along


hore


, have


attracted


immigrants


from


ndian


Ocean


ittora


(Arabs)


east


Shinyanga


and


Mwan


Regions


(Sukuma)


well


Rwanda/Burund


(Tutsi)


north


aire


(Tabwa


Goma)


west.


This


elaborated


in the


owing


ub-section.


Ethnic


Groups


Today


Nkansi


trict


mult


-ethnic.


Wherea


plateau
Sukuma1


area


occupied


lake


hore


mainly


inhabited


Fipa


majority


variety


and


ethnic


These are immigrants from Mwanza and Shinyanga Regions to the north.


2ra attrertpdr tn the


onuth


bv pastures for their


e lttac a n d


farm


They
Willis


.. ...






groups,
individual


some


The


them


Fipa


represented


extend


between


less


than


one


Mkombe


hundred


north


Kala


south


Lungu,


second


largest


group,


and


Mambwe


south


Wampembe;


Tabwa


Goma


central


area


of the


shore


Arabs


are confined at


Kirando;


and


Sukuma


occupy


pastures


and


farmland


peripherie


villages


north


Wampembe.


Before


we go


any further


important


note


that


character


frontiers


zation


, both


given


above


socio-cultura


been


and


ngu


stic


implified.


Exact


difficult


determine.


because


first


people,


especially


borders


(e.g.


Kala),


speak


a "Creole


type


language


case


Kala


, for


example


mixture


Kifipa


Kilungu.


Second


some


people,


especially


youth,


tend


change


ethnic


identitie


various


reasons.


During


this


research


project,


example,


discovered


dozen Fipa youth


around


(through
Kirando v


cross


vho


-examination)


introduced


over


themselves


Tabwa


or Goma


(from


Zaire)


because


both


Tabwa


and


Goma


who


wealthier


Kirando


than


Fipa.


today


generally


Additionally


economic


hore


dweller


irrespective


ethnic


origin,


stereotype


the


conservative


, superstitious


and


witche


1 4 The Sukuma are


encouraged by the village authorities to occupy the peripheries


of the villages in order to keep their livestock away from the "other people'


crops.


Irnniralv


thp nlItpsaii Fina rnntenrl that ths chnrr< duullorc


ocnnriidk, tha


. w





Commenting


on the


fluid pattern


ethnic


frontiers


area


, Tambila


writes:


There


were


"outsiders"


ethnic


group


situation


stories.


supposed
(Tambila


Such


many


cases


settled down


ettled


[during


region


territory


on the


which
historic


one


can


cannot


century]


which


or members of


hardly


make


compartmentalization


another


write


sense
mply


one


creating
"tribal"


because


there"


1981


The


ethnohistory


of the district


traced


back


, though


with


some


gaps


along


time


chart


beginning


first


ennium


A.D.


The


only


archaeological


source


area


prior
along


current


Tanzania


work


-Zambia


investigations


border


Desmond


Kalambo


Clark


and


colleagues


situated


1960s
the


(Clark


1969


edge


, 1974).
Lake T


The


Kalambo


anganyika


Rift


escarpment


near the


southern


an altitude


150


m (Clark


1969),


about five


from


Lake


Tanganyika


outh


Kala.


Physiographically


and


geologic


Kalambo


Falls


area


closely


related


Kirando


-Kip


Kala


area


, the


shore


localitie


current


research.


The


excavation


Kalambo


produced


most


complete


Acheu


and


tone


tratified


Age


sequence


assemblage,


with


culture


history


racemization


from


estimate


around


200.000


years


(Gowlett


1990)


(initially


dated


around 60,000 years ago,


Clark


1969


, 1974),


present


day


The


present


inhabitant


Kalambo


Lungu,


one





Bantu-speaking


peoples


Corridor


Lungu oral traditions hold that Kalambo was


language


initially


sub-group.
inhabited


by the Fipa.


"If so," Clark argues,


"it would seem that the


Fipa


must


have


entered


valley


some


time


after


eleventh


century,


perhaps


sixteenth


century16


since


earlier date,


if not later, it was occupied by the makers of the


Kalambo


Negroid stock,


industry


made


who


a very


, though
different


most
kind


probably


pottery"


Bantu
(Clark


1974:1).


The Kalambo pottery dates from about


400 A.D.


to about


1000 A.D. and is characterized by bowls and globular pots most


which


undecorated.


Decorated


pots


dominated


mple decorative


techniques, such as grooving and channeling,


hatching,


and


stamping.


Additionally,


false


relief


chevron


designs are found and beveled rims, often externally


thickened,


common.


This pottery type is


related


chronology


and


attributes


both


Urewe


Ware


(dimple-based)


from


interlacustrine


region


Mwabulambo


and


Gokomere


traditions from Malawi and Zambia to the south (Fagan and van


Noten


1964; Fagan


196


Robinson and Sandelowsky


1968; Soper


197


1 b, c).


that


Linguistic
influxes (


evidence


(Nurse


Bantu-speaking


1982


people


Ehret


1991)


, proto-Eastern


suggests


Bantu


from


Niger-Congo


region


settled


nterlacustrine


region (the region bounded by Lakes Victoria,


Albert


(Mobutu),





Edward


This


group


Amin),
formed


Kivu


two


, and Tanganyika)


offshoots


Mashariki


around


and


1000-


Kusini


700


B.C..


between


800-400


B.C.


Later


Mashariki


community


seven


primary


subgroupings:


Lake


Kati;


Upland


Lang
The


Southern


Fipa


, together


Tanzania


with


ombero;


[Lungu]


, Nyiha


and


Corridor


, Nyamwanga,


Mambwe belong to


(Corridor)


group.


The


Corridor


group


arrived


this


ub-region


around


0 A.D.


Ehret


suggests


that


southward


route


proto-


Corridor


Bantu


-speakers


was


"perhaps


initially


along


west


Lake


Bantu


Tanganyika"


population


from the south


(Ehret


1991


southea


or southwest.


50).


stern


This


Implying


that


region


agrees


only


the
ake


with


first
came


Clark


suggestions


also


with


most


oral


accounts


origin


given


inhabitants


of the


corridor region


today


1958


Fagan


Yellen


1968;


1981).


