FRONTIERS OF VIOLENCE IN

NORTH-EAST AFRICA
Zones of Violence

General Editors: Mark Levene and Donald Bloxham

Also available in the Zones of Violence Series

Alexander V. Prusin, The Lands Between: Conflict in the East European

Borderlands, 1870–1992

Mark Biondich, The Balkans: Revolution, War, and Political Violence since
1878
FRONTIERS OF
VIOLENCE IN
NORTH-EAST
AFRICA
genealogies of conflict
since C.1800

RICHARD J. REID

1
 3
Great Clarendon Street, Oxford ox2 6dp
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University Press
in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
© Richard J. Reid 2011
The moral rights of the author have been asserted
Database right Oxford University Press (maker)
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose the same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Library of Congress Control Number: 2010943357
Typeset by SPI Publisher Services, Pondicherry, India
Printed in Great Britain
on acid-free paper by
MPG Books Group, Bodmin and King’s Lynn

ISBN 978-0-19-921188-3

1 3 5 7 9 10 8 6 4 2
 Contents

Acknowledgements vii
List of Maps ix
Glossary and Abbreviations xi

Prologue: The Past in the Present 1

PART I: SETTING AND APPROACH

1. Interpreting the Region 9

Peoples and places 9
Toward a new regional historiography 14
The fertile frontier: borderlands and frontier societies
in north-east Africa 20
2. The Shadows of Antiquity 24
Axum and its aftermath 24
Violence and mythology: the new Zion 26
The Oromo frontier 30

PART II: VIOLENCE AND IMPERIALISM:

THE ‘LONG’ NINETEENTH CENTURY

3. States of Violence, to c.1870 39

Haunted house: the wars of the mesafint 39
The politics of shiftanet: Tewodros and the frontier state 49
Frontiers of faith 59
4. Borderlands, Militarism, and the Making of Empire 66
Resurgent Tigray 66
Frontiers (1): the Mereb River zone 69
Frontiers (2): Sudan and the north-west 78
Pax Solomonia? The restless militarism of Menelik’s empire 84
vi conte nt s

PART III: COLONIALISMS, OLD AND NEW

5. Demarcating Identity: The European Colonial

Experience, c.1890–c.1950 95
Blood on the tracks: Italians on the Mereb 95
Raising the first born: Italian Eritrea, to 1941 99
Somalia divided 108
Choices: Eritrea, 1941–52 114
6. The Empire of Haile Selassie, c.1900–74 129
The anomalous empire 129
From Wal Wal to Kagnew: war, reconstruction,
and reinvention 139
Fatal federation 154
Crises impending 159

PART IV: REVOLUTIONS, LIBERATIONS, AND

THE GHOSTS OF THE MESAFINT

7. Revolution, ‘Liberation’, and Militant Identity, 1974–91 173

The violent state renewed: the Derg 173
Inter-state front lines 177
Emerging markets of violence: ethnic and
nationalist borderlands 183
Tigray: revolution and renaissance 188
Eritrean epicentre 193
8. New States, Old Wars: Violence, Frontier, and
Destiny in the Modern Era 208
The dead governing: liberation and federation 208
The great unexpected war 217
Borderline nation: Eritrea as frontier society 226
Oromia 236
Continuities and transitions: Somalia and Ethiopia 240

Epilogue: Armed Frontiers and Militarized Margins 246

Endnotes 257
Bibliography 287
Index 305
 Acknowledgements

I n terms of the management of this project, I would like to mention at the

outset Christopher Wheeler and Matthew Cotton at Oxford University
Press, and the general editors of the series, Donald Bloxham and Mark
Levene, for their encouragement and helpful comments at the draft stage.
Friends and fellow practitioners who have likewise provided guidance and
inspiration—if sometimes unknowingly—include Christopher Clapham,
Dan Connell, the late Richard Greenfield, Wendy James, Douglas Johnson,
Gaim Kibreab, Alex Last, Tom Ofcansky, Izabela Orlowska, Martin Plaut,
Gunther Schroeder, and Alessandro Triulzi. I must record my profound
thanks to informants, colleagues, and friends from both Eritrea and Ethiopia,
many of whom I cannot name individually but who have contributed inef-
fably to whatever understanding of the region I can claim to have. Some
deserve special mention, however, and in Ethiopia I am grateful to Alemseged
Girmay, Amira Omer, Eyob Halefom, and Gebretensae Gebretsadkan, while
Yemane ‘Jamaica’ Kidane has been generous with time, information, and
opinion. My visits to Addis Ababa were greatly enhanced by Patrick Gilkes’
wonderfully open-handed help and advice. In Eritrea, I am more grateful
than I can say to Alemseged Tesfai, while trips to Asmara would have been
so much less productive without the help of Azeb Tewolde and her staff at
the Research and Documentation Centre, and also of Brook Tesfai, Mekonen
Kidane, Seife Berhe and Zemhret Yohannes. I want to reserve special men-
tion for my two old buddies in Asmara, Abraham Keleta and Eyob Abraha,
who have always made it worthwhile, no matter what. Closer to home, the
team at Chatham House in London—Sally Healy, Roger Middleton, and
Tom Cargill—have provided much in the way of inspiration and expertise.
Then there is my wife Anna, who has lived with north-east African violence
on a daily basis for some time now, more than anyone sane has a right to
expect. I thank her with love and amazement. It is never more important
than in a book on north-east Africa to point out that none of the above bear
v ii i ac k nowle dg e m e nt s

any responsibility for the interpretation which follows; some, indeed, will
heartily disagree with it.
Finally, I would like to dedicate this book to the memory of an old friend
who died before its completion, Amanuel Yohannes. Amanuel taught me
more about the region than anyone else, and especially how to think differ-
ently about its past and its present; I owe him a great deal, though I never
got to tell him just how much.
rjr, london
 List of Maps

1. Physical north-east Africa

Source: J. Markakis, National and Class Conflict in the
Horn of Africa, Cambridge University Press, Cambridge, 1987 xiii
2. The region during the zemene mesafint
Source: J. E. Flint (ed.), Cambridge History of Africa Vol. 5, c.1790–c.1870,
Cambridge University Press, Cambridge, 1976; this version adapted to
R. J. Reid, A History of Modern Africa, Wiley-Blackwell, Oxford, 2009 xiv
3. The region in the late nineteenth century
Source: R. Oliver and G. N. Sanderson (eds),
Cambridge History of Africa Vol. 6,
c.1870–c.1905, Cambridge University Press, Cambridge, 1985 xv
4. Imperial Ethiopia, mid-twentieth century
Source: P. Henze, Layers of Time: A History of Ethiopia,
Hurst, London, 2000 xvi
5. The region in the early twenty-first century
Source: P. Henze, Layers of Time: A History of Ethiopia,
Hurst, London, 2000 xvii
This page intentionally left blank
 Glossary and Abbreviations

AAPO All Amhara People’s Organisation

Andinet ‘Unity’, youth wing of the UP in the 1940s
ANDM Amhara National Democratic Movement
Anyanya Southern Sudanese guerrillas from the late 1950s onward
ascari African troops in the Italian colonial army
banda irregular militia
baria derogatory Tigrinya term for the Nara,
implying ‘black slaves’
blatta title given to learned men, counsellors
BMA British Military Administration
BPLM Benishangul People’s Liberation Movement
dejazmach noble title, lit. ‘commander of the gate’
Derg informal name for Ethiopian government 1974–91,
lit. ‘Committee’
EDU Ethiopian Democratic Union
ELF Eritrean Liberation Front
ELM Eritrean Liberation Movement
ELPP Eritrean Liberal Progressive Party
EPDM Ethiopian People’s Democratic Movement
EPLF Eritrean People’s Liberation Front
EPRDF Ethiopian People’s Revolutionary Democratic Front
EPRP Ethiopian People’s Revolutionary Party
fitaurari noble title, lit. ‘commander of the vanguard’
gada Oromo age grade system
GPLM Gambella People’s Liberation Movement
gult fief
habesha common, informal term for highland Ethiopians and Eritreans
IGAD Intergovernmental Authority on Development
Jiberti Amhara and Tigrinya populations who are Muslim
x ii g lo s sary and ab b rev i at i on s

kebessa Eritrean highland plateau

Kebre Negast Glory of the Kings
lij title for children of nobility, lit. ‘child’
metahit Eritrean lowlands
ML Muslim League
na’ib title of Ottoman governor of Massawa, lit. ‘deputy’
negus king
negus negast king of kings (emperor)
NFD Northern Frontier District (Kenya)
NIF National Islamic Front
OAU Organisation of African Unity
OLF Oromo Liberation Front
ONLF Ogaden National Liberation Front
OPDO Oromo People’s Democratic Organisation
PFDJ People’s Front for Democracy and Justice
ras noble title, lit. ‘head’
sha’abiya ‘popular’, nickname for EPLF
shangalla derogatory Amharic term for peoples to the west of Ethiopian
highlands, in Sudan
shifta bandit, rebel
shiftanet banditry, rebellion
SPDO Sidama People’s Democratic Organisation
SPLM/A Sudan People’s Liberation Movement/Army
SYL Somali Youth League
TFG Transitional Federal Government
TNG Transitional National Government
TPLF Tigray People’s Liberation Front
UIC Union of Islamic Courts
UP Unionist Party
Woyane lit. ‘revolt’, referring to 1943 uprising as well as a nickname for
TPLF
zemene mesafint ‘era of the princes’
 R . N il

R
e

E
D
R.
At

ba
A R A B I A

S
ra

R. Bark

E
R. G

A
a
as
h
R. Mare
sh

Blu
e
Jebel

Nil
Mara

h
Nuba

R. A w as
Mountains

R . Bar o
THE
SUDD
W
hi
le
Ni
R.

R.Omo
She
le

bel
i

R.
Jub
a
I N D I A N
Land over 2000 metres
1000—2000 metres O C E A N
0—1000 metres
0 500 km
R.Tana

0 300 miles

1. Physical north-east Africa

Source: J. Markakis, National and Class Conflict in the Horn of Africa, Cambridge University Press, Cambridge, 1987
 32 °E 36 °E Beni Amer Habab 40 °E 44 °E TIGRAY Provinces of the Gondarine empire

M
Mensa

ar
K A FFA Sidama states

Ba r k a
16°N

ya
Khartoum R. Bilen Keren A far Dromo, Somali and other ethnic groups
Massawa
Blu Land over 3,000 feet
Kassala Baria
e a Zulla Red Sea 0 300 km
am
Kun Ma

Nile

Saho
re Hodeida
b
R.
keze R. 0 200 miles

Y
Te A
T I G R Adowa 48 °E 52 °E
Gedaref Axum

SIMIEN Mocha

bo
Metemma
Karkoy a G u l f o f A d e n

Aze
Falash
Rah Assab
ad R
. Gondar WAG Aden
BE

ite Nile
12°N G E LASTA
L. Tana Obok
12°N
M
QWARA D
Debre E R YEJJU Tajura
Tabor

Wh
R Magdala
E Zeila
D AM WALLO r
M
E
HA Di
AW GOJJAM RA
AG
Ab
ba Bulhar Berbera
iR

a
. m Majerteyn
la Ankober
T u I s a
q
SHOA Harar

a
D a r o d

h
Ab ic hu

c
e
GU

Gi
RA

be
A
RY GE

R.
8°N A 8°N
INN
Jimma

A
DY
TA
i

HA
G us

BA
o ie b
Bonga R. Ar

M
KA
K A F F A
MO

O
DA

AM
D a r o d
S I

AL
W
Omo R.

Sh
n eb a
ele Obbia
r a i
y
B o

R.
w
a
4°N H 4°N
Rahanweyn

I N D I AN

D i j i l
OC E AN
Bardera

Juba
Mogadishu

R.
32 °E 36 °E 40 °E Brava 48 °E 52 °E

2. The region during the zemene mesafint

Source: J. E. Flint (ed.), Cambridge History of Africa Vol. 5, c.1790–c.1870, Cambridge University Press, Cambridge, 1976; this version adapted
to R. J. Reid, A History of Modern Africa, Wiley-Blackwell, Oxford, 2009
 30 °E 35 °E 40 °E 45 °E TIGRE Provinces and regions

Atb
KAFFA Sidama states
BOGOS

ara
I t t u Oromo, Somali and other ethnic groups
Keren
Ethnic/ linguistic limits of Somali

R.
Omdurman Khartoum Massawa
HAMASIEN Menelik‘s territorial claim of 1891
Kassala B a ria 15°N

a
Asmara Red Coastal territories of Somali clans which concluded

am
Blu
15°N M
ar ‘Treaties of protection with Britain’, 1884–6

e
eb

un
R. Anfilla Sea Battles involving Ethiopian troops

Ni

K
Tekeze R. (Ital.) 0 300 km

le
TIGRE Adowa Edd
Gedaref Axum 1896 (Ital.) 0 200 miles

D a ( A
50 °E
Mocha

n a f a
Karkoj SIMIEN
Gallabat Amba

k i r )
1888 Assab
1887, 1889 la sh a Alagi
Gondar Fa (Ital. 1882) Aden

Blue

l
Ra WAG 1895
ha d R A

e Nile
. D EM BY

Nile
Obok G u l f o f A d e n

BE
L. LASTA Wichale

G
Tana Tajura (French 1884–5) C. Guardafui
QWARA

EM
Roseires Debra YEJJU (French 1884–5)

Whit

D
Jibuti

ER
Tabor Magdala
ER WOLLO Zeila
ED
M Oromo

SHANQUL
BANI
W AWSSA

A
A GOJJAM
AG A

M
Bulhar Ras Hafun
bba i R

H
. BASSO Dir Berbera

A
10°N 10°N

R
A
Fashoda

Didessa
Imbabo Ankober q
1882 Jigjiga s a D a r o d
WOL L E GA SHOA I
1900 i
So Addis Harar
o

R.
ba o m .
Ababa sh R

Gi b
tR
O r

GU
. a

O
A

e
ar RY Aw

RA
B

R.
o R. A

o
N

G
GE
IN

O r o m
A It tu l
DY

A
J i mma
HA TA i IIIig
BA Aruss

D
G
o je
b R. AM

E
Ak Bonga K
ob

N
o KAFFA O

O
R.
Pib

AM

a
Ba

AM
D D a r o d
or
hr

SI

AL
R.
el–

W
Omo R.

Wa
Jeb

bb
eS
el

ha m Jayd
be
lle 1903 Obbia
5°N 5°N

e
(Hobyo)

y
n

i
a

w
r o) o
o
om

a
B
(O r aw

H
D
a R. R a h a n w e y n
L. Rudolf
S
I n d i a n

D a r o d
D i j i l
O c e a n
Bardera
Mogadishu

Juba R
L. Albert

. Brava
30 °E 35 °E 40 °E 45 °E 50 °E

3. The region in the late nineteenth century

Source: R. Oliver and G. N. Sanderson (eds), Cambridge History of Africa Vol. 6, c.1870–c.1905, Cambridge University Press, Cambridge, 1985
 N SAUDI
ARABIA
Red Sea

SAM
HAR
BOGOS
Keren Dahlak Is
Massawa
SUDAN Kassala Arkiko YEMEN
M ar HAMASIEN
e Zula

b
Blu

Gura Senafe
e N

Alitena
Ta k a z z e
TIGRAY Adwa AGAME
eil

Metemma Debre Abbay

(Gedaref) Makale
SHIRE TEMBIEN
SEMYEN Cheleqot Amba Alagi
Derasge
Lake Ashangi Assab
Mal Chew
Gonder
Azezo Dankaz
Gorgora ANGOT
QUARA DEMBIYA
Emfraz
Lalibeia
YEJJU
DJIBOUTI
Aw a Tajura
Lake Debre Tabor LASTA sh
Tana WOLLO DJIBOUTI
Zeila

GOJJAM Dessie

Debre Markos
SOMALIA
Abba MANZ Sala Dingay
y
Debre Birhan Ankober Diredawa
WOLLEGA SHOA Aliyu Amba
Angolala
Addis Alem Harar Jijiga
ADDIS ABABA Mieso

ENARYA
Webe
GURAGELAND
Sh

Jimma Lake Zway

eb
el

BALE
le

KAFFA
Awasa
Omo
GAMU OGADEN
GOFA Lake Abbaya
SUDAN (Margerita)
Lake Chamo SIDAMO

BORANA
Lake Turkana
(Rudolf)
0 100
K E N YA SOMALIA
km

4. Imperial Ethiopia, mid-twentieth century

Source: P. Henze, Layers of Time: A History of Ethiopia, Hurst, London, 2000
 N
SAUDI ARABIA
Red Sea

Dahlak Is
ERITREA Sana′a
“Independent since 1993”
Hodeida
YEMEN
SUDAN TIGRAY Mocha
Assab Aden

Lake Tana AFAR Gulf of

DJI BOU T I Aden
Bahr Dar

AMHARALAND
BENI - SHANGUL
Berbera
DIREDAWA
Hargeisa
OROMIA ADDIS ABABA HARAR SOMALILAND

GAMBELA

OROMIA
SOUTHERN
SOMALI
PEOPLES’
STATE

SUDAN

REPUBLIC of
0 1000
K E N YA SOMALIA
km
Addis Ababa and Harar are city states; Diredawa is a separate administrative region.