There


seems


controversy


however


between


linguistic


and


makers of


archaeological
the Kalambo


evidence


industry


regard


. Who


who


other


were


words,


inhabited


southeastern


shore


ake


before


sixteenth


century


? According to


Clark


74),


makers


Kalambo


of Fipa.


Falls


f this


industry


were


is the case


neither


who were


Fipa
they


direct


ancestors


and what happened


proto


arrived


-Corridor


in the


Bantu-speakers


region around 0 A.D.


who


Answers


, according


these


Ehret,


questions


(2





Ethnographic


Fipa


N.D.


sing


evidence


between


genealogy


places


seventeenth


Tafuna


coming of both


I eighteenth
chiefdom ol


Lungu


centuries


shore


Kalambo


(Clark


Lungu,


region


74).


Clark


(from


Based


suggests


that


the southwest)


Lungu


in the


genealogy


arrived


seventeenth


Milans


century
dynasty


which


ruled


first


chiefdom


Ufipa,


Willis


suggests


that


Fipa


migration


occurred


some


time


around


700


A.D.


The


origin


Fipa


began


with


a man


called


Ntaatakwa


"the


Unnamed


One"


who came


from


the southwest


near


Lake


Mweru


what


now


northern


Zambia


1968


, 1981).


The


Tabwa


Goma


from


aire


Arabs


from


coast


migrated


Kirando


and


neighboring


during


between


what


century


as a result


Unyamwezi


ong


north


distance


Zaire


trade


west,


Zambia


east


south


(Tamb


refugee
1988).


and


1981


running


The


Uswah


(Indian


1979).


from


remaining


nter-ethnic


ethnic


Some
wars i


groups,


Ocean


ttoral)


Wagom
Zaire


Sukuma


came


(Manyes


and


twentieth


-century


immigrants.


The


former


(Sukuma)


mixed


farmers


(Ha)


reta


attracted


traders


who


mainly


run petty


pasture:
shops a


where


along


ater


eastern


shore of the


ake.


clear


populous


from


have


this


ong


overview


history


that


Nkansi


Fipa


most


District.


these


niefeP ,r. i I-fl


Fnllnirr;nn


Fnri I


mnF+I\I


+~nm


rd rl f n ne


Ilnnn






Political


Orqani


nation


The


political


history


of the


Fipa


begins


with


coming


Ntaatakwa


founder


Milansi


chiefdom.


The


oral


traditions


Milansi


holds


that


Ntaatakwa


sent


five


sons


found


ages


became


and


minor


govern


chiefs


other parts


of the


who continued


country
regard


These


sons


reigning


then


chief


ansi


as their


"father"


1968).


About


middle


eighteenth


century


, during


reign


third


chief


Milans


, Ntseka


some


'invaders'


came


from the north.


These


people,


said to be of Tuts


origin,


usurped


chiefdom


Fipa,


introduced


Twa


ruling


dyna


which


continued


rule


Ufipa


into


beginning


this


century


(Popplewell


193


1968


Wright


198


The


Tuts


believed


organ


zation


have


similar


brought


that


with


them


more


concept
northerly


political


Bantu


states


of Karagwe,


The


two


Buganda,


Tuts


relatively


Ankole


kingdom


strong


, Bunyoro,


consolidated


chiefdom


and E
them


Nkan


3usoga
selves


(the


is 1968).
gave rise


source


name


District)


Lyangal


that


were


stronger


than


others.


The


two state


(Nkansi


Lyanga


were


constant


struggle
growth


(which


two


turn


states


enhanced


tica


and


throughout


economic


eighteenth


century


This


and


state


much


affairs


was


first


half


interrupted


nineteenth


appearance


century.
in Ufipa


of the


Ngoni,


a warrior people


from


southern


Africa


who


rapidly




several


years,


during


which


time


Twa


chiefs


and


their


followers


have


taken


refuge


caves.


After


death


their


leader


, Zwangendaba


Ngoni


quarreled


split up and the different bands


eft Ufipa


in various directions


(Willis


1968


, 1981).


Although the Ngoni occupation lasted less than a decade it


marked a definite watershed in Fipa social and


The


two


chiefdom


began


political


concentrate


history.
internal


developments instead of the military struggles of the


pre-Ngon


period.


The defeat by


Ngon


and,


few years


after,


Bung


forced Fipa to participate more actively in Swahi


trade for the


purpose of acquiring firearms (the Bung defeated them because


they


firearms


(Willis


1976).


Their


involvement


trans-


continental


geographic


trade


location


paid


(between


them,


thanks


Unyamwezi


strategic


north


and


Kazembe to the south


both


important trans-continental trading


centers) and abundant ivory resource.


This in turn


enabled them


acquire


firearms with


which


they


conquered their neighbors


and got war captives who were exchanged for more firearms


other


manufactured


goods


1981


Wright


1982).


last


quarter


nineteenth


century,


Nkans


became


southwestern


strongest
Tanzania.


chiefdom


only


liffe describes it as


Ufipa


"one of the most


elaborate chiefdoms"


, deserving to be called a state


because


"was


more


stratified


, had


more


precise


borders,


and


was





other


Kapufi,


polities


who


plateau"


reigned Nkansi


(Iliffe
about


1979:24).


thirty


years


Mwene


(ca.


(Chief)


1860-90)


made


iances


with


coast


traders


and


reported


have


had


Arab


'prime


minister'


1880s


(Iliffe


1979)1


"Mwene"


governed


, literally


with


meaning


help of


"the


sub-chiefs


one"


"the


known


omnipotent"


as "Walasi"


(Mias


ngular).


beginning


century


, Kirando


was


under


Mwenekandawa


(Manyesha


1988)


. The


prosperity


chiefdom


remained


uninterrupted


unt


coming


colonialism


by the end of the


nineteenth century.


Settlement


Pattern


Ufipa,


many


other


places


Tan


zania


was


variably


affected


'villagization


program


1970s


which


introduced


, among


other


things,


new


settlement


plans,


land


distribution


and


land


patterns.


this


reason


, therefore,


contemporary


settlement


pattern


Ufipa


valid


indicator


past


situation.


reflects


modern


urban


plans


introduced


government


official


The pre-villag


action


pattern


however


reconstructible


through


both


oral


traditions


and


history


Additionally


archaeological


survey


retrieve


settlement


units


most


of which


today lost


middle of


wilderness.


combination


these


methods


show


us that


Fipa


used


have


compact settlement


village


form.