5. The region in the early twenty-first century

Source: P. Henze, Layers of Time: A History of Ethiopia, Hurst, London, 2000
This page intentionally left blank
 Prologue
The Past in the Present

I n the first few years of the third millennium, the region of north-east
Africa is as enmeshed in conflict as it has been for several decades. Somalia
is engulfed once more by violence, the result of both factional fighting and
the Ethiopian invasion; Ethiopia itself is politically tense, its government
confronted with armed insurgency among the Oromo in the south and the
Ethiopian Somali of the east; the Sudanese peace agreement looks impossibly
fragile, and violence has flared in contested areas between north and south;
Eritrean and Djiboutian forces are engaged in a standoff across their com-
mon frontier; and—the epicentre of so much regional conflict—the Eritrean–
Ethiopian border remains highly militarized, with armies representing two
competing national missions glaring at one another with malicious intent. It
is this last conflict which best illustrates the analytical parameters of this
book. Eritrea and Ethiopia, the two key state-level actors in the account
which follows, appeared to be on the brink of renewed war in the early
twenty-first century: scores of thousands of troops were dug in on either side
of a border which was as torturous as it was disputed, and the governments
in Asmara and Addis Ababa seemed as far away from any kind of rapproche-
ment as they had been at any time since the Algiers Agreement in December
2000 supposedly ended the original conflict. In fact, while the Algiers
Agreement had ended a battle, or series of battles, it had not addressed the
causes of the war itself. That war had begun, ostensibly at least, in May
1998—but as this book will argue, it had been going on for some consider-
able time before that, interrupted by brief periods of armistice, quiescence,
and even, apparently, sporadic goodwill. Its genealogy, in fact, can be traced
back through the twentieth century and into the nineteenth. It is clear that
what happened in May 1998 was only the latest manifestation of a war which
had been going on for a very long time.The conflict represented the crystal-
lization of a number of intertwined historical grievances, the opening up of
2 p rolog ue

so many fault lines, and constituted another episode of extreme violence in

a regional cycle of conflict stretching back more than two centuries.
In the early years of the twenty-first century, beyond the northern high-
lands themselves, the Eritrean–Ethiopian war had had a profound effect on
the wider region, either sparking new, proxy, conflicts or joining up with
extant ones. It was in Somalia in particular that the implications of the war
between the two countries were clearest, and where the violence inherent
within their respective political cultures was most dramatically manifest.
Somalia had imploded in the early 1990s, but in the first years of the twen-
ty-first century violent instability there had taken on new forms. In a world
increasingly defined by post-9/11 allegiance, Eritrea sought to shore up the
Union of Islamic Courts in Mogadishu, a government which appeared to
have at least some semblance of local support but which the United States
believed sponsored, and provided a base for, Islamic terrorism. When at the
end of 2006 Ethiopia invaded what remained of ‘Somalia’, the US—after
initially, so it is claimed, advising against such an act—came to support it,1
and the stated objective was to overthrow the pernicious al-Qaeda-suc-
couring Islamic government. Within a few months, entirely predictably, an
insurgency had erupted against the Ethiopians, while the Eritreans were
widely believed to have been arming the insurgents themselves. A proxy
war had begun, compounding the stand-off on the Ethiopian–Eritrea bor-
der itself and—temporarily at least—replacing actual fighting along that
frontier. At the same time, Eritrea found itself threatened with the ‘terrorist-
sponsor’ label by the US. This was a ludicrous accusation by a misguided
administration.Yet what was certainly true was that there appeared to be no
regional crisis which the Eritrean government would not exploit for for-
eign policy gains against Ethiopia: it reached out to the Southern Sudanese
government, raising the issue of Eritrean support for South Sudan’s seces-
sion from Khartoum, even as the Comprehensive Peace Agreement itself
appeared increasingly fragile; it at first provided military support to, and
then sought to build alliances among, a range of Darfur rebel factions which
for a time at least—like their Somali counterparts—saw Asmara as a safe
haven. Eritrean President Isaias Afeworki saw himself as a peace-broker; his
critics detected the whiff of megalomania, and at best saw him as a trouble-
some meddler. Meanwhile Ethiopia provided succour to armed groups
opposed to the Eritrean state; Eritrea, in addition to their Somali connec-
tions, trained and armed insurgents among the Oromo and anyone else in a
position to direct their fire at the government in Addis Ababa.
 p rolog ue 3

The governments of Eritrea and Ethiopia themselves were in increasingly

precarious positions at home, their power based upon greater or lesser degrees
of violence or the threat of it against sizeable sections of their own popula-
tions. Governments—indeed entire state apparatuses—which had grown out
of guerrilla movements were ever less secure and drawn to various forms of
violence in the attempt to stabilize themselves: in both Ethiopia and Eritrea,
the ‘new’ political elites were imbued with a militarism and a faith in armed
force which had been forged in mountains and barren places during their
armed struggles.Yet these ‘new’ elites had much in common with their polit-
ical forebears in that their visions of statehood had been conjured up on the
periphery and carried through blood sacrifice to the embattled centre to be
made flesh. Given to armed adventure and intrigue, these ‘new’ governments
now sat, trembling and suspended, at the centre of a web of interlinked con-
flict, linked by almost imperceptible, but remarkably resilient, threads to the
past as well as to the wars of the present, and indeed those to come.
In Ethiopia, the coalition government of the Ethiopian People’s
Revolutionary Democratic Front (EPRDF) was faced with a growing armed
insurgency in the Ogaden, the Somali-dominated ‘Region 5’, where the
Ogaden National Liberation Front (ONLF) demanded the right of self-de-
termination from Ethiopia. There was also an ongoing insurgency among
the Oromo, led by the Oromo Liberation Front (OLF), and among an array
of smaller ethnic groups along the western border with Sudan.These peoples
remained violently irreconcilable to the Tigrayan–Amhara order at the cen-
tre of Ethiopian political life; elections of a kind might be staged to present
the image of pluralism and accountability, but the violence at the heart of the
Ethiopian political system remained overt, whether in Addis Ababa itself or
in the outlying provinces. Prime Minister Meles Zenawi was no more will-
ing than his predecessors to brook dissent, whether ideological or ethnic.The
system of ethnic federalism now in place was only the latest, albeit markedly
sophisticated, in a line of political strategies to secure some degree of unity
and cohesion across the Ethiopian region stretching back to the zemene
mesafint, and indeed earlier. In Eritrea, meanwhile, armed irreconcilables
were likewise gathering, within and on the borders, albeit on nothing like
the scale of the insurrections confronting the Ethiopian Government. More
importantly for Eritrea, in the immediate term, was the marked degree of
militarization in political and cultural life, as the spectres of wars past, present,
and future haunted public discourse, and ensured the maintenance of a cap-
tive army of young men and women whose lives were being sacrificed to the
4 p rolog ue

service of an ungenerous state. Levels of oppression were heightening year

on year, as the Government became ever more wary of enemies both within
and without; Eritreans were told daily that the Ethiopians could attack at any
time, but increasingly there were neither panem nor circenses as a reward for
the resultant loss of freedom which—contrary to Jefferson’s dictum—flowed
directly from apparently eternal vigilance.
The region was characterized by a series of interconnected conflicts, a
spatial network of violence which—it was increasingly clear—would need
to be understood by those who would promote ‘peace’ across north-east
Africa and the Horn.2 Yet what was perhaps less commonly recognized was
the antiquity of many of these wars, and their profoundly historical nature:
in other words, there were temporal as well as spatial networks of violence,
indicating a depth which was rather more difficult to grasp than mere sur-
face spread. It will be argued in this book that while novelty can indeed be
identified in the contemporary situation—most obviously in terms of the
political contexts with which much violence as well as interethnic and
interstate relations have occurred—there are much clearer lines of continu-
ity than discontinuity in the modern history of the region. The key crises
and the major fault lines—the frontiers concept developed in the first chap-
ter—have deep roots, and much of what has happened in the early years of
the new millennium represents a continuation of, or a reversion to, well-
established precedent in terms of intra-regional relations, at the level of state,
ethnicity, and locale, however shifting those definitions may be. Cyclical
patterns of hegemony and subjugation, and the creation of militarized com-
munities and identities based on alternating ideas about defence and aggres-
sion, are discernible over a period of two centuries or more, and make it
possible—indeed essential—to talk of genealogies of conflict and identity
over this timeframe. Claims for novelty—for which ‘revolutionary’ libera-
tion movements have an especial propensity—should not be overlooked;
but nor should they blind us to the deeper roots of groups and movements
and ideas, even in the face of prevailing vernacular wisdom.
It has become clear in the course of the late twentieth and early twenty-
first centuries that an appreciation of la longue durée in terms of the history
of violent conflict in this region is long overdue; certainly, several contem-
poraneous and cumulative dynamics have produced an era of continuous
conflict since the late eighteenth century.This era has been characterized by
markedly intense, cyclical, and increasingly institutionalized violence
between, and systematic persecution and/or marginalization of, particular
 p rolog ue 5

population groups. This is not, then, a particularly happy book, but there is
no reason why it should be: the region’s troubled present is rooted in a trou-
bled past, and an appreciation of the longue durée is at the heart of this
project. The region of north-east Africa is conceived here as stretching, on
a south–north axis, between the southern Somali coast and the Eritrean
shoreline, between Mogadishu and Massawa, and encompassing the vast
range of mountains and plateaux known as the Ethiopian Highlands and its
adjacent lowlands; and this region is distinctive vis-à-vis many other parts of
Africa both temporally and spatially. In terms of timescale, it is possible to
trace patterns of conflict over a remarkably long and continuous period,
owing chiefly to the relative wealth of source material, both indigenous and
external, relating to the area. While the actions, ideologies, and visions of
literate Semitic-speaking elites were recorded in chronicles and ecclesiasti-
cal records from the middle of the first millennium ad onwards, foreign
visitors penetrated the states and societies of the region much earlier, and in
greater numbers, than most other parts of Africa, especially Europeans from
the sixteenth century. In spatial terms, too, the region is noteworthy for it
has an historically rooted, geopolitical coherence—notwithstanding its cul-
tural, linguistic, and environmental diversity—which means that the com-
petition for space, resources, and political power within it was remarkably
intense. This was a region in which a series of interlocking competitions
linked the Tigre and Tigrinya, in the north, to the Amhara further south, to
the Oromo and Somali still further south, and in turn linked these to the
Afar in the east, and the Sidama and others in the west and south-west.
Ultimately, indeed, it was this very competition which would give birth to
the modern nation-state—or empire-state, perhaps more accurately—of
Ethiopia at the end of the nineteenth century, and in that sense it might be
argued that ‘Ethiopia’, insofar that it is at base the culmination of so much
spatially defined conflict, is indeed much older than we might assume.
This page intentionally left blank
 PART
I
Setting and Approach
This page intentionally left blank
 1
Interpreting the Region

Peoples and places

It is important at the outset to provide the framework—in terms of physical
geography and human settlement—within which our discussion is situated.
It will be clear that the integrated history of this complex region cannot be
understood without first appreciating its environmental and ecological
diversity, and the socio-economic systems which have resulted. These sys-
tems are evidently subject to continual change and adaptation, and it is not
our intention here to provide a ‘static’ picture, for just as peoples and com-
munities migrate, so too do identities and senses of belonging, according to
time, space, and circumstance—this, indeed, is a theme which will surface at
various stages in our story. Nonetheless, it is important to attempt a ‘snap-
shot’ of the human protagonists in our narrative and their physical environ-
ment, before turning to the interpretation of these—interpretation, that is,
both by outsiders and by the peoples themselves. The book is concerned
with a broad corridor of conflict which has the Ethiopian Highlands (and
their Eritrean extension) as its centrepiece; it incorporates these highlands
and their escarpment and lowland peripheries, stretching between the mod-
ern Eritrean Red Sea coast and the southern, western, and eastern foothills
and lowlands of present-day Ethiopia. In particular contexts, the Sudanese
and Somali frontier zones must be considered, but only insofar as these
relate to the core themes of ethnic and religious conflict, the irruption of
external influences, and the violent state- and empire-building which takes
place in the central and northern Ethiopian Highlands in the course of the
nineteenth and twentieth centuries.
The region thus defined comprises remarkable diversity in terms of
physical terrain, vegetation, and climate; geographical difference explains
much of the conflict and competition under examination, and indeed these
10 set t i ng and ap p roac h

physical borders—the demarcations imposed by nature, as it were—are

doubtless the most important in the first instance. Particular geographies
have facilitated particular economic systems, and thus have reinforced
notions of ethnic uniqueness and difference. In the broadest terms, the vol-
canic fertility of the highlands—where rain falls relatively abundantly—is in
contrast with the dry, hot coastal lowlands of Eritrea, Djibouti, and Somalia,
the Nile floodplains of eastern Sudan, and the savannah scrub of northern
Kenya. In Ethiopia, these surrounding hot lowlands of desert and steppe,
known as Quolla, reach to 1500 metres around the foothills of the great mas-
sif. Beyond this is a sub-tropical zone, or Woyna Dega, which reaches to
around 2500 metres and which is cooler, wetter, and contains pastureland.
Beyond this still is the high mountainous zone, which is temperate and wet,
known as Dega. The Ethiopian tableland itself is a natural centre of central-
ized civilization, protruding skyward to dominate north-east Africa, and has
the greatest concentrations of population, although it is characterized by
natural divisions in the form of river beds, valleys, and deep gorges which
have separated communities at key moments in the region’s history. An
important feature of this landscape is the distinctive mountainous amba,
with sheer sides and flat tops. Most dramatically, the Rift Valley splits the
highlands into two, western and eastern, zones—a landscape which has had
a profound influence on Ethiopian history.
Eritrea’s highland plateau—the kebessa—is an extension of the Ethiopian
Highlands, and is flanked in the west by lowland plains, which stretch toward
Sudan and are extremely fertile; and to the north and east, by a hot, arid
coastal plain which links up with Djibouti. The two broad Eritrean zones,
therefore, are the kebessa and the lowlands, or metahit, although there are
important sub-regional differences in habitat and culture within these. The
Somali landscape is characterized by mountains in the north and north-east,
which abut the eastern Ethiopian highlands, but more generally by hot
savannah with irregular rainfall. This is a region which has been especially
prone to drought, and water scarcity has had a huge impact on human rela-
tions across the Ethiopian–Somali and the Ethiopian–Kenyan borderlands.
But even the well-watered highlands have experienced periodic drought,
and states and societies have been vulnerable as a result, which has coincided
with, and indeed spurred, massive political and social change. There are a
number of examples of this through the period under examination. Food
shortages in the north in the 1860s were both the symptom of prolonged
violence and economic collapse, and the cause of further socio-political
 i nte rp ret i ng the re g i on 11