This


believed





long


-held


cultural


trait


1979).


The


concentrated


age


settlement


probably


resulted


from


need


communa


labor


force


produce eleu


(finger)


millet


their


taple


since


time


immemorial


(Willis


1968


:979).


This


settlement


which


system


involved


addition


ding


enhanced


fortifications.


Nearly


ective


defense


pre-colonial


Fipa


villages


were


defended


stockade


and


sometimes


deep ditches.


excellent


example


a defensive


ditch


one


Ngon


called


under


Kantalamba


Zwangendaba


which


n the


was


dug


1840s.


obstruct


Minor ditches


attacking
, which are


common


over Ufipa,


were


dug


keep


wild-pigs


away


from


planted
grounds


fields.


near


Village


perennia


were


sources


usually


water


located


springs,


open


streams


Lake


Tanganyika).


Settlement


pattern


varied


from


linear


circular


amorphou


, depending


landscape


(Thomson


1881


1979).


Although the contemporary houses are rectangular


with two bedrooms and a living room (a plan that G.


I


, usually
Poplewell


observed


1930s


well),


tradition


both


Fipa


and


Lungu


built


circular


house


with


conical


roofs.


The


house


were


poies


with


or without


mud


plastering


with


thatched


roofs.


places


where


building


trees


were


scarce


, people


used


recycle


pole


(also


Popplewe


1937).


The


house


exteriors


were


(and


some


them


are)


painted


decorative


designs,


usually


horizontal


ayers


alternating


rnlnrc


rft on


ranrl


~IOhlfllAI


nr nonritrir


(1rinnru, iIr\


rlaPCrwnc





Farmers


also


build


food


-storage


structures


(barns)


especially


cereals:


along the shore.


eleusine


These


also


millet


constructed


plateau


from


and


wooden


rice


posts


and mud.


Sub


sistence


The


contemporary


inhabitants


Nkansi


District


sub


farming,


herding,


fishing


and


trade.


Petty


income


also


earned


precious
weaving,


through


minera


and


part-time
s hunting,


occupations


and


crafts:


such


as honey


potting,


ecting,


and


mat


iron-smithing.


Farminca.


Farming


Fipa


most


important


occupation.


Some


travelers


(Thomson


farming


1881)


skills


and


who


crossed


expressed
devotion tc


Ufipa
their


farming.


nineteenth


admiration


Their


over


farming


century


Fipa


technique


involved the


use of green


manure


formed


into


cones


(Popplewell


1937


1979),


called


ituumba


Kifipa--a


technique


today


mited


farming.


The


cones


were


prepared


collecting


and
the


heaping-up green grass


heaps with earth.


(cleared


Completed


from the


cones


field'


measured


) and
about


covering
a meter


high


and


meter


wide


base.


They


were


planted


eleusine


millet


first


year


following


year


they were


leveled


ridges


were


made


on which


maize


, beans,


millet


, groundnuts


prepared


and


same


cassava


time


were


adjacent


planted.
> the old


new


one


plot


was


eleusine


millet.


This


process


continued


through


the


third


vear





Planting


and


harvesting


crops


past


was


done


women


while


preparation


fields


was


shared


between


sexes


(Popplewe


1937).


this


strict


division


labor today.


Sim


farming


techniques


and


crops


found


along


shore and on the plateau.


Before


introduction


of the


current


staple
rice18


crop
have


food


(from


over


been


crops,


Asia)

Ufipa


cassava


around


was


cultivated


maize


(from


eighteenth


eleusine


South


century
a cereal


sub-region


America)


staple


believed


around


since


A.D.


(Ehret


1991).


Eleusine

environmentally


farming te

destructive


chnique


, especially


Ufipa


trees.


sometimes


involve


cutting tree


in new areas


burning them to ash.


The


farm


then


hoe


chemical


-ploughed

fertilizer


and


planted.


used


Since


crop


neither


product


manure


unit


area


usua


drops


gnificantly


after


exhaustion


fert


zer"


and


humus,


often


after


third


fourth


consecutive


harvest.


Consequently,


farmer


looks


a new


wooded


plot


and


same


process


repeated.


shifting


technique


cultivation


employed


places


where


farm-land


local


people hold that


as wel


as sorghum,


groundnuts, and sweet potatoes are indigenous crops.
are wild varieties of some of these crops. Wild rice fo
Kipanga, and wild banana, known as Kulumbaleza, use


eleusine,


This is because:


banana,
) there


>r example still grows around
ed to be common in forested


areas on the plateau.


(2) These crops have been grown in the area for years, and


oral accounts give no external origin.


Foreign crops for them included maize,


cassava (brought by a British


*


tm U -


(District


Commissioner))


II~ -


from


America,


S U


-- R ..- aa a aaI A A ap a a


"corridor"


* =* ^


,, i, LI-..-LL L,, I::::


r1A l





plentiful.


where


and


when


land


scarce


farmers


adopt


crop


rotation


and/or


field


fallowing.


Farming


Nkansi


District


principally


subsistence


(food'
rainfall


rather than


amount


for commerce


and


distribution)


in good years


farmers


often


(dictated


manage


produce


surpluses


eleusine


rice


, maize


beans


and


sugarcane.


With


exception


sugarcane,


rest


these


crops
beans


usually


exported


exported


outside


Zaire


and


District.


The


Burundi;


rice


maize


and


and


eleusine


sent


towns


north


country.


Anima


commonplace


keeping.


Nkans


The


District


owing
cattle


domestic
, goat, s


animals


;heep,


donkey


pigs,
kept:


dogs,


chicken


mainly


Fipa.


The


ducks


by the


division


pigeon.


Sukuma


Two species of cattle are


Ankole


historical


sense


type
that


kept


Ankole


cattle


were


brought


into


Fipaland


Tutsi


mmig rants


in the


eighteenth


century.


Sukuma


Zebu


as they


cattle


migrate


are
from


presently being
Shinyanga and


brought
Mwanza


north


following


pastures


and


farm


land.


Fishing.


Fishing


extensively


practised


lake


and


perennia


nvers.


The


chief


catch


lake


Stolothrissa


tancanicae


(~I)


and


Laciolates


staDDersii


(Mikebuka).


These


caught


night


using


umination


method.


kerosene


pressure


amp,


fitted


with


reflectors


rlirt,+


in ki- ant


tha a sitnr


e.;nAl


aF


zebu.


r( yr~n nn


Ir


Ir rnn





fish


haul


accumulate


into


under the


canoe.


lamp;


The


then the


process


may


fishers


lift the


repeated


net and


as many


times


as the


fishers find


necessary.


past,


fishers


used


a wood


place of


pressure-lamps.


The


woods were


carried


on a small


ron grate


fixed


at the front of the canoe.


Exchange.


Both


loca


long-distance exchange


network


traced


back


tory


Ufipa.


Gray


(1957)


suggests


and


as this


each


illustrates


(chapters


and


commercial


other


relationship


between


ndian


ocean


littoral


and


region


around


Lake


Tanganyika


probably


existed


as early


as the


no information


first


millennium


regarding


A.D.


interaction


Although
between t


there


little


coast


Lake


region


for the


following


eight


centuries


there


title


doubt that


Fipa


it continued


were


directly


(Gray


tied


1957).


into


By the


nineteenth


long-distance


trade


century
network


control


their


Nyamwezi


neighbors


north


1979).


The


co-ordinate


geographic


trade


position


between


Ufipa
ndian


enabled


Ocean


region


ttora


, the


Atlantic


Ocean


littoral


, and


interior


centers


such


zembe


and Katanga to the south and Unyamwezi to the north.


The


prosperity


Ufipa


chiefdoms


, especially


Nkan


and


Lyangalile,


owed


much


this


trade.


The


trade


involved


ivory


red color"


(palm oil),


and


slaves


bartered


with


arms


bead


blue cotton cloth


and some broad cloth.


middle


Pottery with TIW attributes. faunal remains, a corner bead and an iron nail


8)19





nineteenth


century


most


trade


traffic


involved


boating


along


across


Lake


Tanganyika


from


Marungu


Zaire


of the


as it was known


boats


"were


then)


about six


(Tambila
fathoms


1981


Gray


[10.8


1957).


long,


Some


but had


sails"


(Gray


1957


29).


Kirando


(and


a lesser


extent


Kisumbi


the south) were


most


important


trading


ports


along


southeastern


was


shore


much


Lake


Tanganyika.


harbor20


than


Kirando's


was


attraction


fert


hinterland


and


comprising


The


relatively
twenty-five


islands at Kirando and the


large
ethnic


fact


population


groups"


that


living


(Tambila


there


1981


narrow


:75).
both


Kirando


and


Kisumbi


thus


demanding


only


few


hours


rowing


, put


two


ports


better


competitive


position


compared to other ports.


This


only


trade


presence


permanent


"indigenous


legacie


which


Arabs


include


they


call


themselves


Kirando


also


people


who


continue


barter


with


their


neighbors


across


(especially


Burundi


and


a certain


degree Zambia


Zaire).


Exported


good


include


grown along the shore,


beans (from the plateau),


honey (from the


neighboring
exchange


wilderness),


they


import


and


beer


dried


fish


(Primus


(especia
brand),


DaCaa).


construction


materials


(portland


cement


, roofing


meta


sheets


, etc.),


printed


cotton


cloth


(Zairean


vitenge)


sugar


electronics.


Ac f r 2C nnnr hhrhnr (Idsn and nrrnt'rtsd ic rnnrsrnsri Itirandn rannnt


-





Occasional


crafts.


These


include


those


trades


that


performed


specialists,


mostly


as part-time.


Ironworking.


natural


Fipa


-draft


were


renown


furnaces


smelting,


mnclunau,


plus


refining
discussed


furnaces


length


called


vintenawe.


chapter


, not


Since


much


subject


said


here.


indigenous


iron


production


in Ufipa


stopped


n the


1930s


, mainly


due


ts suppression by the colonia


government


(Wright


1982


1985).


It was revived


n the


1950s


, as requested by the colonial


administrators


hoes due


(Wright


1982


, following


World War


, 1985)


a decline


The reviva


due


supply


European


lasted only for three


a combination


factors.


years
Some


smelters
practicing


simply


backed


smelting


the enthusiasm
about a decade


after
before


having
(Wright


stopped
1985).


Others,


this


each


has


earned,


thought


that


government's
consequently


request


, pun


was


a mere


trick


smelters.


meant


should


identify


and,


remembered


that


two


decades


the


same


government


banned


indigenous


production


on the


pretext


that


was


hazardous


to the environment.


Those


who


broke


had


been


caned


and/or


imprisoned21


smithing,


however


, ha


survived


present.


Smith


either


manufacture


sma


tools


such


knives


, spear-


heads


, adze


axes


, and


bush-knives


from


scrap


meta


repair


metal


objects


including


hoes,


cooking


pots,


buckets


, and


many





other


teams.


The


skill


learned


through


apprenticeship,


usually


among


kinfolks


often


fro m


grandfather


grandchild.


"outsider"


also


acquire


knowledge


paying a


"", which may be a goat, cattle,


money,


or labo-


Potting.


This


a common


specialized


craft


Nkansi


District.


mainly


undertaken


women


who


usually


learn


from


their


kinfolks


(mother


aunt,


sister)


and


friends.


Potting


remains


predictable


a part-time


market


although


activity


occas


because


onally


does


yields


not have a
attractive


incomes


potters.


sketry


. Basket weaving


men


activity.


Baskets


made
wood


from


(twig


variety
palm le


materials


aves etc.


including


included


bamboo


this group are


, palmetto,


basket


traps


for fishing


which


made


from


twigs,


reed


or bamboo.


Matting.


Mats are made from


two


different


materials


and


gender
mats (


division


is based


ukili/iamvi)


on the
often


type
woven


material
women


used.


where


Palmetto
as reed-


mats (msengele/mlago) are made by men.