insecurity; the major famine of 1888–92 drove much political violence in

that critical moment in the region’s history; the prolonged drought across
the Sahel belt in the 1970s and 1980s is crucial to understanding the instabil-
ity of the Ethiopian state and the political cultures which emerged around
liberation struggle. The pressures facing pastoralists and farmers alike, and
the broader competition over resources, have fuelled a great deal of conflict
across the region.
These various landscapes are inhabited by—indeed have been instru-
mental in the creation of—a vast array of ethnic and linguistic groups. In
terms of human populations, clearly, the ‘zone of violence’ envisaged here
involves primarily the Tigrinya of highland Eritrea and northern Ethiopia,
the Amhara of central Ethiopia, and the Oromo and Somali of southern and
eastern Ethiopia. However, a range of other, invariably smaller, groups, nota-
bly in south-west Ethiopia and along the western flank of the corridor—no
less important, and in certain contexts even more important than their bet-
ter known neighbours—need also to be considered, including Gurage,
Welayta, Sidama, Hadiya, Kambata, Kaffa, Beni Amer, and Beni Shangul.
Most commentators have categorized the peoples of the region as belong-
ing to one of three major blocs—Semitic or Semitized, Cushitic, and Nilotic
or Sudanic—according to a range of variables, including language group
and genetic origins. In a handful of cases, these divisions are a little mislead-
ing, as some population groups in fact straddle more than one such category.
The Nilotic groups are strung out around the edges of our zone, from the
Kunama and Nara in the western lowlands of Eritrea and adjacent Sudanese
and Tigrayan borderlands, to the various groups along the Ethiopian–
Sudanese frontier zone, including the Berta, Anuak, Koman, and Beni
Shangul.These groups have been historically referred to in pejorative terms
by highlanders—the Nara, for example, have been called in derogatory
Tigrinya baria, implying ‘black slaves’, while the Beni Shangul have com-
monly been referred to as shangalla in Amharic, although the term is applied
to all ‘black’ peoples to the west. Some practice mixed farming, though
many remain primarily pastoralist, and many are Muslim.
The Cushitic category contains a vast array of peoples, spread across the
region. Some pre-date Semitic populations by some distance. Thus the
Agaw are clustered north of Lake Tana and west of the Takkaze river, and
are also represented in Eritrea by the Bilen around Keren. Then there are
the Beja: although in various contexts, both past and present, the ‘Beja’
are referred to as a distinct ethnic group—such as in Eritrea, where they are
12 set t i ng and ap p roac h

also known as Hedareb—in other contexts it is more properly a linguistic

term. Beja-speakers inhabit the north-east African coast from Egypt to
Eritrea, and inland occupy swathes of eastern Sudan and northern and
western Eritrea. Today they include, among others, some Tigre (the second
largest cultural and linguistic group in Eritrea after the Tigrinya), sections
of the Beni Amer, and Hadendowa ethnic groups of Eritrea and eastern
Sudan. Most are Muslim agro-pastoralists. The Saho and the Afar, or
Danakil, are Cushitic, largely Muslim, groups straddling central-eastern
Eritrea and north-east Ethiopia, while the largest single Cushitic grouping
in Ethiopia is, of course, the Oromo, who have a common language but
who have diversified in culture and economy. While largely pastoralist in
origin, migrating into the highlands from the sixteenth century onward,
they have become urban dwellers and settled farmers in some areas, just as
others retain their pastoral roots; some Oromo are Christian, others have
absorbed Islam. In the broadest terms, the more southerly Oromo tend to
be Muslim—in Hararge, Arsi, and the Gibe region—while those in Welega,
Wollo, and Shoa are Christian; some have become so assimilated into high-
land habesha1 culture that they speak Amharic as their mother tongue—one
of the challenges for the emergence of a coherent Oromo consciousness.2
Yet in some areas migrating Oromo have themselves absorbed and assimi-
lated conquered groups, notably in the Gibe region of the south-west. In
sum, there is no singular ‘Oromo experience’ but rather many such experi-
ences, although as we shall see certain key narratives have been extracted
from Oromo history, and a discernible Oromo ‘culture’ identified and clar-
ified. The Somali are also Cushitic, associated primarily with pastoralism in
the great plains of the east, although again others have become farmers.
Most Somali are Muslim, one of the key dynamics in the region’s history.
In the fertile south-west of Ethiopia are other Cushitic language groups,
such as the Konso, and others are clustered together under the term Sidama;
these include Kambata, Gimira-Maji, and Janjero. They inhabit a complex
cultural and ethnic zone which has been much prized by central state-
builders owing to its natural wealth; hoe agriculture is common, while
there is a mixture of Christianity, Islam, and indigenous belief systems. In
this area, mention should also be made of the Omotic cluster of commu-
nities—peoples living along the Omo River—which comprises remarka-
ble linguistic and cultural diversity, including the Kafa grouping. There is
no common religion here and peoples are both pastoralists and grain
cultivators.
 i nte rp ret i ng the re g i on 13

The Tigrinya of highland Eritrea and Tigray, and the Amhara of the great
Ethiopian massif comprising Wollo, Gondar, Shoa, and Gojjam in Ethiopia,
continue to be the main shapers of political destiny to the present day.They
have much in common in the way of culture and custom, but there are
regional differences amongst the Amhara, and between the Tigrinya north
and south of the Mereb River; nonetheless Tigrayans and Amhara have
competed for political dominance of the region for a millennium or more.
They are primarily plough farmers, while also keeping some livestock, and
represent the central Semitic bloc. The Semitic or ‘Semitized’ bloc also
includes a range of other groups, including some Tigre-speakers, of which
the Beni Amer are the largest grouping, the Rashaida along the Red Sea
coast, and ‘pockets’ amidst Cushitic populations, notably the Adari of Harar,
and the Gurage south of Shoa. Some of these have become ‘Semitized’
through extensive intermarriage as well as through political and military
absorption, much as swathes of the Oromo population have become assimi-
lated. Although highland Semitic culture is associated with Christianity,
sizeable numbers of Amhara and some Tigrinya are also Muslim (known as
Jiberti), as are the inhabitants of Harar and most Gurage.
While a broad historical overview would suggest ongoing rivalry between
the Tigray-Tigrinya and Amharic segments of the Semitic bloc, and the
struggle of the Cushitic and Islamic ‘south’ against the domination of the
Semitic and Christian ‘north’, it is also the case that the various cultural,
ethnic, and linguistic fault lines of the region—the zones of identity, as it
were—are intertwined and overlapping. It is true, moreover, that ethnic
identity has been flexible and fluid, historically—doubtless it is ever thus,
and this should not be especially startling—and that an instrumentalist
interpretation of ethnicity would suggest that particular identities have
become solidified in the last century or so owing to the formation of the
modern state of Ethiopia. In other words, the Amhara as a hegemonic group
have become rather clearly defined as a result of the state-formation process,
and indeed a number of communities—across the south and south-west,
notably—have in turn sharpened their own ethnic edges in response to
Amhara dominance.Yet I would argue that regardless of whether particular
groups expanded or contracted according to political circumstances, such
identities did indeed have deep roots and defined characteristics which were
less concerned with biological descent than with cultural traits—in the case
of the Amhara, for example, the Amharic language and Orthodox Christianity,
with its attendant cultural apparatus.There are institutions and accoutrements
14 set t i ng and ap p roac h

associated with being Amhara which are as important to understanding the

longevity and depth of the ethnie as any sense of who exactly is Amhara at any
given point in time. In sum, it is not that ethnic categories are formed
purely according to the instrumentalist model, it is rather that such group-
ings contract or expand according to circumstances, and that such an epi-
sodic process represents a shifting frontier of identity and cohesion.
There can be little doubt, too, that marginalization produces indignant
cohesion, often compelling groups with loose structural and cultural affili-
ations into larger communities. This can be seen, for example, along the
Ethiopian–Sudanese border, a zone of sporadic, low-level conflict through-
out the twentieth century. The 1980s witnessed the emergence of Berta
militia in Beni Shangul, and Anuak militia in Gambella: in the latter prov-
ince, Nuer and Anuak competed for political dominance, a struggle which
was heightened in the 1990s with the arrival of Amhara, Tigrayan, and
Oromo settlers. This was the latest stage in a cyclical push-and-pull experi-
enced by the peoples of the region between Sudan and Ethiopia, an experi-
ence which has clearly led to sharpened self-definitions on the part of the
communities themselves—manifest, for example, in the creation of the
Gambella People’s Liberation Front and the Benishangul People’s Liberation
Movement (dominated by the Anuak and Berti respectively).3 These devel-
opments have served to increase the ‘visibility’ of the ethnic periphery across
north-east Africa—scholarly analysis since the 1980s has certainly tended to
focus on the overtly ethnic aspect of much regional conflict.

Toward a new regional historiography

Several broad groupings and thematic approaches in the literature are par-
ticularly important. In terms of the older literature on Ethiopia—up to the
1970s, broadly—there is what we might generally term the ‘Greater Ethiopia’
or Ethio-centric school, which ultimately saw both Ethiopia and the wider
region from the perspective of Addis Ababa. Within the essentially benevo-
lent imperial state conceptualized by such writers came not only the recently
conquered but generally pacific Cushitic (including Oromo), Nilo-Saharan,
and Omotic peoples of the south and south-west, but also the incorporated
Somali of the Ogaden and Eritreans in the north. Much of this older work
was concerned with the state-building travails and successes of the eighteenth
 i nte rp ret i ng the re g i on 15

and nineteenth centuries. This is in overall contrast to the more recent

literature, which has tended to focus on the twentieth century, and the
period after the 1950s and 1960s in particular. There are exceptions to this
general trend—the work of Donald Crummey, notably, and the necessarily
elongated projects of archaeologists—but little new work has been done on
the pre-1900 era for some years now.‘Greater Ethiopianists’ were concerned,
above all, with the essential continuity of ‘Ethiopian’ (or at least Semitic)
civilization from the early centuries ad onward, the resilience and adapta-
bility of that civilization, and—notwithstanding the trauma wrought upon
sections of the population by Solomonic mythology—its assimilationism.
This older generation of scholars was also characterized by its tendency to
particularize Ethiopia and Ethiopian-ness, to the extent that academics who
‘did Ethiopia’ regarded themselves rather self-consciously as a breed apart
from other Africanists, and were regarded as such by the latter. The two
rarely intersected.4 Scholars of the region were in thrall to the mountainous
mysticism of Ethiopian antiquity and to the supposedly rarefied environ-
ment facilitated by literate Semitic culture.
Increasingly, set against this was the more ethnically or regionally based
nationalist literature which began to reflect the aspirations of those very
groups (or elements within them)—chiefly Eritreans, Oromo, and
Somali—who kicked against the centrism of ‘old Abyssinia’ and in so
doing were in fact demanding a new perspective on the region’s history.
The most noteworthy branch of this was the emergence of Oromo nation-
alist literature,5 but there were others and, increasingly, anthropologists
(and to a lesser extent historians and political scientists) worked among
the smaller minorities of the southern and south-west fringes: those on
the borders of Kenya and Sudan and squeezed between self-possessed
hegemonic blocs. In many respects this amounted to an examination of
the peoples living on Menelik’s southern battlegrounds of the late nine-
teenth century. This coincided with studies of the ethnic federalism of the
post-Derg state, some of which sought to historicize Ethiopia’s multi-
ethnic challenge; at the same time, however, those writing on Eritrea and
Somalia soon had new tests of their own. Somalia collapsed in the early
1990s, descending into an apparent chaos which was hotly contested by a
series of ‘warlords’ and which was only partially alleviated with, first, the
emergence of Somaliland as a comparatively stable would-be state in the
north, though lacking international recognition at the time of writing;
and, second, the emergence of the Islamic Courts Union, who were
16 set t i ng and ap p roac h

considered threatening enough to warrant military action by Ethiopia at

the end of 2006. But, basically, Somalia—the supposedly heterogeneous
ethnic nation in Africa par excellence—had ceased to be. Further north,
while the first few years of Eritrean independence created many new
admirers and caused many an enthusiastic thesis to be written on the vir-
tues of EPLF-led nationalism, the war with Ethiopia, and the political
oppression which swiftly followed, led scholars to either wonder where it
had all gone wrong or to write ‘told-you-so’ pieces, depending on where
they had positioned themselves in the 1980s.
In developing its particular interpretation of the region’s history, then,
this book is building upon, or has been inspired by, a range of scholarship
both old and new. Several key works need to be noted at the outset, most
of them clustered, not coincidentally, between the mid-1980s and the early
1990s—a period of radical rethinking about the contours of, and approaches
to, north-east African history.6 One scholarly collection of major signifi-
cance in the reinterpretation of Ethiopian history was The Southern Marches
of Imperial Ethiopia, published in 1986 and edited by Wendy James and
Donald Donham, which signposted a number of important shifts.7 The fol-
lowing year, John Markakis published his immensely valuable regional
study, National and Class Conflict in the Horn of Africa, which, while it is now
a little dated in certain aspects, was extremely prescient in drawing out the
interconnectedness between, and striking array of, conflicts emerging
across the north-east African region, in Ethiopia, Sudan, Somalia, and
Eritrea.8 The scholarly deconstruction of the Ethiopian region continued
apace, and two further studies were significant. The first, though perhaps
less well-known, was Bonnie Holcomb and Sisai Ibssa’s provocative The
Invention of Ethiopia, which gained in intellectual courage what it may at
times have lacked in scholarly rigour.9 It was among the first academic
shots fired on behalf of the Oromo, studies of which were beginning to
demand a revised history of the Ethiopian imperium. John Sorenson’s
equally provocative and stimulating Imagining Ethiopia argued essentially
for the demystification of ‘Ethiopia’, and a more honest analysis of what it
represented to various parties both within and outside the country itself.10
Finally, what was in effect the sequel to James and Donham’s Southern
Marches—Remapping Ethiopia—was published in 2002, in which James and
Donham were joined as editors by Eisei Kurimoto and Alessandro Triulzi.11
The title encapsulated the spirit of the body of work which had emerged
over the previous 15 years.
 i nte rp ret i ng the re g i on 17

Meanwhile, again, feeding into this revisionism was the burgeoning lit-
erature on, most notably, Eritrean and Oromo nationalisms—very different
beasts, no doubt, but each responsible in their own way for the remapping
exercise taking place through the 1980s and 1990s. It is inadvisable at this
stage to attempt even a brief survey of this canon of work—we will have
occasion to examine and utilize it in due course—but suffice to say it was
of immense importance in pioneering the unravelling of the many mythol-
ogies at heart of Ethiopianism. Inevitably, of course, such revisionism has
created mythologies of its own, not least the result of the aggressive opti-
mism which lay at its core, and in different ways Eritrean and Oromo
nationalist revisionism has each encountered serious problems—the former
because of the failures of the Eritrean state,12 the latter because of the deep
fissures within Oromo studies and the Oromo struggle13—but this scarcely
detracts from its historiographical significance.
Taken as a whole, this body of literature, disparate though it was in many
respects, signified a radical shift in regional historiography, typically charac-
terized as moving away from centre-biased analysis and toward the ‘periph-
eries’, borderlands, and the constituent yet marginalized parts of the old
Ethiopian empire. The current work builds on the critique initiated by
these volumes of the ‘centrist’ or ‘Greater Ethiopia’ school, exemplified by
Donald Levine’s glibly optimistic thesis—recently revised but essentially
unchanged a quarter of a century after it first appeared—that Ethiopia has
been and can continue to be a contented amalgam of ethnicities, cultures,
and languages.14 My own view is rather more sympathetic to that expressed
by Ernest Gellner, who, echoing Lenin’s remarks about Tsarist Russia,
asserted that ‘[t]he Amhara empire was a prison-house of nations if ever
there was one’.15 But this was an overly simplistic, if pithy, summation, and
our understanding of Ethiopian history, politics, and culture—and that of
the surrounding region—has become rather more sophisticated since the
1980s—at least, it has changed quite dramatically. In part, this has been a case
of scholars catching up with events on the ground, certainly in the early
1990s with the overthrow of Mengistu in Ethiopia and Siad Barre in Somalia,
and the appearance of an independent Eritrea. In many respects the ELF
and then the EPLF in Eritrea had led the way in this exercise in regional
‘remapping’, violent pioneers whose military success had forced the most
dramatic reconfiguration of political reality and scholarly received wisdom
alike; in many respects, perhaps, they have never really been forgiven, in
certain quarters at least. At any rate, while just a few years earlier most had
18 set t i ng and ap p roac h

been fixing their gaze, essentially, on Addis Ababa and Shoa, and seeing
the region from the viewpoint of this particular centre, now scholars and
commentators witnessed the encroachment of the so-called ‘periphery’ on
the centre, and understood that the history of the region was rather more
nuanced than had often been supposed. The Ethiopian empire-state would
be dissected by scholars and soldiers alike.
It is wholly understandable, and very welcome in many respects, that so
much work has tended to focus on the modern era; it has, however, come
at a price, namely a neglect of the deep past and la longue durée. Pre-colonial
African history more generally has, of course, experienced something of a
dramatic decline since the 1970s, and no doubt scholarship on north-east
Africa reflects that trend. In north-east Africa, of course, the concept of
‘pre-colonial’ is problematic. It has no applicability in Ethiopia itself, unless
we wish to talk of the various subject peoples of the south and west in these
terms, a people who have only recently begun their struggle against Amhara-
Tigrayan domination—and many do, of course. More conventionally, work
on ‘pre-colonial’ Eritrea is virtually non-existent, for various reasons,
although a recent special edition of the Eritrean Studies Review on ‘Eritrea on
the eve of colonial rule’ suggested what might be done.16 Archaeological
work in Eritrea is in its troubled infancy.17 Meanwhile, those interested in
the pre-colonial history of the Somali must continue to rely on the most
recent edition of Ioan Lewis’ classic monograph.18 As for Ethiopia itself,
work on the pre-1900 period has been reduced to a trickle. Donald
Crummey’s recent weighty study of land and society was the product of
many years’ research,19 while the 1986 Donham and James volume con-
tained some important insights into the late nineteenth-century peripheries
of imperial Ethiopia.20 The Oromo in particular have begun to be studied
in these terms, not coincidentally in the era of ethnic federalism in Ethiopia
and of continued Oromo insurgency.21 Many have attempted to historicize
this conflict, sometimes very self-consciously—although it is interesting
that this was never really attempted in the Eritrean context,22 doubtless
reflecting the liberation movement’s own somewhat inhibited view of the
past. Certainly, it is as though the big states had been ‘done’, and there was
a brief flurry of attention directed toward smaller, less ‘visible’, but increas-
ingly armed ethnicities in the Rift Valley, in northern Kenya and southern
Ethiopia in the early and mid-1990s. Contributors to Fukui and Markakis’
examination of ethnic violence in the Horn, or Kurimoto and Simonse’s
volume on age systems across the region, often rooted their analyses in the
 i nte rp ret i ng the re g i on 19

nineteenth century.23 As for the dramas of the nineteenth-century Ethiopian

highlands, the last word, it seemed, had gone to the generation working on
these in the 1960s and 1970s, with only fast-moving, cinematic retellings
appearing on bookshelves, such as, most recently, Philip Marsden’s version
of the Tewodros story.24 However, exceptionally, and rather more impor-
tantly, James McCann’s work on Ethiopian environmental history incorpo-
rated the nineteenth century, and—as was only appropriate to the
subject-matter—carried the analysis into the late twentieth century,25 and
Izabela Orlowska’s doctoral research on Yohannes IV should be mentioned
here, too.26 Clearly, it is not that nothing has been done, but that there has
been so little of it.The problem is that, as a result, there has been a failure to
more deeply historicize some key phenomena so apparent across modern
north-east Africa, not least warfare.
I have argued elsewhere that the role of war in Africa’s deeper history
has been a comparatively neglected subject to date, but that the Ethiopian
region is rather better served than most.27 In our region, warfare has been
given its historical place in a number of studies concerned with the expan-
sion of the Solomonic state, particularly bellicose (and successful) mon-
archs, and the necessity of defence against a range of external enemies,
culminating in those from Europe. In much of the Ethio-centrist literature,
indeed, all roads lead to Adwa, and in almost no other corner of pre-
colonial Africa has a battle been awarded such nation-building significance
as to be virtually European.28 More generally, Ethiopia since the ‘Solomonic
restoration’ has been defined as a militaristic polity, smiting enemies and
imposing itself—and its grand (written) narrative—on the highland envi-
ronment. Some excellent work has indeed been produced on soldiers,
weapons, and battles, and the current study owes much to them. But these
older celebrations of both sword and pen in the great sweep of Ethiopian
history were rarely concerned with analyses of social and political violence,
or with the economic, cultural, and ethnic dimensions to war which was
itself hardly problematized. It is also true that, as noted above, the role of
war and violence in the pre-1900 past has largely been set aside in favour
of more contemporary analyses of violence, more often undertaken by
anthropologists than historians. Once more, this has reflected concern for
actual events. The collapse of the Ethiopian and Somali states—temporary
in the case of the former, rather more enduring for the latter—in the early
1990s was at least in part brought about by localised, anti-state, ethnically
cohesive (or ethnically inspired) violence; as the 1990s progressed, a host of
20 set t i ng and ap p roac h

small but often deadly armed organizations grew up alongside older

movements such as the Oromo Liberation Front to challenge the existing
order. Meanwhile the supposed pax between Eritrea and Ethiopia proved
to be a mere suspension of hostilities.
A great many of these conflicts have been examined in considerable
depth by specialists, and again this book owes much to their analyses.
However, while these studies often lay stress on the essential novelty of such
violence—environmental catastrophe, the failure of the state and of state-
building projects, and the conjunction of arms and newly radicalized ethnic
and/or ‘national’ identity, all since the 1970s—it is my contention that these
conflicts are in fact much more deeply rooted in the past than might be
assumed, and are indeed intrinsic to understanding the political and social
evolution of the region. In essence, like some previous scholars of north-east
Africa, I too believe in continuity; unlike them, however, I do not perceive
a continuous state, but rather a continuous state of violence upon which is
founded the modern polity—whether Ethiopia, or Eritrea, or indeed Sudan
and Somalia—which is in turn defined by violence.The nation-states which
have emerged in the early years of our millennium are built less on negoti-
ated settlement—although inevitably this is a core component of state-
building—and rather more on cyclical conflict whose periodic cessation is
only part of the political and social equipment seen as necessary to the con-
solidation of those states. Such equipment has been refined over the last two
centuries or more; and, within the states themselves, actors who see them-
selves as marginalized continue to draw on past experience and to consider
violence as the natural means to ‘resolution’.

The fertile frontier: borderlands and frontier

societies in north-east Africa
The title, ‘frontiers of violence’, alludes to the conceptualization of the
region’s history over the two centuries which is at the core of this book. It
is contended here that the region can be characterized in terms of tecton-
ics—a mosaic of fault lines and frontier zones, shifting borderlands which
are not ‘peripheries’ but which have defined the very nature of the states
and societies themselves. As communities compete over ideologies and
resources, frontiers of violence have opened up across the region: sometimes
 i nte rp ret i ng the re g i on 21

the communities pre-date the frontiers, which are thus formed by expanding
polities, at other times the frontiers have emerged first, and serve to forge
the communities. In that sense we are interested in the ways in which par-
ticular societies have grown up within and because of the violent frontier—
Oromo expansion, for example, or the liberation struggles in the north in
the twentieth century—and also how other (often larger) states have been
fundamentally shaped by the experience of these frontiers: i.e. states are
ultimately defined by their turbulent borderlands, which are thus not
‘peripheral’ but are seedbeds, zones of interaction which are as constructive,
creative, and fertile as they are destructive and violent. Thus the concept of
the fertile frontier is true often in a literal sense—borderlands are the result of
societies and cultures pushing into resource-rich lands—but it is especially
true in the sense that out of these frontiers come new ways of being, of
organizing, and self-perceiving and perceiving others. The fertility of the
frontier—the vitality of violence, if we prefer—is crucial to understanding
the region’s modern history; tectonics helps us rethink the political con-
figuration of the ‘Greater Horn’ over la longue durée in terms of fault lines, at
times dormant, at other times explosive, as states and societies and cultures
invent and reinvent themselves according to current political and economic
exigencies.
Although it is hoped that this model may be more widely applicable to
other parts of the world which are clearly historically turbulent, our region
does have some unique features—or at least a combination of dynamics
which renders the region both particularly violent and markedly creative:
its peculiar historical experience of contemporaneous European imperial-
ism and African imperialism; the very experience of ‘African-on-African’
imperialism in the form of modern Ethiopia; particularly deeply rooted
(and literate) ideologies of governance, ethnic election, and social organi-
zation (among the Tigrinya, Amhara, and Somali); as a result, especially
clearly defined and historically rooted (even if at times shifting) notions of
ethnic community; the presence of competing global faiths; remarkable
geographical and thus cultural diversity within a historically discrete and
coherent zone of interaction; proximity to the strategically and commer-
cial vital southern Red Sea zone; climatic variability, thus giving rise to
fierce competition over resources; large-scale and long-term population
movement across the region.While many other regions have had a number
of these features in place, few have all of them, and this is what renders
north-east Africa so dynamic and its frontiers of violence so fertile.
22 set t i ng and ap p roac h

The analysis presented here owes much to Frederick Jackson Turner’s

seminal thesis—that the moving frontier had fundamentally defined the
history of the United States, shaped its political and social structures, and the
ways in which Americans thought about themselves—and the criticisms that
have been made of it over several decades. My own interpretation has like-
wise drawn inspiration from Dietrich Gerhard’s revisiting of the Turner the-
sis in 1959,29 and, above all, from Kopytoff ’s work on the African frontier and
political culture.30 The study of borders and frontier zones has long been, in
different ways, of interest to scholars of north-east Africa, but there is no
doubt that in recent years it has attracted more attention.31 I have attempted
to build on some of Kopytoff ’s core ideas in relation to north-east Africa,
which provides an excellent laboratory for the purpose. It is important to be
clear about our typologies, which are basically three.The first is the squeezed
zone between states, contested areas between the expanding imperial fron-
tiers of competing polities. This zone produces militant identities and politi-
cal cultures and frontier mentalities of its own, built on regional fault lines.
Second, we have deep-rooted ethnic or ‘national’ boundaries, which can be
both internal and external, and which can shift continually—in terms of
categories of population and the location of the frontier. Thirdly, there are
the physical frontiers and borderlands between environmental and economic
zones (and even within these); and these clearly make a contribution to the
formation of the first two typologies. Periods of economic and environmen-
tal stress rendered these frontiers highly volatile and competitive zones, where
human suffering but also economic opportunity and political creativity co-
existed. It is also the case, of course, that insofar as borders are ‘shifting’—al-
though it is argued in this book that the broad frontier zones remain
remarkably consistent over two centuries or more—some borders or fault
lines are ‘dormant’ at particular moments while others become ‘live’. This is
resonant, clearly, with the idea of identities also shifting. The margins make
the metropole: they fundamentally influence and shape it, place overt and
covert political and economic and cultural pressures on it, while, of course,
often the metropole itself is from the margins originally, or otherwise a direct
product of it. The very nature of society in north-east Africa, thus, is defined by its
rough, contested edges. The border reflects the essence of the state itself, in
Ethiopia and Eritrea, and indeed Sudan.
Frontiers, then, frequently produce highly militarized societies—especially
when we recognize the frontier not simply as a no-man’s-land between
mature polities, but as a fault line and, often, a contested zone. Such a zone
 i nte rp ret i ng the re g i on 23

of conflict and competition—and this does not preclude the zone being
a conduit, or indeed a ‘land of opportunity’—produces particular kinds of
societies and degrees of conflict; the fault line is representative of deeper
political tectonics, and this is clearly demonstrated across north-east Africa.
Again, as will become clear, I will argue for the interpretation of the
Eritrean–Ethiopian frontier zone as the epicentre of much of the region’s
violence. It is not, I would suggest, overly deterministic to argue that the
fault line does indeed have a profound impact on the societies and com-
munities which grow up there: the Eritrean frontier zone, for example,
demonstrates this very well. Here, we can see the national liberation move-
ment of the late twentieth century as the product of a much deeper com-
petitive and contested environment. More broadly, Ethiopia and its
immediate neighbours constitute a region which may be described—echo-
ing, but in many respects distinct from, Gellner’s idea about the prison-
house of peoples—as a mosaic of fault lines and frontier zones. This violent
competitiveness, and importantly its antiquity, in turn explains the strength
of ethnic identity and group consciousness across the region. Moreover, as a
region relatively free of the processes of colonial invention, it offers us
insights into African political and ethnic evolution which might be possible
elsewhere, were it not for the fact that scholars are unable to study local
constellations for the bright lights of colonialism. It is also the case that the
frontier zone itself exercises a profound influence on the wider region—it does not
merely exist of and for itself. Metropoles are frequently defined by what hap-
pens in their adjacent frontier zones—it is not merely a matter, as Koptyoff
implies, of pioneers from the metropole going forth and building new soci-
eties there. These fault lines are merely the surface manifestation of deeper
political tectonics, and thus shape profoundly the kind of ethnies and com-
munities and nations which press in on them from every direction. In this
context do we thus recall the work of Fredrik Barth on how ethnic com-
munities define themselves according to their boundaries, and what lies
beyond them32—and, I would suggest further, what happens within those
frontier zones and corridors of competition. In these places, political and
cultural creativity flourish and violence has a vitality that is sometimes
unpalatable to the squeamish observer; to be sure, such creativity has often
had brutal results, but it has been a crucial part of the process of invention
and formation and, ultimately, construction as well as destruction.
 2
The Shadows of Antiquity