Boat


building.


Fishers


built


canoes


--made


joining


pieces


timber


as opposed


dug-out


canoes


which


common


among


other


small-scale


fishers


East


and Central


Africa.


accounts


logs


canoes.


testify


that


The


past


boat


fishers


ding


was


used


brought


Wagoma


from


Zaire


towards


end


last


century


(Manyesha


1988).


Boat-building


became


an unavoidable


option


rn~~~~~~nr aA tr rn n t k ann a ra n~ -t r a I a nn.kani ananrt


C:n^ -


~mnF


SkA a a ~ rA FC


~ma


I





precipitous,


was


easier


transport


timber


than


ready-


made canoe from there to the


lake.


Other


activities that


usually


performed


under


different


; category
specialists,


(woodcraft),


include


making


mortars


and pestles.


Huntina


and


honey


collecting.


Hunting


used


a very


important


business


before


1970s


when


Government


became


more strict about it


in a campaign


to protect endangered


species


including


rhinoceros


, elephant,


cheetah,


and


leopard.


Most


often


hunters


were


honey


collectors


well


, the


latter


being


a serendipitous


activity.


With


hunting,


some


hunters


have


taken


honey


collecting,


and


now


owly


moving away from depending on


natural


occurrences


beehive


maintenance


artificial


ones.


The


new


beehives


made


either


from


bark


some


trees


from


two


split-open,


dug-out


pieces of


logs.


Beliefs


The


belief


system of


Fipa


past involved


spirits


and
leve


ancestor


At the


cults.


Supernatura


top was Leza


(God)


beings were
the creator


conceived at three
with whom human


beings
(Mwao,


communicate


singular)
"hosts"


who


were


directly


Leza's


Next


messengers.


rank


They


were


Miao


resided


various


rocks


, precious


including
pebbles,


forests


etc.


Miao


island


could


animals


, big


communicate


trees,


with


~t a a- alan~aAa L~ ...a' a n .A~ II., ..aka 4.1 a


*


A


~HA IW ~~ CEL


~IIL~An


~ L~ A I r


AlA ACA rj


A A ~ A A


P


+nA





supernatural


being


were


zimu


(ancestral


spirits).


These


were


divided


into


good


and


evil


spirits.


The


latter


were


specifically


referred


Viswa


(Kiswa.


singular)


Good


spirit


were


believed


people


who


ived


and


died


happily


whereas


iswa


were


those


people


who


were


mistreated


relatives


or fellow


human


beings


during


their


lifetime;


so, they


came back to avenge.


Viswa often were thought to be the source


sufferings


iving


people


(Manyesha


1988


also


"Ritua


tes"


, chapter


Today


Christians


remaining


over


(Catho


Muslims


percent c
(Diocese


and


inhabitants


Sumbawanga


traditionalists.


Chri


Ufipa


1985).
stianity


The


was


brought


Ufipa


by the


White


Fathers


missionaries who came


way
They


(Popplewell


their


first


1937


ssion


Sumbawanga


Diocese


southeastern


1985).


shore


Kirando


1888


Kala


189


said


that


Fipa


accepted
subjects


Christianity


rapidly


Ufipa


very


became


regalia


readily


a Christian


sale"


reports


people,


(Iliffe


that


[chief]


1979:232).


Although


this


less


may


an exaggeration,


resistant


Christianity


argument
testified t


that


other


Fipa


were


writers


(Popplewe


1937


Willis


1981


Manyesha


1988).


Rev.


Bertsch


"Notes


on Karema


Diocese"


attributes


state


affairs


fact that:


Ufipa


paganism wa


without


a rea


between


parts


- a





or taught


also


that


by o
there


fficia


were


ministers.


never


any


might


secret


mentioned


societies


Nyao
other


Cewa]


countries


which


Africa


make


(Rev.


paganism
Bertsch.


so strong in
i. 68, quoted


n Manyesha


1988:13).


Islam


was


brought


Swah


traders.


According


account


collected


Manyesha,


Islam


Nkan


District


was


brought
Mwene


Mwiny


Kapufi


(ca.


man


1860


from


and


Tabora


1890).


, during


said


that


reign


upon


arriva


Kirando


, Mwiny


became


a friend


Sumai


another


immigrant
changed h


from


Kigoma


name


who


was


(Manyesha


ater


converted


1988).


This


account


and


bears


some


contradictions


, leading


one


doubt


credibility.


example


, it


very


likely


that


Sumail


was


already


Muslim


before


met


Mwiny


or before


came


Kirando.


First,


name


Sumaili


is very


kely


Islamic


rather


than


either


Chr


stian


or indigenous Fipa


and second,


fact


that


Sumaili


came


from


Kigoma


around


end


nineteenth


century,


a time


when


Islam had already began


n Kigoma,


suggests that he was


already


Mus


Ora


traditions


collected


during


this


study


however


suggest
traders


that


some


Islam


decades


came


before


Kirando


through


Christianity,


Arab


probably


(Swahili)


around


middle of the


nineteenth century.









CHAPTER


LITERATURE


REVIEW


IRON


CENTRAL


METALLURGY


EAST


AND


AFRICA


Students


traditionally


divided


ndigenou


Africa


African


into


two


metallurgy
sub-regions


have


when


discussing


history


metallurgy:


north


Sahara


desert,


including the Mediterranean Sea littoral,


the Nile


Valley,


and the Red Sea coast; and 2) south of the Sahara desert (sub-


Saharan Africa),


including West,


East


Central and South


Africa


(van


der Merwe


1980


Kense


1985


Miller


and van der


Merwe


1994).


This


division


derives


from


dea


that


metallurgical


history


of the


two


sub-regions


different.


example, the north, especially Egypt (the Nile valley and the Red
Sea coast), experienced an elaborate bronze technology whereas


the south did not (Kense


1985).


Moreover,


in Egypt (and the


Middle


East


that


matter),


there


was


long-delayed


development


two


four


ennia


from


time


iron


smelting
was used


was


first


regularly


performed


(ca. 5000 B.C.) to the


1000 B.C.) (Waldbaum


1980).


time


iron


This was


largely
could


due
meet


the
the


already
needs


established


bronze
metal


industry


was


which


concerned


(ca.