Axum and its aftermath

The Axumite empire has cast long shadows over the region’s history, and
these can be understood from two perspectives. First, there are the very
tangible ways in which the ideological, political, and ethnic nature of Axum
had repercussions long after the state’s demise, and of course the manner in
which it actually collapsed likewise had a profound impact on the shape of
political and social relations across the region. Second, the undoubted intel-
lectual and material legacy of Axum has been claimed and contested by a
number of successive regimes seeking—in different ways—both continuity
and legitimacy. The partial mythology—and it is only partial—persists that
‘Ethiopia’, in its medieval, early modern, and more recent manifestations, is
the successor to Axum, and the rightful inheritor of its physical and intel-
lectual legacy. Yet even within this collective inheritance, tensions have
existed over whether the southern Semites, the Amhara, can really lay any
claim to what was in essence a northern Semitic, Tigrinya civilization; cer-
tainly, few groups south of the tenth parallel would consider themselves to
have been part of the supposedly seminal cultural experience that was
Axum. Lately, however, a new challenge has come—as in so many other
respects—from north of the Mereb River. Eritrean independence and con-
sequent nationalist revisionism has meant the split of the Axumite inherit-
ance across an international frontier, and a violently contested one at that,
and the assertion that Axum was at least as much an ‘Eritrean’ as an ‘Ethiopian’
civilization, notwithstanding the fact that these appellations are meaningless
for the period under discussion. Recent discoveries of pre-Axumite sites in
Eritrea caused much excitement among archaeologists,1 and apparently
some bitter resentment on the Ethiopian side of the border: Eritrea’s extant
archaeological inheritance came under attack by Ethiopian soldiers in 2000,
 th e shadows of ant i qu i ty 25

who attempted to destroy an important pre-Axumite stele at Metara.2 It

seemed that the battle over boundaries, identities, and regional hegemony
had also become one of who had a stronger claim to antiquity—a curious
juxtaposition alongside the revolutionary militancy of the former guerrillas
in power in Asmara and Addis Ababa.
Axum had been one of the great civilizations of the ancient Eurasian
world, with its roots in an indigenous agrarian culture with links across the
Red Sea to southern Arabia between the fifth and third centuries bce.
Axum’s mercantilism had a violent aspect, and royal inscriptions from the
fourth century onward attest to military campaigns across the Red Sea,
leading to the establishment of military garrisons in southern Arabia, to
wars against the Nubian kingdom of Meroe in central Sudan, and to the
conquest of various peoples in the immediate vicinity of Axum itself. Ezana
was the self-styled ‘king of kings’—an epithet which would be inherited by
the emperors of Ethiopia many centuries later—while Axum itself, domi-
nant between the second and eighth centuries ce, was in truth one of a
number of states vying for power in the southern Red Sea region. It was a
distinctively Semitic, highland civilization which made much of military
glory—few do not, of course—but which also, in time, made much of its
proselytizing Christian mission across the wider region. In religious centres,
such as Debre Damo, a new monastic zeal emerged around the principles of
communal living, obedience, and self-discipline and was introduced into
the provinces of Wag and Lasta, south of Tigray;3 Christianity was also grow-
ing among the non-Semitic Agaw speakers of northern Wollo. Already by
the middle of the first millennium ce, the interaction between Christians
and non-Christians in the northern highlands had an ethnic element to it,
and the notion of what it was to be Axumite was being formed very sharply
along the lines of religion, ethnicity, and language.
These definitions would become all the more marked in the centuries
which followed, the period of Axum’s slow, apparently inexorable decline.
Out of the violence and chaos of Axum’s collapse emerged the Zagwe
dynasty, itself Agaw—a problem in terms of posterity, as we see below. In fact
the Zagwe era, beginning in the mid-twelfth century, was one of the most
remarkable in the region’s early history.4 The Zagwe presided over an ener-
getic Christian expansion through the northern and central Ethiopian high-
lands, as well as a notable commercial and cultural interaction with Egypt
and the wider Middle East; the dynasty was also responsible for the creation
of some of Ethiopia’s most remarkable architecture. Despite their considerable
26 set t i ng and ap p roac h

achievements, however, weaknesses in the Zagwe state had begun to appear

by the early thirteenth century. Unable to forge broader unity, the Zagwe
were also undermined by their own repeated succession disputes, which in
turn encouraged the emergence of anti-Zagwe movements among the
Semitic-speaking Tigrayans and Amhara; they were, in many ways, the ‘natu-
ral’ ethnic and cultural enemies of the Cushitic, Agaw-speaking Zagwe, and
certainly in terms of their own self-image the rightful claimants to the
Christian Axumite inheritance. The most potent challenge ultimately came
from the Christian community in Shoa, which had grown prosperous from
the eastbound trade routes and which—critically—had the support of the
local church. The Shoan rebellion started in 1268 and was led by Yekuno
Amlak, who won a series of battles across Lasta and Begemedir culminating
in the death of the last Zagwe king in 1270.Yekuno Amlak declared himself
ruler and set about creating a new moral and ideological order.

Violence and mythology: the new Zion

The non-Semitic origins of the Zagwe would allow the insurgents of the
1260s and subsequent chroniclers to talk of their illegitimacy. While the
Zagwe had sought legitimacy by creating the myth that they were descended
from Moses, and assiduously sought moral authority through impressive
churches of rock, none of this impressed later chroniclers, who probably
destroyed much of the dynasty’s own literary endeavour and depicted the
Zagwe as lacking any ‘Israelite’ connection. In order to assert its own legiti-
macy, the new dynasty claimed to be a restoration of the ‘Solomonic’ line.
It sponsored the narrative at the centre of the Kebre Negast (‘The Glory of
the Kings’5) that the rightful rulers of the Christian kingdom must be
descended from the union of King Solomon and Makeda, the Queen of
Saba, or Sheba; the Zagwe had represented an illegal hiatus, while the newly
restored Solomonic kings were associated with the ancient and glorious
kings of Axum. In real terms, it seems fair to say that only insofar as the mon-
arch was once again a Semitic-speaker can Yekuno Amlak’s advent to power
be considered a ‘restoration’; but the Kebre Negast was a powerful vindication.
Its central point was that the ‘greatness of kings’ was related to possession of
a sacred emblem (namely the Ark of the Covenant), and a blood connection
with a God-anointed elect. While some scholars have dated at least some of
 th e shadows of ant i qu i ty 27

the text to the sixth century, it is more likely to have originated in the thir-
teenth century as a piece of anti-Zagwe propaganda, and was redacted in the
early years of the fourteenth century. In many respects, it is a composite
work, incorporating oral and written traditions from both the Ethiopian
region and the Middle East, but above all it was used to bathe the new state
in military glory and legitimacy, and hail the ‘restored’ lineage. In fact, the
early decades of the new dynasty’s existence were precarious, with Yekuno
Amlak facing much opposition, and it was only under his grandson, Amda
Tsion (1312–42) that the state was able to establish itself firmly in the Christian
heartlands and indeed embark on expansion.6 The Kebre Negast was therefore
crucial in terms of legitimization, for who could argue with a kingship
descended from Solomon, and his son Menelik, who had taken the Ark of
the Covenant to Ethiopia? Thus was made explicit the covenant between
God and Ethiopia, which was now Israel’s successor, the new Zion; habesha
Christian highlanders were developing a distinctively chauvinistic view of
themselves and their geopolitical environment. Yet there was perhaps more
to it than meets the eye.The ‘final’ redaction of the text was undertaken by a
team of Tigrayan monks in the early fourteenth century, at a time when the
north was in revolt against Amda Tsion; one argument, therefore, is that the
Kebre Negast was actually a Tigrayan script, a northern riposte to southern
‘usurpation’, namely that of the increasingly ambitious and hegemonic
Amhara.7 Tigray, after all, had a much stronger claim to the Axumite inherit-
ance than any other parvenu Christian province.
Whatever the case, the Kebre Negast provided the foundations for a
uniquely habesha polity; the ‘nation’ thus born in its narrative was superior
to all others in faith, in culture, in blood, and, blessed by God, Ethiopians
now had a mandate to carry fire and sword among pagans and Muslims
alike across the surrounding area. Their violent expansion was just and
righteous, a form of holy war which defined the Solomonic state in the
centuries to come. Much of this war was documented in glorious detail in
royal chronicles, composed in the style of the Old Testament and redolent
of the trials and triumphs of the early kings of Israel.8 These were doubtless
inspired, of course, by the Kebre Negast itself, which is a violent book: soon
after becoming king, Menelik I—also David II—
waged war wherever he pleaseth, and no man conquered him, but whosoever
attacked him was conquered, for Zion himself made the strength of the enemy
to be exhausted. But King David II with his armies and all those who obeyed
28 set t i ng and ap p roac h

his word, ran by the chariots without pain, hunger or thirst, without sweat and
exhaustion . . .

In this way, successful wars are waged against enemies as far flung as Egypt
and India.9 Righteous violence became legitimized through Solomonic
mythology and came to occupy a central role in habesha political discourse
and action, as real and as vital to nineteenth- and twentieth-century rulers
as it was to early Solomonic monarchs. From the outset, moreover, the
Solomonic state was wedded to a militant monasticism, which pushed for-
ward the frontiers of Christianity as the state itself expanded. Monks and
monasteries were as much at the forefront of habesha imperialism as soldiers
and governors.
The Christian state increasingly defined itself against a range of external
enemies, and indeed its rivalry with the Islamic states and societies to the
east and south-east was prolonged, and episodically violent. In the first
instance it was concerned with the control of trade routes which fanned out
from the port of Zeila on the Gulf of Aden, and around which a series of
Muslim settlements were founded, reaching as far as south-east Shoa. Islam
had been gradually introduced to north-east Africa via the Somali coast,
linking the region with Muslim trading networks which were stretching
out across the western Indian Ocean from the Persian Gulf.10 The early
Solomonic state had undoubtedly accrued considerable benefits from work-
ing with Muslim merchants—notably in terms of the slave trade—but the
trade routes themselves were soon the focus of fierce competition between
Christians and Muslims, and Amda Tsion’s victories toward the east meant
that Muslim settlements were required to recognize his suzerainty. Islam was
disunited in the region—many Muslims were content to accept Solomonic
hegemony, as it meant commercial gain in some respects, while there was
also bitter rivalry between settlements as well as ethnic diversity—and the
key sultanate of Ifat, under the Walasma dynasty, was ultimately overcome in
the 1330s. But a radical anti-Christian element survived in Ifat, and sparked
into life in the late fourteenth century when an abortive coup attempted to
overthrow the sultan, Ali, who was seen as overly friendly with the
Ethiopians. Ifat, indeed, collapsed temporarily under the weight of these
disturbances; the Walasma family took refuge for a time in Yemen, but they
later returned to found a new settlement further east, close to present-day
Harar. Meanwhile, by the early fifteenth century, another Islamic state, Adal,
had become established nearby—in the area of modern north-west Somalia,
 th e shadows of ant i qu i ty 29

southern Djibouti, and north-east Ethiopia—but out of range of direct

Christian control, if not of their marauding armies which visited regularly
in the course of the fifteenth century.
Adal would become the focus of concerted Islamic resistance against
habesha domination of the region.11 The Ethiopian defeat of Adal’s ruler
Muhammad in 1516 led to a crisis for the sultanate: as the capital was trans-
ferred to Harar, Ahmad ibn Ibrahim al-Ghazi, a prominent soldier, emerged
as kingmaker, having Sultan Abu Bakr killed, putting his brother in his place,
and having himself titled Imam.The broader context was the rise of Ottoman
power in the Red Sea region, and with it a more confident, aggressive Islam,
and contemporaneously a weakening of Ethiopian power stemming from
problems at the centre through the late fifteenth century. Imam Ahmad was
a messianic leader, successfully preaching the need for jihad against the
Christian state among the Somali and Afar—although quite how much
these Cushitic soldiers were moved by religious fervour is a matter of debate,
and environmental factors may have been at least as important. He urged
self-discipline and abstinence, and his militant message—the purification of
Islam in the region, and the wresting of the Ethiopian state from the unbe-
lievers—received support (and doubtless inspiration) from imported mis-
sionaries, adventurers, and activists from the Arabian peninsula. The links
between nascent Somali Islam and Arabia were clearly already significant. In
the course of the 1520s, Ahmad—also known as ‘Gran’, the left-handed—
reorganized Adal, revitalized the ruling dynasty, and set about the systematic
recruitment of the Somali, who had become increasingly Islamized over
several decades; and in 1529, he led his jihad into the Ethiopian highlands
armed with Ottoman guns and temporary Muslim unity. As his army
advanced, it was joined by Cushitic communities hopeful of regaining their
independence from Solomonic control, and success came quickly: by the
mid-1530s, the Islamic forces had swept into the Ethiopian heartlands. The
Ethiopian emperor, Lebna Dengel, however, remained at large, and fought a
rearguard action until his death in 1540; and before he died, he succeeded in
sending a desperate message to the Portuguese, his new allies with whom he
had been in contact since around 1520 and who had themselves recently
arrived in the southern Red Sea area. The Portuguese contributed a small
force equipped with firearms and Ahmad’s increasingly over-stretched army
was decisively defeated in 1543. The Christian church was thus preserved,
and the image of ‘fortress Ethiopia’ underlined. But much damage had been
done. In the 1550s, as Galawdewos tried to refashion the Solomonic state
30 set t i ng and ap p roac h

with the remnants left to him, Ethiopia was characterized by a loss of con-
fidence as well as much material damage and immeasurable human losses.
Certainly, the events of the sixteenth century were burnt into the ‘popular’
and ‘institutional’ memories of Ethiopians, as we see in Part II; they also cast
a shadow over future Christian-Muslim relations, however economically
interdependent adherents to the two faiths were. But Christian Ethiopia
remained conscious of its precarious position surrounded by Muslim ene-
mies—something which would echo down to the present day in terms of
Ethiopia’s internal politics, and its relations with neighbouring states, nota-
bly Somalia and Sudan.
Between the sixteenth and the eighteenth centuries, the shrunken
Ethiopian state attempted to stabilize and consolidate, and for a time—
during the Gondarine period—it did so. But centrifugal tendencies, an
over-powerful regional nobility, and lack of clear succession rules weakened
the Solomonic state, which was also fundamentally transformed by the
arrival of the Oromo. The Oromo ‘appeared’ in the habesha line of vision in
the aftermath of Ahmad’s jihad, were of increasing significance in the remak-
ing of the Solomonic state at Gondar, and were instrumental in the demise
of the monarchy and the forging of a new era—and ultimately, of a new
political reality. Although the Oromo migrations were only one factor in a
crisis-laden and bloody sixteenth century, their impact would endure.12
Oromo played many roles, and came in many guises; their past, indeed their
present, is contested, and it is to their story that we must now turn.

The Oromo frontier

There can be few peoples in African history who have been as misunder-
stood, and indeed as misrepresented, as the Oromo—or pejoratively ‘Galla’ in
the older literature. They have been, arguably, even more demonized by
Ethiopian chroniclers of various hues and over a longer timeframe than the
Somali, historically the other great rival ‘bloc’ confronting the Amhara in
north-east Africa. The tone was set early on by the monk Bahrey, whose
‘History of the Galla’ was produced in 1593.13 In many respects Bahrey’s
‘History’ tells us rather more about the problems of the Solomonic state than
it does about the Oromo, as the latter are juxtaposed throughout against the
society within which Bahrey was writing. He explained Oromo success as
 th e shadows of ant i qu i ty 31

being due to the failure of the Ethiopian state to properly mobilize its
resources, for example; he bemoaned the fact that Solomonic society com-
prised a number of privileged classes which no longer did any fighting, or
indeed much of any use at all, while the Oromo were dedicated to warfare
and moved ‘as one’ into the highlands, an unstoppable force both numerous
and ferocious. Bahrey exaggerated both their numbers and their ferocity, but
he created images that endured. By the nineteenth century it was de rigueur
for European travel accounts to contain historical prologues outlining the
savage invasions of the ‘Galla’, which gripped the European imagination and
became a central part of the narrative which described the Christian king-
dom as a highland fortress resolutely defending itself against savage neigh-
bours.There was something in the story that reminded some writers, at least,
of the decline of Rome, and Henry Salt could not resist the temptation to
consider the Oromo as the region’s Goths and Vandals. Yet they were also
Philistines to Ethiopia’s Israelites, and indeed ‘the feelings of the Abyssinians
towards the Galla partake of the same inveterate spirit of animosity which
appears to have influenced the Israelites with regard to their hostile neigh-
bours’.14 In another source, the Oromo were the ‘savage hordes’ which had
‘rent asunder the once powerful empire of Abyssinia, and arrested . . . the
progress of Christianity, civilization, and refinement’.15 They had their admir-
ers—Victorian Britons were ever keen to applaud noble savagery—and in
the 1860s Blanc considered the Wollo Oromo ‘a fine race, far superior to the
Abyssinian in elegance, manliness, and courage’;16 but rather more common
was the view expressed by Markham in 1869, who declared that the Oromo
were ‘untameable people, resolved either to conquer or die . . . They are a
cruel set of bloodthirsty robbers’. They had thrown ‘a once civilized people
more and more into barbarism and anarchy’, and the resultant violence had
‘plunged the wretched country deeper and deeper into anarchy’.17 Thus
were the Oromo made scapegoats for much that was held to be ‘wrong’ with
highland state and society in the nineteenth century.
The generally accepted account18 is that the prolonged period of destruc-
tive warfare in the first half of the sixteenth century had weakened the
Christian kingdom’s defences, leading to its shrinkage and consolidation
around a northern core, particularly under Sarsa Dengel between the 1560s
and the 1590s, and facilitating the gradual advance of groups of pastoral
Oromo into the southern highlands. Sarsa Dengel withdrew many of his
southern garrisons—which were increasingly becoming mere islands of
Solomonic ‘authority’ in the midst of Oromo populations—and redeployed
32 set t i ng and ap p roac h

them in the north, thus reducing the level of resistance the migrants might
otherwise have encountered. The Cushitic Oromo originated in the broad
zone between the grasslands north-east of Lake Turkana and the south-east
Ethiopian foothills, and in the course of the late sixteenth century they
began to move into the southern third of present-day Ethiopia. They also
moved in the vicinity of Harar. They were in search of better pastureland,
although their point of origin suggests that they also practised some agri-
culture; almost certainly, their spread—like that of the Somali—was related
to overpopulation and overgrazing. There was a significant number of
Cushitic speakers—mainly Somali—in the Adal forces of Ahmad ibn
Ibrahim. It seems reasonable to suppose that these jihad-ist fighters were in
fact part of a larger Cushitic movement. Many converted to either
Christianity or Islam, depending on the states and societies they encoun-
tered; however, it is the Muslim Oromo who are frequently portrayed in
contemporary and later sources as ‘barbarous hordes’, closing in on Christian
‘fortress Ethiopia’. Certainly, the advance of the Oromo into the central
highlands was to some extent related to the shift in Amhara settlement to
the north and west, but this was pull as well as push: the Amhara state was
repositioning itself toward the Red Sea coast, seeking trading contacts with
the Ottoman Turks at Massawa. Overall, it was a process of relocation that
was reflected in the creation of the royal capital at Gondar in the 1630s.
As they moved, they both assimilated others, and were themselves assimi-
lated.19 There was nothing uniform or monolithic about the movement
itself, of course: Oromo identified themselves as belonging to one or other
of the two main branches, namely Borana (west, south, and centre of the
Oromo population) or Bayratu (in the east), and then as members of one of
the numerous clans within those branches, and as members of particular
age-sets and lineages. But what they shared was a highly complex, sophisti-
cated, and distinctive system of social and cultural organization; it was a
system which would have an enormous impact on the peoples of the region,
including the highland cultures of the Amhara and Tigray. Initially, their
culture had its greatest impact in the military sphere, particularly through
their complex age-grade system, known as gada.20 Gada was remarkable for
its pervasiveness, affecting all aspects of society; and although it is as much
as anything else a philosophical tradition and the embodiment of a set of
cultural values, it is most commonly associated with a militaristic pastoral-
ism within which a man could gain honour and influence in society by
attacking neighbouring societies and killing. The gada system defined male
 th e shadows of ant i qu i ty 33

activities in eight-year segments. The pastoral nature of Oromo life meant

the necessity of a loose, decentralized, egalitarian society which was ulti-
mately led by officials who were elected by the particular gada rank respon-
sible for government.The qallu, leaders who represented the forces of nature,
exercised a loose authority over religious and political matters; they vali-
dated the leadership of the gada council drawn from a list of approved indi-
viduals. Oromo men, in order to reach the qallu grade, needed to have gone
through the key aspects of life, notably marriage and military service. Success
in the latter led to the former, very often, and so violence was periodically
exercised on neighbouring peoples in order to permit young men to prove
themselves. Certainly, one of the enduring historical images of the Oromo
is of militaristic, proud, pastoralists, given to bursts of controlled violence;
the influence of gada, arguably, was such that it led over the longer term to
a wholesale change in the organization of highland Christian society after
the sixteenth century.21
The mammoth clash between the Solomonic state and Adal had left the
way clear for Oromo expansion, especially into central Shoa which had
been situated on the fault line between Ethiopia and Adal, and which fol-
lowing the fighting constituted a depopulated borderland. Indeed Shoa
now became one of the key battlegrounds between the Amhara and Oromo,
and the Amhara-dominated parts of Shoa constituted a centre of anti-
Oromo sentiment and propaganda. Sarsa Dengel was more or less con-
stantly engaged in campaigns against the Oromo throughout his reign, and
although he won victories against them in pitched battles in 1569, 1573,
and 1578, in the longer term these were largely irrelevant: ‘they kept on
coming’, in Levine’s words.22 In the 1550s, Oromo were beginning to move
into both the former Christian provinces of the south, and the former
Islamic sultanates of the Rift Valley; by the end of the sixteenth century,
they were present in the heartlands of both Ethiopia and Adal. In time, they
moved into the west and south-west, where Oromo pioneers encountered
extant state systems which had been tributary to or influenced by the
Solomonic state; here, the Oromo were influenced by what they found,
and their pastoral egalitarianism was inverted in favour of agricultural and
commercial monarchies.23 These Oromo kingdoms asserted themselves
over the Sidama and Omotic peoples of the region, and would remain
essentially independent until Menelik’s campaigns of territorial expansion
in the 1880s and 1890s. Other groups of pioneers headed for the uplands,
their movement and settlement facilitated by local depopulation and war-
34 set t i ng and ap p roac h

weariness; as they encountered different physical environments, socio-

political differentiation developed. Again, there was no uniformity of
response to or interaction with the communities they met. Some Oromo
continued as pastoralists, others took to sedentary agriculture; many prac-
tised both. While large numbers of Oromo came to identify closely with
the host society, others held themselves apart, or only selectively adopted
certain aspects of the host society and culture. Importantly, where Oromo
pioneers settled and became Christian, it was increasingly rare for them to
act in political unison as Oromo, a description which only had meaning
below the public horizon—often only at the level of family; in time, they
intermarried with the Amhara and other Semitic groups, and such cultural
integration often led to the dissipation of the cultural frontier in the public
space of the Ethiopian polity.
It was certainly clear, in time, that the Oromo were permanent settlers,
especially across Shoa, Gojjam, and Wollo, where they continued to settle
through the seventeenth century. Strategies of cohabitation were required.
In the early seventeenth century, Emperor Susenyos implemented one such
strategy which aimed at the integration of the Oromo into the political life
of the Solomonic state; it was a project fraught with danger, although
Susenyos’ confidence doubtless stemmed from the fact that he had in fact
lived among the Oromo and had married the daughter of a gada official. He
sought to integrate Oromo military units into his army, and indeed used
these to subdue rival branches of the Oromo, such was their disunity.24 In
the late seventeenth century, Iyasu sought to bring more Oromo into impe-
rial service, and in this way did Oromo come to fill ever more prominent
roles in the habesha political and military establishment. On one level, it was
a highly successful policy of co-option on the part of the Amhara; but from
a different perspective, it was the Oromo who had co-opted the Solomonic
state, becoming a force to be reckoned with at the centre of imperial power
in the eighteenth century, and uneasily sharing power with the Amhara and
the Tigrinya—a situation which has persisted, in different guises, down to
the present day.
In Gondar and its environs there was a swelling Oromo population—
farmers, soldiers, courtiers, senior figures at the Gondarine court. New
Oromo populations flowed in on the eastern flanks of the kingdom, while
the state expanded into new borderlands north and west of Lake Tana.
While they may have been associated with recent disasters, they also repre-
sented a new and dynamic element in Ethiopian society. Oromo leaders
 th e shadows of ant i qu i ty 35

were increasingly powerful, as the monarchy came to depend on them to

shore up its position against the Amhara nobility. In the second half of the
seventeenth century, for example, Fasilidas had forged alliances with the
Oromo elite at Gondar during a theological controversy within the
Orthodox Church which had isolated the monarchy from many Amhara
noblemen. In the 1720s, Bakaffa had attempted to purge the court nobility,
and used newly armed Oromo as a counterweight to the increasingly unre-
liable imperial guard, as well as deploying them in the suppression of regional
revolt.25 But it was the long-lived, powerful matriarch Mentewwab—
Bakaffa’s consort and, after his death, the head of an extensive political fam-
ily known as the Qwarrana—who simultaneously buoyed the ailing
monarchy and sowed the seeds of its destruction by allying it with the pow-
erful Wollo Oromo through a series of political marriages.26 This intricate
network of alliances was supposed to protect the monarchy from a range of
enemies, within the court and beyond Gondar—many of whom questioned
the legitimacy of the Qwarrana, who had Agaw roots—but even Mentewwab
came to realize the dangers of making the Oromo such an integral part of
the political centre. She realized the power of Oromo king-makers too late,
attempting to develop a Tigrayan counter-balance to Oromo power through
Ras Mikael Sehul in the north. But in fact Mikael was himself able and will-
ing to take advantage of widespread (and perennial) habesha fears of immi-
nent decline, and playing upon the image of Gondar as the ailing heart of
empire, riddled with vice and infidelity, and, most important of all, full of
Oromo soldiers, courtiers and administrators who had no right to be there.
It was, at least in part, the ethnicization—indeed, racialization—of political
catastrophe which finally crushed the medieval Ethiopian state, and ushered
in the period of overtly violent political culture known as the zemene
mesafint; it also set the tone for much of what was to follow in terms of the
region’s political development.
This page intentionally left blank
 PART
II
Violence and
Imperialism
The ‘Long’ Nineteenth Century
This page intentionally left blank
 3
States of Violence, to c.1870

Haunted house: the wars of the mesafint

The conflicts which erupted in the 1760s and 1770s were a long time com-
ing. They finally shattered the façade of unity—of political and ethnic
homogeneity—much trumpeted at the apex of the medieval Ethiopian
state. Yet although the ‘basic’ narrative is reasonably well-known,1 this
remains an era on which much work needs to be done in order to fully
understand its rich social, political, and ethnic dimensions and implications.
The zemene mesafint was the crystallization of many of the evolving crises
experienced by the habesha polity dating to the mid-sixteenth century,
including the ever sharpening religious frontier between Christianity and
Islam, the regionalism and centrifugal forces which would ultimately
destroy the Ethiopian state, and the migration of the Oromo.The processes
and dynamics which were unleashed from the late eighteenth century,
moreover, would come to define the region’s history long after the sup-
posed ‘reunification’ of Ethiopia in the middle of the nineteenth century.
The patterns of violence, and the reasons behind such violence, exist to the
present day.
There is much we know about the zemene mesafint, but there is much we
do not; and it remains the case that scholars rarely connect the events of this
era to those of a later epoch.Viewed in bold colours, the century and a half
between the 1770s and the 1920s is supposed to represent some kind of
linear progression toward Ethiopian ‘modernity’, the latter manifest in the
functioning state fashioned by Menelik and Haile Selassie. The zemene
mesafint, from this perspective, is treated as a form of temporally determined
‘pre-modern savagery’, in much the same way as escalations of violent con-
flict in the nineteenth century have been regarded in other parts of the
continent. It is history as prologue—albeit a necessary one, as otherwise the
40 v i ole nc e and i mpe ri al i sm

essential modernity of the Ethiopian state-building exercise at the end of

the nineteenth century cannot be emphasized to the same extent—and
what happens next is, again, linear progression.The zemene mesafint is bloody
and terrible, and, at least implicitly, against the natural order of things; and
then comes ‘restoration’ and ‘reunification’, and Ethiopia not only ‘survives’
but expands. In some respects, the Ethiopian ‘model of modernity’ follows
that set up for European colonial rule elsewhere on the continent.
This was an era of the nobility’s dominance over the monarchy, and one
during which ‘Ethiopia’, in its medieval and early modern form, did not
exist in any practical, tangible shape—save for the fact that an emperor still
existed, lingering in Gondar in an impoverished state while politics rushed
on around him. Importantly, however, ‘Ethiopia’ did indeed survive as an
idea, powerful in the imagination—and it was this which caused contem-
porary chroniclers to despair about the awful illegitimacy of the violence
which defined the era, and lament that ‘the Kingdom has become con-
temptible to children and slaves . . . a laughing stock to the uncircumcised . . . a
worthless flower that children pluck in the autumn rains’.2 Imperial author-
ity was paralysed by intrigue and by an impossibly entangled web of tactical
alliances and counter-alliances. Whatever power was still wielded by the
emperor Iyoas, and indeed of the ageing matriarch Mentewwab, was
contingent upon the advantages which others could derive from royal com-
mands. Politics was increasingly a matter of ethnicity and of region. The
royal family, with Agaw roots, had sought to shore up its position through
intermarriage with other powerful groups, not least the increasingly impor-
tant Oromo of Wollo province.When it was feared that an ethnic imbalance
might destabilize the delicate arrangements at Gondar, redress was sought
elsewhere, notably in the form of the Tigrayan leader Mikael Sehul, who
had been appointed governor of the north. But Mikael was no more inter-
ested in the restoration of Solomonic glory than the Oromo nobility now
so central to the functioning of imperial power. He too sought political pre-
eminence, and indeed had the advantage—as governor of Tigray—of access
to coastal commerce and thus the ability to acquire firearms.3 Further south,
the Amhara may have been resentful of his being Tigrayan, but Mikael was
able to play on the fears harboured by traditionalists and conservatives that
Ethiopia was facing disaster, and at any rate he was, for many, a better option
than any of the Oromo now flooding the streets of Gondar as soldiers,
courtiers, and administrators. In Bruce’s indubitably embellished account,
Mikael at one point proclaims the ‘Galla’ ‘the enemies of [this] country . . .
 state s of v i ole nc e, to C. 18 7 0 41

I am much deceived if the day is not at hand when [the emperor] shall curse
the moment that ever Galla crossed the Nile’.4
Among the Oromo themselves there were tensions between those who
were ‘assimilated’ into the Solomonic state and those who were not, and
such strains exploded into open war in 1766. Out of desperation, and fear-
ing that Gondarine society would be destroyed by intra-Oromo strife, Iyoas
called on Ras Mikael to save the state. It was a plea which highlighted the
insoluble dilemma faced by the monarchy: unable to withstand the power
of regionalism, they had turned to the clearest manifestation of potent
regionalism to protect the Solomonic inheritance. He responded by mar-
shalling his army and defeating, first, the unstable royal faction, followed by
the Wollo Oromo themselves. Bruce recorded: ‘[T]here was no safety but in
Ras Michael’, who was now possessed of ‘supreme power, both civil and
military’.5 In 1769, Iyoas, quickly becoming fearful of the Ras’s strength,
ordered Mikael back to Tigray and attempted to form a coalition of anti-
Tigrayan forces, loyal Oromo included; Mikael refused, marched on Gondar,
and had Iyoas murdered. It was arguably the most significant coup d’etat in
the region’s history until the removal of Haile Selassie in 1974. In effect, the
Gondarine era was brought to a close, and in many respects the Solomonic
state was finished. In Gondar itself, emperors would come and go, some dal-
lying longer than others, but each one merely the titular head of an empire
which to all intents and purposes no longer existed. The zemene mesafint
may have begun as a ‘civil war’, but it became rather more than that, mor-
phing into bloody competition between territorially and, increasingly, eth-
nically defined polities, from the region of the Eritrean highland plateau
(the kebessa) and adjacent lowlands, down through Tigray and the central
Ethiopian Highlands. It was an era in which the power of Tigray grew,
portentously connected in no small measure to the slave and arms trades of
the Red Sea, and during which the importance of the Oromo in altering
the ethnic and cultural shape of the central highlands was profound. In the
meantime the Amhara provinces sought to hold their own against both; and
in the longer term, it was the ability of Shoa to supplant Tigray, marginal-
izing the latter province in political and economic terms, which defined the
later decades of the nineteenth century.
While the 1770s were characterized by a series of schisms within the
increasingly decrepit Solomonic order, in the course of the 1780s a faction
of families known as the Warrashek became prominent.6 They came from
the Yajju region of Wollo, and were ultimately descended from a group of
42 v i ole nc e and i mpe ri al i sm