(Sassoon


1963)1


sub-Saharan


Africa


iron


smelting


began


developed
prolonged


spatially,


delay


well


(Tylecote


complexity,


1975


Schmidt


1978b;


without


199


Finally,


closely


during
linked


Bronze


with


Ages,


rest


north


Mediterranean


was


more


world


culture


technology


than with


sub


-Saharan


Africa


Snodgrass


1980).


review


chapter


focu


sub-


Saharan


Africa.


begin


with


debate


about


origin


iron


meta


urgy


sub


-Saharan


Africa.


provides


background


owing


genera


scusslon


that


covers


spatia


and


temporal
southern


metallurgica
neighbors to


distribution


Africa.


The


techniques


north


east


iron


meta


urgy


chapter
practiced


in eastern


ends


with


Fipa


south and west and


central


and


review


and


shows how


their


they


relate


rest of the


continent.


Origin


Iron


Metalluraqy


sub-S


aharan


Africa:


Schoo


Thought


The


archaeological


iron


artifacts


come


from


Middle


East.


The


reported


sites


include


amarra


northern


Sassoon also adds that bronze, in some


respects, has several advantages over


"Until it has been carburized and tempered," he notes,


"pure wrought


it comes from the furnace is not harder than aood aualitv bronze."


iron


Furthermonr






where


four-sided


instrument"


about


long


been


found


n a grave


is dated to


around


5000


B.C.


Another


early


Tepe


pherica


have


northern


been


found


Iran


from


where


three


a habitation


small


eve


, nearly
, dating


between 4600-4100 B.C.


(Waldbaum


1980).


nickel-content


analysis


conducted


some


these


objects


determine


whether


they


were


made


from


smelted


meteoritic


Iron


showed


that the


Samarra


object


was


smelted


those


from


Tepe


were


meteoritic


(Waldbaum


1980)


Although


this


result


indicates


that


iron


smelting


probably


began


seven


lennia


ago,


Iron


Age


Middle


East


begin


four


ennia


matter


when


iron


ceased


considered


material


"this


precious


for making


[the


became


tools and


Age]


first


accepted


weapons.
reached


as the


According


fruition


predominant
SWaldbaum,


about


10th


century


B.C.


large


region


stretching


from


Greece


Leventine


coast,


around the


9th century B.C.


Mesopotamia,


somewhat


ater


Europe


and


region


farther


east"


(Waldbaum


1 980:82


Africa


earliest


iron


objects


meteoritic


they


come


from


two


sites


Egypt:


Gerzeh


(nine


beads


Armant


ring)


dating


between


3500-3100


B.C.


(Waldbaum


1980)


When


comes


melting,


incident


melting


evidence


also


comes


from


Egypt


with


find


from






and


2565


Abydos,


-2440


dating,


B.C.)


pectively,


Sixth


Fourth


Dynasty


2345


Dyna


-2181


(ca.
B.C.)


(Waldbaum


1980).


Iron wa


not regularly produced


in Egypt unt


seventh century B.C.


(Snodgrass


1980).


smelting


sub-Saharan


Africa


started


more


than


millennia


ago.


The


earliest


evidence


including


furnace


come


tuyeres1
(Okafor


around


199


sixth


century


from


Taruga


Okafor


(Tylecote


1993


both


Dimm


and


Nigeria,


Niger


with


Nsukka


dating
a date


mid-ninth


century


(Calvocoressi


and


David


79),


some


sites


in Gabon


, including Otoumbi,


Moanda and Oyem,


dating


between


seventh


and


second


centuries


(Schmidt


1985


1989;


Peyrot


1992


Oslisly


199


Other


early


sites


include


Buhaya,


Tanzania


, dating


between


500-


200 B.C.


(Schmidt and Child


1985)


Meroe


Sudan


, dating


200


B.C.


(Trigger


1969).


There are


also


some


early


dates


from


Rwanda


and


Burund


which


confirmed


earliest


ub-Saharan


ninth


Africa.


These


century B.C


come


from


Miramba


Gesiza


Rwanda


, Mubuga


and


, dating t
Rwiyange


Burundi


, dating


between


mid


-fifteenth


eighth


centuries


B.C.


(van Grunderbeek


1992).


The


question


how


sub-


Saharan


Africa


developed


acquired


knowledge


ironworking


or meta


urgy


in the


place


existed


since


time


amateur


students


metall


urgy--


missionaries


, explorers,


and


travelers


--during


(ca.





emerged:


one


believes


external


origin,


and


second


ieves


loca


invention.


was


According
imported


external-origin


into


sub


-Saharan


school


Africa


iron


from


metallurgy


eastern


Mediterranean.


Different


routes


have


been


proposed


from


eastern


Mediterranean


to Carthage,


from


there


across


Sahara


, to


Wes


t Africa;


Meroe


from


there


West


East


Africa


Aksum


from


there


East Africa


a direct


maritime


route


from


eastern


Mediterranean


Africa.


Many,


however,


think


that


first


route


most


viable


assoon


196


pson


1985


Merwe


1994)


. They


argue


that


knowledge


ironworking


was


brought


western


Mediterranean


Phoenician


colonialists


sometime


after


1000


B.C.


Then


, by


sixth


century


B.C.


Berbers


north


Africa


who


"must


have


learnt


from


Carthagian


were


establishing


these


south


Sahara


time


iron-s


melting


would


have


reached


southern


side


Sahara


during


first


few


centuries


B.C."


assoon


196


179)


evidence


been


found


that


supports


these


various


routes


whole


external


-origin


model.


Nonethe


followers


this


school


hold


tenacious


that


is reasonable


believe


that


knowledge of


metal


working


was


introduced


-Saharan


Africa


from


outside


, despite


Discussion on this subject can be found in several sources,


(1952)


Arkell


(1961.


1966),


- -JI


Sassoon


(1963),


Shinnif


including Mauny


(1966)


, Diop


--





paucity


archaeological


evidence


those


areas


that


might


have


acted


conduits


spread


this


technological


knowledge"


(Miller and van der Merwe


1994:8)


. The debate to


people


Holl


puts


matter


faith"


(Ho


199


331) rather than a scientific discourse.