Oromo families which had married into the imperial court in the course of
the sixteenth century. Although originally from a region in Wollo that had
been exposed to Islamic influence, the leading members of the Warrashek
faction were Christian by the late eighteenth century—although names
such as Ali clearly show an Islamic influence. Ali Gwangul was in many ways
the founder of the dynasty, establishing hegemony over the rump Solomonic
polity in the 1780s and 1790s. By the early years of the nineteenth century,
Ras Ali II of the Warrashek was the dominant figure, based at Debre Tabor
to the east of Lake Tana. Ali’s mother, Menen, was a powerful and charis-
matic figure, and belonged to an Ethiopian tradition of influential matri-
archs stretching back several centuries. In essence, the Warrashek represented
the dominance of the central Amhara provinces, the old heart of the
Gondarine empire, but new frontiers and conflicts were opening up.
Warrashek hegemony was increasingly challenged from the north, notably,
Tigray under Wube from the 1830s. For a time, Wube was a serious con-
tender for the Solomonic inheritance—had he won a major encounter
against the Warrashek at Debre Tabor in 1842, he might have been able to
claim it—but he failed, and retreated back north. Ali’s Amhara domain was,
for the time being, more powerful than Tigray, a fact at least tacitly acknowl-
edged by Wube; yet Tigray remained a major, and increasingly important,
player in habesha politics.
In addition to this essentially ‘northern’ struggle, challenges came from
the southern mountains in the form of Gojjam, lying south of Lake Tana,
and Shoa, further south-east.7 These historically Amharic provinces were
also deeply infused with Oromo immigration. Gojjam and Shoa were well-
positioned to gain access to the rich lands of the south—later the great prize
of imperial expansion—while Shoa in particular was also able to reach
toward the Red Sea trade at its southerly end from the early nineteenth
century. Gojjam and Shoa—the former under Berru and his son Goshu, the
latter under Sahle Selassie, grandfather of Menelik—began to accumulate
both political and economic muscle in the early decades of the nineteenth
century, and were the key rivals in the southern-central bloc of the zemene
mesafint. The Shoans in particular showed themselves interested in the
Solomonic restoration, an increasingly potent notion as the first half of the
nineteenth century wore on, and indeed Sahle Selassie was using the title
Negus by the early 1830s.Yet the more potent attempt to consolidate impe-
rial power and claim the Solomonic inheritance would in time come from
an ostensibly unlikely direction—the province of Kwara, closer than Shoa
 state s of v i ole nc e, to C. 18 7 0 43

to the old imperial heartland at Gondar, but a peripheral, borderland terri-

tory nonetheless, located in the foothills between the Ethiopian Highlands
and the central Sudanese riverine plains. There, Kassa Haylu—later
Tewodros—would gradually gather strength.
It is axiomatic to suggest that contemporary European sources must be
treated with caution in this context; ‘Western’ observers had a tendency to
exaggerate and misunderstand the level of violence they were witnessing,
and to emphasize above all else the prevalence of war, or what passed for it,
among savage tribes endlessly at one another’s throats.8 What they were
witnessing, in fact, was vicious total war, ruthlessly rational in economic and
political terms. It was cyclical economic war aimed at the capture of agri-
cultural resources and livestock, or the destruction of these so that enemies
could not benefit from them. This form of economic war was especially
virulent during periods of climatic instability, drought and hunger—as in
the 1860s, notably, as well as later in the nineteenth century. While pitched
battle between ‘professional’ armies appears to have been common enough,
there was scarcely any differentiation made between combatant and non-
combatant when it came to the targeting of particular communities by
those armies. Peasant farmers were deliberately attacked, periodically dis-
persing communities into hills and forests away from vulnerable positions
on open ground, or else pushing them together into fortified settlements.9
Regional chiefs pursued terroristic tactics against populations in enemy ter-
ritory in order to achieve hegemony and secure resources.Violence was not
something that happened on distant appointed fields, but was lived by ordi-
nary Tigrinya, Amhara, Oromo, and a host of others across the region.
Europeans were drawn to the depiction of the Oromo as savage hordes,
destroying a once-great civilization, as we saw in Chapter 2. Nonetheless, it
is clear enough that particular divisions of Oromo across the north and
north-east highlands were frequently the decisive military and political fac-
tor throughout the decades of the zemene mesafint. In Gojjam, Lasta, and
Wollo, the ethos and organization of Oromo forces, and their deployment
of substantial corps of horsemen, made them formidable enemies as far as
various Tigrayan and Amhara overlords were concerned.10 The latter, indeed,
frequently required Oromo support in order to make good, through armed
force, political claims over the longer term. Thus were the Oromo involved
in key struggles across the region, including Tigray itself. Plowden, well-
acquainted with mid-nineteenth century realities, noted that the Oromo of
Wollo and Shoa ‘are always in a state of more or less active warfare’, and
44 v i ole nc e and i mpe ri al i sm

recorded one Oromo as sneering, ‘ “What do you Amharas know of fight-

ing? It is the same to us, and as regular, as breakfast and supper are to you.
What was a horse made for but to fight on, and a man made for but to die
like a man when his time comes?” ’11 As Blanc suggested some years later,
‘[e]very Galla is a horseman, and every horseman a soldier; and thus is
formed a perfect militia, an always ready army, where no discipline is
required, no drill but to follow the chief ’.Yet Blanc was careful to point out
that there was no such thing as a single ‘Oromo nation’, and that they were
deeply divided according to lineage and region;12 this was to the advantage
of the habesha rulers who were frequently able to co-opt Oromo groups
into their own wars.
Between the 1800s and the 1840s, as competition heightened between
Tigray, the broad Amhara zone comprising Gojjam, Gondar, and
Begemeder, and Shoa with its particularly large Oromo population, suc-
cessful armies required ever larger numbers of both foot soldiers and
horsemen, as well as of firearms, and thus ever greater access to the
resources to fund these forces.13 The cumulative effect was a dramatic
militarization of political culture and discourse which has defined the
region subsequently. As Gobat wrote in 1832:
The supreme judicial and executive power, including both the administration
of the civil law, and military offices, are intimately connected in Abyssinia,
being deposited in the hands of a single individual; for all the governors are
both civil judges and military chieftains.14

For the broader population, the outcome was episodic suffering on a large
scale, with chronic insecurity of life and property across the central and
northern highlands, and as huge refugee movements away from zones of
conflict brought about severe economic hardship. Evidence from through-
out the period indicates that non-combatants, including (indeed especially)
women and children, caught up in the violence were likely to be either
killed or enslaved.These were not merely the clashes of noblemen and their
professional armies on appointed fields, rather this was terroristic social vio-
lence on a dramatic scale. Soldiery itself was indeed becoming ever more
‘professional’ in the course of the nineteenth century, certainly insofar as
there were ever larger numbers of men whose time was largely dedicated to
armed service; Plowden wrote extensively of a distinctive and powerful
military class.15 Armies themselves were supplemented by large numbers
of scavengers, fortune-hunters, and sundry hangers-on, all armed, if less
 state s of v i ole nc e, to C. 18 7 0 45

impressively than the warrior class, and predisposed to inflict damage on

communities where profit was to be had as a result. Clearly, too, cultures of
violence developed whereby the chase, the action, was an end in itself; bru-
tal and brutalized, the perpetrators of the violence, which spread out beneath
the grand political ambitions of the central figures of the period, were evi-
dently in thrall to the notion of gedday, killing, for its own sake. The milita-
rization of broader society was both cause and effect: all men carried arms,
wrote Plowden, which was necessary even among a numerous ‘middle class’
of landholders who needed to protect themselves from marauding soldiers
and nobility alike.16
In the mid-1840s, one traveller recorded grimly that
the whole country was always more or less in a chronic state of war. If the two
ruling princes [Ras Ali, west of the Takkaze, and Wube in Tigray] were not at
war with each other, they were either fighting their neighbours or quelling
rebellion among their own subjects.17

It seems possible to argue that it was through such violence that ideas about
both sovereign statehood and the political utility of armed force were devel-
oped in their modern Ethiopian context—much as, for example, the wars
of the princes in sixteenth-century Europe gave rise to military profession-
alism and laid the foundations of the nation-state. If it is the case that ‘[t]he
origins of Europe were hammered out on the anvil of war’,18 then it holds
true of much of Africa, too, and certainly north-east Africa. Plowden’s
description from 1844 of Ras Ali, Oromo by birth, and controlling a vast
realm of Amhara territory from Gondar to Harar, illustrates the essential
point that ‘the spear and buckler have outweighed the law’,19 echoing
Machiavelli’s assertion that ‘war is just when it is necessary’ and subsequent
debates in Europe about the legal framework required to facilitate war.20 As
Plowden explained, ‘[t]he chief power being entirely military, the soldier
occupies the principal place; the Ras, and all the great men of the country,
are of that class, and have absolute sway’.21 It was a precarious and unstable
existence, as recognized by Ras Welde Selassie in a revealing (if somewhat
embroidered) conversation with Henry Salt around 1811:
Even should I be successful [against Gondar] as I have every reason to expect,
still I have not the means of ensuring any permanent settlement of affairs. It
is my most heartfelt desire to see the King reinstated in all the dignity of his
office but how can I give him ability? And all rule in this country depends on
the energies of a chief himself—or how can I prevail [?] when I am obliged
46 v i ole nc e and i mpe ri al i sm

to return to my own Government (from which a long absence is absolutely

incompatible with my interests) affairs reverting as they have so often done to
the old system.When I march into the country they will, as before endeavour
to appease me by every humiliation they can employ and for a time agree to
all my measures, but no sooner is my back turned than they plunder the King,
refuse their proper tribute . . .

‘Besides’, he tells his European interlocutor, with (one imagines) a weary

sigh, ‘I am old now, and when I am dead all will come again to confusion’.22
While the central and northern highlands were embroiled in these con-
flicts, rapid change was also taking place in the south, with the emergence
of, and fierce competition between, several key Oromo states situated both
at the end of vital trade routes and in richly productive areas.23 State-
formation here came about through a series of dynamics. Alongside the
Omotic kingdoms of Yam (or Janjero), Kaffa, and Walayta, further north
there were the Oromo monarchies forming in the first half of the nine-
teenth century, including the so-called ‘Gibe’ monarchical states (founded in
the area of the Gibe river) of Limmu-Enarea, Jimma Abba Jifar, Gera,
Gomma, and Guma, there were also the Oromo principalities of Leqa,
Qellam, and Naqamte. The Gibe monarchies were the product of complex
processes of socio-political transformation: Oromo groups migrating into
the area absorbed both people and ideas, and gradually made the shift from
pastoralism to sedentary agriculture. This meant a long-term dilution of
egalitarian social organization based around gada, the age regiment system,
and the creation of social differentiation and political hierarchies similar to
the extant political systems of the Ethiopian Highlands. Wars of expansion
led to the enhancement of the position of war leader (abba dula in the gada
system) which in several cases evolved into a monarchical institution.
Additionally these were polities which fed off the increasingly lucrative
trade routes running between the lower Nile basin and the southern Red
Sea. Certainly, by the early nineteenth century, key states which had emerged
as a result of these dynamics included Jimma Abba Jifar—named after its
founder king, Abba Jifar I, who reigned 1830–5524—and Enarea, which
reached its height under Abba Bagibo (reigned 1825–61) before being
eclipsed by Jimma.25 Another dramatic example is the state created by Jote
Tullu in Wallaga in the later nineteenth century, which used military power
to exploit the Ethiopian–Sudanese frontier commercial network. On the
eastern side of the Great Rift Valley, Harar, the successor to Adal and a city-
state which was becoming a major centre of Islamic influence, successfully
 state s of v i ole nc e, to C. 18 7 0 47

resisted repeated attempts by local Oromo to overrun it, and indeed man-
aged to convert some Oromo in the vicinity to Islam.
Thus did great territorial borderlands open up across a huge area of mod-
ern northern and central Ethiopia, characterized by a series of ethnic as well
as multi-ethnic polities in competition with one another over space, econ-
omy, ideology, and belief systems. In a process of violent symbiosis, these vari-
ous borderlands and frontier zones both cardinally defined the communities
in and around them, and in turn were shaped by adjacent polities. It was a
vortex of competition characterized by frequent minor armed clashes, and a
handful of major battles. Campaigns which often stretched over months and
years had a devastating impact on local populations, which were not only the
target of the violence but which had to endure the garrisoning of troops on
their lands for long periods at a stretch. Although it may be a characteristic
of the region’s deeper past, too, there can be little doubt that the nineteenth
century witnessed a growing gulf between military and peasantry, and a
growing belief among the latter that the former were rarely to be trusted,
and only to be welcomed cautiously and often under duress. Above all, this
was an era in which violence had primacy in political culture and in which
militarism was both the means to an end and an end in itself in public affairs,
thus creating models for political interaction according to which the region
continues to function.Violence and militarism in nineteenth-century north-
east Africa became self-perpetuating and cumulative.
Before the zemene mesafint fell into historiographical obscurity some
twenty years or so ago, there was debate over the nature of the conflicts
themselves—notably, whether centred on class, region, or ethnicity. While
‘classic’ accounts such as that by Abir couched their analyses in terms of
‘Galla’ (Oromo) domination and Tigrayan and Amhara struggles against it,26
Crummey warned against the ‘inadequacies of ethnicity as an explanation
of politics in Christian Ethiopia during the Zamana Masfent’, and instead
placed emphasis on ‘the social framework of political power’.27 Class, he
asserted, was crucial—particularly in terms of struggles between the nobil-
ity and the monarchy, and within the nobility itself. Much of this discussion
needs to be understood in its intellectual context, namely the then-current
conversation concerning Ethiopian ‘feudalism’ and its similarities to medi-
eval European socio-political structures.28 There can be little doubt that
social status (and the mobility which characterized it) was a key factor in the
pursuit of conflict; violence was often driven by ‘class’, loosely defined, in
terms of access of economic and political resources. Regionalism drove
48 v i ole nc e and i mpe ri al i sm

conflict, too, for the same reasons—nor were regions always defined in
‘ethnic’ terms: witness the ongoing intra-Amhara violence, notably the
often bitter competition between Shoa, Gojjam, and Gondar, and that
between Tigrinya communities on either side of the Mereb River. But
ethnic, cultural, and linguistic consciousness was also critical, and increas-
ingly so; the analysis presented here rests on the argument that the militari-
zation of politics involved the utilization by a range of groups and
communities of various forms of identity (class, ethnie, territory), and that
violence both was driven by and helped enhance, these cohesions. It is
certainly the case that ideas about ethnie, nation, and cultural community
were at the forefront of violent conflict, and both grew up within and were
defined against the frontiers which define our story.
This period witnessed an expansion of Red Sea commerce, and in par-
ticular the expansion of the north-east African slave trade with the height-
ened demand for slaves in the Arabian peninsula and the Middle East.29 The
marked increase in slaving violence across the north-east African highlands
and into their lowland environs in the course of the nineteenth century had
a profound influence on state-formation in the region. In the middle of the
eighteenth century, Jeddah was already a major destination for slaves from
the region, both habesha and Sudanese, although the former, especially
females, reportedly fetched higher prices on account of their supposed
physical beauty. Habesha slavers themselves preyed on the Sennar border-
lands, and on Oromo and shangalla communities further south;30 the Gash-
Barka lowlands constituted another important hunting ground in the
mid-nineteenth century.31 But this was no one-way traffic: the north-west
borderlands of Ethiopia were vulnerable to ‘Arab’ slave-raiding incursions,
while along the western frontier Oromo and other slave raiders were regu-
lar visitors.32 The so-called ‘Baria’, observed Parkyns in the mid-1840s, regu-
larly raided for slaves into the northern highlands.33 Slaving warfare was one
of the key forms of the economic war of the nineteenth century; it was also
violence which was racially motivated and justified. The armies that preyed
on dispersed mixed-farming communities along contested borderlands rep-
resented an ever more complex and large-scale military organization, which
in itself represented expanding and ever more sophisticated state-formation.
Such statehood increasingly defined the parameters of both warfare and
economy across the region.
Slaves, cotton cloth, and horses were traded at markets along the Sennar–
Ethiopian borderlands; Metemma was an important commercial centre.34
 state s of v i ole nc e, to C. 18 7 0 49

Slave merchants made their way via burgeoning trade routes to the Red Sea
hinterland, often across considerable distances, for example the ‘Sidama’
groups, reported Gobat, were heavily involved.35 In the 1770s, Bruce noted
that the settlement of Digsa, in Akele Guzay, was a key market town for
slaves, especially children, from whence many were presumably dispatched
coastward.36 Adwa in Tigray was likewise a key transit point for slaves: Salt
reckoned about a thousand passed through annually, some of whom were
sent onward to Massawa, and others to smaller ports further north, as traders
sought to avoid the customs duties levied by the na’ib.37 Shortly after acced-
ing to power, Tewodros had famously abolished the slave trade as part of his
putative ‘modernization’ programme, but by the end of the 1850s the com-
merce was thriving38—indeed it is unlikely that Tewodros’ proclamation had
made any real difference. Slaving continued to be a central component
within the larger umbrella activity of warfare, and remained so in the 1890s,
when Menelik’s soldiers took slaves in the north-west during skirmishes
with the Mahdists;39 in the 1870s and 1880s, meanwhile, swathes of the
kebessa, especially Hamasien, had been devastated by the slaving which
became a lucrative sideline during the Ethio-Egyptian conflict. Many of
these, again, made their way to Arabia, Wylde recalling that ‘when I first
went to the Red Sea as British Vice-Consul with headquarters at Jeddah,
the Hedjaz was full of Abyssinian females . . . that had been taken from the
Hamasen’.40 Against the violent background of the zemene mesafint, then,
regional commerce was thriving—and while it is clear that recurrent con-
flict was frequently inimical to the successful flow of trade and economic
development more generally, there can be little doubt, by the same token,
that the promise of economic gain (whether the control of trade routes or
the capture of resource-rich territory) drove a great deal of the violence of
the nineteenth century.