The crux of this school is the hypothesis that in order for


iron


technology


(which


complex


pyrotechnology)


develop it must be


(relative


iron


preceded
smelting)


less complex,


pyrotechnologies


low-temperature


such


production of copper, lead,


, glass, cement, and/or k


n-fired


pottery and terra-cotta (Wertime


1980;


Kense


1985


Kense and


Okoro


1993).


Iron technology would emerge from the use of


ores


fluxes


Furthermore,


instance


copper


lead


i copper
smelters


and


would


ead


have


smelting.
became


acquainted with


kilns,


bellows, and


fluxes--all


devices


used


iron


smelting.


Based


on this


hypothesis,


sub-Saharan


Africa


could not have invented iron technology


because the sub-region


does


not


have


conclusive


evidence


non-ferrous


pyrotechnologies


that


predate


iron


production


(Kense


1985;


Phillipson


1985)


In their words, Kense and Okoro writes:


"The


chief


obstacles


accepting


indigenous


origin


African


iron-working


include


lack


evidence


any


pyrotechnologies


in sub-Saharan Africa


predating the


beginning


iron


production,


absence


cultures


demonstrating


transitional


state


between dependence on stone and then iron


Fhr ;+r +bPhnhlhn;F 'J I k~TP;P 'J nrl +ka F~~+ ~knm k~F th ~n





found


site


that predates


beginning


of iron


production


Near East" (Kense and Okoro


199


456).


The


loca


invention


school


base


their


argument


on several


reasons.


First,


there


no evidence


an externa


source.


Andah


(19


79)


example


argue


most


informative


criterion


determine


diffusion


technology


space


would


similarit


between


'donor'


and


recipient'


technologies


But,


as Schmidt


observes


ronworking


Saharan


Africa


doe


seem


have


affinity


with


that


eastern


Mediterranean.


words


"the


technology


Kagera


Region


possibly


Taruga


so distinctive


from


what


we know


from


Europe


Middle


East


that


feel


that


most


reasonable


technology


hypothesis


for the


independent


origin


of th


invention


preheated


Africa"


smelting
Schmidt


1983


434).


Sec


ond


demonstrated


both


technological
ethnographically


dexterity


and


variability


archaeologically


sub-


Saharan


Africa


have


many


people


(Schmidt


1981


198


hmidt


Avery


1978


Okpoko


1987


David


. 1989


believe


that


African


ronworkers


must


have


been


authors


and


mere


copyists


a foreign


technology.


After


observing


uniquen


Cameroon


Nichola


Mafa


Iron


David


smelting
colleague


technology


conclude


North


that,


As there


is no record of


uch a group of techniques


anywhere


admire


else


loca


genius


invention


unknown


implied
inventor


must


(David


.9nnt*h 'nf





Third,


some


evidence


early


copper


working


(Calvocoressi


David


1979


Grebenart


1987


continue


found


which


cha


enge


validity


pyrotechnologic


hypoth


esis


used


externa


-origin


school.


With


evidence


there


now


reason


entertain


alternative


hypothesis


autochthonous


iron


technology


Africa


Schmidt


(1983:43


argues.


However


finds


referred


Schmidt


here


have


undergone some


reassessment


1988)


and


Nonetheless


they
with


as old


1000


B.C.


they


were


being the


thought


re-examined


date


copper


melting


, one


still


say


that


copperworking


preceded


ronworking


for three


four


centuries


when


compared


with


-century


B.C.


date


from


Taruga


and


sukka


(Tylecote


197


Okafor


and Ph


1992


Okafor


1993).


Fourth


, the


universality


pyrotechnologic


hypothesis


also


questioned.


apply


this


model


Saharan


Africa


, according to Diop (1974) and Andah (19


79),


an example of


importation


cultural


founded


Western


and


differ


mode


ecologica


with


that


which


often


background


sub-


Saharan


work


upon
Africa


which


Si


because


they


iron ores


et al.


(1988) have


reexamined the Niger data and have found that


unquestionable


over


2500


evidence


for copper smelting dates to around 1000 B.C.


reported


earlier


Calvocoressi


as opposed


David


Grebenart


(1987).


s The thesis that the
technology is based la


emergence of iron technology was
iraelv on a research which T. Wer


Radomir Pleiner conducted in Iran


in the


1960s (Wertime


preceded by copper


time,


1968).


. Smith, and
Their method


n i -lt 1,r An' r n- n a n1 .- ;:, r e-t, 1- i, A .. n-n ,ll *. mt ,- -


/-^4_ tnnt- r^f nnC;


J r


v


nuw~ rlmn r+r


I


1






more


abundant


sub-Saharan


Africa


and


more


easily


recogn


izable


than copper and tin,


Andah


argues


that,


metallurgy


copper
bronze
region


could


bronze


working.


have


working
[Moreove


beginning


characterized


different


meteoric


Andah


ron]


(1981)


the working
methods" r


(Andah
further


begun wi
or even
r.1 in di


without


ssing


preceded


fferent


Age


different


such


through


copper


parts


may


have


form


as experimenting


been
iron
with


79:141).


suspect


genuinen


this


mode


, especially


regards


app


cation


sub-


Saharan


Africa


. To


this


is another


version


"Hamitic


myth"


aimed


demonstrating


superiority


Caucasians.


protests
outside,


that


usua


is wrong


from


claim


north


that


across


ideas


Sahara


peoples


from


stimulated


generated
production


most


or the


major


earlier


developments


working


pertaining


and


early


copper


food


West


Africa]"


(Andah


1981


who


is right


who


wrong


There


is not


enough


evidence


allow


one


provide


objective


answer


question.


that


more


data


needed


(Okafor


199


Holl


1993).


meta


Meanwh


urgica


we need


experiments


ntensify


aimed


archaeological
understanding


research


how


might


have


been


smelted


initial


Africa.


Attention


should


directed


toward


under


standing


cultural


and


ecological


factors


that


might


have


played


gnificant


roles


past


iron





cause indigenous development of iron technology
regions in the continents, as Okpoko proposes:


in the


various


The


production


prehistoric
smiths in


times


of
and


steel
the


sociopolitica


sub-Saharan


significance


and


economic


Africa


metals


ives


and
the


African


peoples


reassessment
beginnings of


from


ancient


times


diffusionist


iron


technology


critical


theories


parts


Africa...


indeed makes no sense to depend so much on dates


attempts at tracing the so called


routes


the knowledge of iron smelting spread to


There


is now


need


look


cultural


' through
parts of


which
Africa.


environment


within each society which made possible


the take-off


such


technology


(that


internal


demands


for the


iron produced; sources of the


skill


necessary


raw


processes


materials;


which


exploitati


and


occurred


materials;
)n and r


type
each


people


with


3finement


metal-working


society)


(Okpoko


1987:224).




Iron Metallurgy in East and Central Africa


Early


Evidence


The earliest evidence for iron smelting


n East and Central


Africa


comes


from


nterlacustrine


region


where


various


metallurgical


materials,


including


furnaces,


tuyeres, slag,


wood


charcoal


findings


, iron ores,


indicate


and


that


ron artifacts have been found.


inhabitants


These


nterlacustrine


region,


using


ocally


available


materials


such


hematite


and


limonite ore.


refractory


clay,


wood,


and charcoal,


were


able


--






first


millennium


B.C.6 (Hiernaux


Maquet


1956


van Noten


1985


van


Noten


Raymaekers


1986


1988


Schmidt


78a


1981


Grunderbeek


1981


Schmidt


and


1985).


Some


other


early


ironworking


have


been


reported


throughout
ironworking
Tchissanga,


eastern


locale


on the


central


with


slag


Congo


coast


southern


objects


, dating


Africa.


was


third


Recently


reported


century


from


B.C.


(Denbow


1990).


Nkese


western


Usambara


mountain


northeastern


Tanzania


a furnace


, slag,


tuyere


fragment


were


A.D. (

found


found


(Schmidt


Mkiu


, dating


1988).


between


Slag


south


heap:
es S


first


and


alaam


third


tuyere
, dating


centuries


fragments

to between


were
the


first


fourth


centuries


A.D.


(Cham


1988


, 1994


Schmidt


et al.


1992


. Further


outh


Mozambique,


tuyere


fragments,


ranging


from


second


third


centune


A.D.


were


found


Matola


Enkwa


zonjan


Leaves


tund


(Mora


1984


nclair


1991


nclair


et al.


1993)


Several


ronworking


sites


marked


with


furnace


remains


tuyere


fragments,


slag,


sma


objects,


such


as nng


sections


tenth


iron


rods


centuries


or strip


Malaw


range
(Davi.


time


from


on-Hirschmann


third


and


Mosley


1988;


Juwayey


i 1993).


These


include


Nkope,


Mwabulambo,


Some earlier dates have also been recorded in the


region,


for example,


145 bc at Rwiyange


in Burundi, 68


95 bc at Gasiza


in Rwanda, and


r- r -


a -1t


-Sl -a -. .


, nn -a- tL~l n* A r. .. L kal .n~Mt- mlnrakl S -n nnn a.r fla-. t* n mn IS a -*






Kayerekere,


Kaziwiziwi.


Kamguze


Some


the second century


A.D.


area


Chombe


copper


have


been


Shelf


iron


bead


found


, Kasumanguwo,


s dating
Mabveni


Ngo


probably


mbabwe


(Huffman


fragmentary
thumb-piano


198


iron


keys,


Large


tools


that


quantities


as a razor


date


fifth


slag,


, spear


century


bloom


point,


A.D.


, and


ring


have


been


recovered


Kapirimbwe


Zambia


(Phi


pson


1968a


, b)


. The


earliest


Africa


known


have


been


preserved


found


meta


Divuyu,


(iron


and


Botswana


copper)
, dating


southern


between


550 and


730 A.D.


, Nqoma also


Botswana


dated


between


850-


1090 A.D.


(Denbow


1990) and Broederstrom,


South Africa


, dating


to the fifth century


A.D.


(Denbow


1990).


Having


seen


early


evidence


now


examine


how


iron


meta


urgy


became


known


East


and


centra


Africa.


Origin


ronworkinq


East


and


Centra


Africa


The question of


acquisition


meta


urgy


mplie


tempora


ogica


begin


patial
discu


lineality


ssion


--characteristics


from


where


which


oldest


make


evidence


ocated.


"logical"


approach


an i


inherent


problem


that


promotes


tendency


think


terms


ngle


origin


This


in turn


lead


dichotomy


"inventors


-a-vi


copyists"


diffu


sion


becomes the


only


explanation


change.


This


tendency


clearly


evident


when


one


reviews


way


people


have ex


plained


cultural


changes


in eastern


central






Scholars,


for example,


have


traditionally


begun


with


interlacustrine


metallurgy


region


when


sub-Equatoria


dealing
Africa


with


that


pread


iron


where


earliest


evidence


have


bee n


found


Archaeologists


have


speculated over how the technology got or developed there from
the time when the first evidence were found (Leakey et al. 1948;


Hiernaux


and


Maquet


1956;


Hiernaux


1959


Posnansky


1966)


The


most


popular


assumption


was


that


technology


was


brought
people.
search


into


interlacustrine


This contributed greatly to


origins


1960s and early


Bantu-speakers


1970s (Guthrie


196


1967;


region


Bantu-speaking


ntensification


linguists
Greenberg


196


Oliver


1966; Ehret


1972).


It later became accepted that


Bantu-speakers originated


n the Niger-Chad Basin (Ehret


1982a;


1991


Nurse


1 982).


claim,


verified.


however


, has


been


n order for archaeologists to do so,


archaeologically
they need to find


some


evidence


migration


route


ironworking


, including


from


Cameroon,


along


Equatorial


alleged
Guinea,


Bantu


Gabon,


Central


African Republic,


and Zaire.


But research in the area has


progressed very slowly as it is


"hampered by thick vegetation,


erosion


rapid


deterioration of


archaeological


remains"


(de Maret


1990:109; also see Eggert 1993).


Meanwhile


claim untested.


archaeologists
As late as the


continued


linguists'


1980s Schmidt argued that "the