The politics of shiftanet: Tewodros

and the frontier state
The period between the 1850s and the 1890s was a formative one in the
modern history of the region, although there was considerable continuity
from the early nineteenth century. It was an era which witnessed the culmi-
nation of a process of militarization in terms of Ethiopian political culture
50 v i ole nc e and i mpe ri al i sm

originating in the mid- and late eighteenth century.The story of Tewodros41

in many ways encapsulates the story of Ethiopia over the past two centuries
or more: the advance of the ‘periphery’ on the ‘centre’, with the latter thus
fundamentally defined by the former; the level of institutionalized and
cyclical social and political violence; and, closely connected to the first
point, the prevalent but much misunderstood role played by shifta in the
evolution of the region’s political culture. Born Kassa Haylu in around 1818,
Tewodros was from an increasingly unstable and contested borderland
between present-day Ethiopia and Sudan—namely, the lower-lying Kwara
area to the west of Lake Tana and Gondar. A religious education made him
deeply pious, but also critical of the established church, as he understood it.
His early life was played out against the backcloth of the zemene mesafint:
apparently in some way related, on his father’s side, to several key regional
overlords of the early and mid-nineteenth century, enabling him to claim a
degree of status—the habesha nobility was an intricate web of inter-familial
relationships—he served in the household of one of them, that of his half-
brother in Kwara. Normally, such service should have led to his being
rewarded with land—a portion of a gult estate—on which he could have
sustained himself as an aspirant noble. On the death of his patron, however,
Kassa received no such reward, and following a period in the service of
Goshu of Gojjam, he ‘rebelled’ and became shifta, established practice for
frustrated or disenchanted nobility.
The shifta phenomenon deserves some brief consideration at this point,
crucial as it is to an understanding of the political culture of the region. We
will see later in the book how important shifta are at particular times and in
particular places—from the Ethiopian region of the nineteenth century, to
the politically turbulent Eritrean borderlands in the 1940s and 1950s, to the
supposedly irredentist Somali movement in northern Kenya in the 1960s
(known as ‘the Shifta war’). It will be clear that the term shifta itself has
changed over time, according to context; as Crummey explains, the word is
derived from the Amharic shiftanet, meaning ‘banditry’, which itself stems
from shaffata, ‘to rebel’.42 It was an ambiguous term, referring both to com-
mon banditry as well as (often in romanticized hindsight) to heroic aristo-
cratic rebellion; indeed it seems that over time the word became more
associated with the former and less with the latter. Across north-east Africa
in the course of the twentieth century, it found its way into English usage
and came to mean ‘any armed band at odds with the state’ as well as low-level
criminal behaviour, and was used derogatively to refer, for example, to the
 state s of v i ole nc e, to C. 18 7 0 51

guerrillas of the Eritrean liberation movement in the 1980s.43 In the nine-

teenth century, shifta were both bandits and rebels, depending on the beholder;
they took to the bush or mountains and defied authority for a range of rea-
sons, whether quick pecuniary gain through banditry or political disenchant-
ment and ideological revolt.44 Their presence in frontier zones and borderlands
enhanced the perception of those places as dark and dangerous, possessed of
a sinister fertility from which grew challenges to the order of things.
Shiftanet, rebellious activity, flourished throughout the nineteenth century,
and indeed was both a cause and a symptom of the zemene mesafint. Ras
Mikael—later the self-appointed guardian of the Solomonic order—began his
career as a shifta in the 1750s; in the following century, both Tewodros and
Yohannes came to power through shiftanet. Thus, being shifta in some senses
appears to have been something of a political and military apprenticeship, in
which aspirant leaders acquired skills at arms and developed strategies of com-
mand and organization; arguably, the same has proven to be the case of the lib-
eration fronts in the later twentieth century. Shifta were fundamental to the
evolution of habesha political culture.They were both cause and symptom, again,
of political instability; they were often the direct manifestation of the ‘fron-
tierism’ which characterized and shaped political discourse across the region,
and they represented, in different ways, the centrifugal forces with which
would-be centralizers continually had to grapple.
The young Tewodros was effectively a product of the north-west fron-
tier.45 Early skirmishes with the Egyptian forces as they consolidated their
control over northern and central Sudan in the 1840s had had a profound
impact on his understanding of regional dynamics, and indeed of the world
beyond.A major defeat at the hands of the better trained and better equipped
Egyptians in 1848 taught Tewodros some valuable lessons in the military
arts; his motley bands of variously motivated shifta were no match. This was
a defeat which doubtless convinced Tewodros of the absolute primacy of
armed force, a lesson which would be a crucial one, too, for future genera-
tions of those who, like Tewodros, sought the capture of the state from the
neglected borderlands and oppressed peripheries. Certainly, Tewodros was
convinced—and would remain convinced, until taking his own life in the
face of an inexorable British advance twenty years later—that armed might
was right, and that there was no political problem which could not be
approached through the use of mass, concentrated force. Following the
defeat by the Egyptians, there were a series of notable victories over key
regional overlords closer to home, in the course of the early 1850s.46 His
52 v i ole nc e and i mpe ri al i sm

growing success won him the respect of the Warrashek, who acknowledged
him in his position as governor of Kwara province—a hollow gesture, as he
was already effectively in control of it; he went on to consolidate his posi-
tion, accumulating more governorships, and even marrying the daughter of
Ali, the Yajju prince. His success was founded on a brilliant combination of
tactical and strategic intelligence, passionate leadership, and the occasional
stroke of good fortune, as well as errors by his enemies. By the early 1850s
he was in a position to challenge the key leaders of the northern and central
highlands, defeating first Goshu, and then, in mid-1853, his former patron
Ali at the battle of Ayshal; in some respects, this ended Yajju pre-eminence,
and thus the era of princes itself. However, Tewodros had yet one major
fight to win—which he did in early 1855 when his army crushed that of
Wube, governor of Tigray and Semien, and erstwhile competitor for the
Solomonic inheritance.47 His subjugation of Wollo and Shoa, in the eastern
highlands, lent further apparent legitimacy to his coronation as Tewodros II,
King of Kings. His belief in the need to establish a monopoly on the use of
force, and to deploy that force in the direction of political challenges, seemed
vindicated. What followed was an aggressive renaissance of the Solomonic
ideal, around which would be built the notion of a perennial, unified
Christian kingdom which would be regionally dominant.
Sheer armed force would also serve to resolve the issue of Tewodros’
dubious ancestry, for the problem was that his lineage was weak. Many
regarded him as a usurper, and for much of his career as ‘king of kings’ a
great many contemporaries refused to accept that he was of Solomonic
descent.48 Yet his claim of descent from Fasilidas was accepted—again, his
extraordinarily effective violence ensured it—and he was crowned by the
(Egyptian) head of the Orthodox Church. His very choice of throne name
suggested an amalgamation of hubris and insecurity, alluding to a long-
standing Orthodox belief that a monarch named Tewodros would come to
save the kingdom, and rule for a thousand years. It was a popular fiction, but
the historicization of violent conflict was the defining feature of nineteenth-
century habesha statehood, and was underpinned by ideas about destiny and
inheritance.49 These ideas—resting at the heart of the Ethiopian world-
view—awarded Ethiopia a belief in a grand narrative arc, which in turn was
translated into an aggressive potency as an empire-state. It would be among
the most dangerous legacies to the modern era.
In truth, Tewodros’ command of his realm was never more than slender,
for all its righteous violence; rebellion was frequent, and his response to it
 state s of v i ole nc e, to C. 18 7 0 53

ever more bloody. Again, this was conflict not simply between ever more
professionalized soldiery, but escalating violence between militarized and
armed peasant communities, so that the lines between combatant and non-
combatant became ever more blurred. While Tewodros’ soldiers were sent
against the troops of defiant chiefs, they were also sent against ‘civilian’ set-
tlements who quickly learned to defend themselves against soldiers from
whatever direction. Within months of his coronation, there were revolts in
Gojjam, in Tigray, in Begemeder. ‘My people’, he reportedly told the mis-
sionary Henry Stern, ‘are bad; they love rebellion and hate peace; delight in
idleness, and are averse to industry; but, if God continues to me my life . . . I
will eradicate all that is bad, and introduce all that is salutary and good’.50
His transferral of illegitimacy onto those he aspired to govern is notewor-
thy; at any rate, his professed policy of purification would be a gory one
indeed. Stern’s assessment is colourful:
The King’s relentless severity towards rebels and traitors does not, however, in
the least damp the aspiration for power, or the passion for dominion. Men and
women are continually scourged and mutilated; whole regions of wild hordes
are sent to desolate and lay waste suspected and disaffected districts; whole
clans are proscribed and outlawed; and yet all these extreme measures and
sanguinary edicts fail to enforce obedience, or to win the nation’s fealty.51

Stern, writing in the early 1860s, opined that, while Tewodros was unable to
prevent continual rebellion, he had had some success against ‘theft and mur-
der’, and ordinary criminality.52 A near contemporary, Dufton, advised his
readers: ‘There are those who think they have described the man when they
have stigmatised him as an inhuman despot, a bloodthirsty tyrant, a Nero, a
worse than King of Dahomey’; but, no, exclaimed Dufton,Tewodros ‘is not all
devil!’53 Later assessments were less sympathetic. The weight of a huge and
more or less permanent army was enormous on land and people, and the
military revolution proved self-consuming;54 indeed, in a curious twist,‘peace’
was actually anathema to the shifta state, for, once created, such an army could
not be unmade, and violence became an end in itself. ‘From 1860 to 1868’,
wrote Blanc, ‘he seems little by little to have thrown off all restraint, until he
became remarkable for reckless and wanton cruelty’,55 as he confronted such
sworn enemies as Tigrayan rebels, the governor of Gojjam, the defiant Wollo
Oromo, and the Shoans—and, of course, as we see below, Muslims.
The emperor, his violence sanctioned by God and facilitated by destiny,
had no option (he told anyone who would listen) but to confront his
54 v i ole nc e and i mpe ri al i sm

wayward subjects with force; he ‘complained bitterly of the conduct of the

Abyssinians in general, whom he styled “a wicked people” ’.56 He could not
control the ramifications of his own violent revolution. Uprisings were
habitual from the early and mid-1860s.57 Moreover, an army had been cre-
ated which was in need of continual action, yet whose tactics rendered ‘suc-
cess’ by any standard impossible, as well as causing the soldiers to be hardly
better off than the peasantry itself. Ultimately, this was a ‘war of extermina-
tion’ between soldier and peasant, with atrocities committed on both sides;
killing was commonly attended by physical mutilation, and the deliberate
targeting of troublesome communities and indeed entire regions was the
key feature of this warfare.The economic effects, as during the early decades
of the nineteenth century, were dire, as harvests were destroyed through
scorched earth tactics, propelling prices for basic foodstuffs skyward.58
Starving and unruly soldiers increasingly preyed on destitute communities
for whatever meagre sustenance they could find,59 with supposedly ‘disloyal’
districts targeted in particular.60 It was a scenario which would be ever more
familiar in the twentieth century.
Military reorganization, above all, was Tewodros’ key aim. In part this was
the direct result of the experience of the increasingly violent political envi-
ronment of the highlands: might was right, and armed force had become
paramount.There is some evidence, too, that the militarization of politics in
the late eighteenth and early nineteenth centuries had produced swollen
militaries which needed to be controlled, and Tewodros sought to do this by
creating a single, unified, ‘national army’. He also attempted to regularize
the payment of a salary to soldiers, and to prevent the looting and requisi-
tioning that was standard military practice. He reformed organizational and
command structures, and sought to equip his entire army with guns.61 There
was also the issue of an increasing hostile external environment, notably in
the form of Egypt in the Sudan and in the Red Sea. Tewodros knew what
guns could do, and sought the import of modern arms—commercial activ-
ity which brought him ever closer into the orbit of external powers, most
obviously Egypt and Europe. Tewodros’ perception of Muslims circling
Ethiopia, slowly strangulating it, meant that overtures to Europe served a
dual purpose, namely the acquisition of firearms and the creation of reli-
gious alliances against common enemies.
Yet his military reforms, and the necessarily political changes they
entailed—the garrisoning and recruitment of troops, the placement of loyal
 state s of v i ole nc e, to C. 18 7 0 55

political establishments in the regions—were met with increasing levels of

resistance, which were in turn fuelled by his land confiscations. In many
provinces he was simply not recognized, and was rejected as an interloping
and over-mighty shifta. His response to violence resistance was even greater
violence, and by the early 1860s he had largely undermined his own best
efforts to both regulate armed force and establish his monopoly on the use
of force. Provincial rebellions were met with severe counter measures—
scorched earth tactics, notably, and the brutal treatment of populations
which he suspected of disloyalty—and much of Tewodros’ reign was, in
truth, spent on campaign. In the final analysis he knew only violent answers
to political problems, and the cyclical conflict which spiralled beyond his or
anyone else’s control in the course of the 1860s seemed to bear out the
lessons of the zemene mesafint—that violence begat violence, and that politi-
cal failure resulted from an inability to control the armies which had become
an intrinsic part of Ethiopian political culture. When in his frustration at
perceived enemies everywhere he attacked and partly destroyed Gondar, in
1866, there was outrage and condemnation. Many called into question, once
again, his background and lineage.
If his early career had been defined by a strange fanaticism, his later life
was characterized by a violent paranoia, even if it has been in the interests
of subsequent story-tellers to exaggerate the tales of madness and bloody
extremism. But there is enough in the contemporary source material to sug-
gest that he soon came to be seen as a violent transgressor, noted for a level
of overt and violent brutality that was generally deemed unacceptable—
destroying churches, burning people alive or starving them to death, mas-
sacring prisoners of war, destroying wholesale the herds of cattle which
were the livelihoods of communities.62 In the words of a local chronicler,
‘everyone held the King in dislike for he had resort to punishments hitherto
unknown in the land’.63 In European texts, he is the embodiment of the
crazed oriental despot, the stuff of Victorian nightmares; but we must be
careful not to lift his violence out of context. His was the violence of a
political culture born of the zemene mesafint, the violence of that era writ
large. An exemplary product of the fertile frontier, he did not imagine there
were enemies everywhere; there were enemies everywhere, but his response—
entailing extreme armed force—merely served to exacerbate extant fissures
and conflicts, as would be the case a century later under the Derg. He may
well have been a paranoid individual, and there is little doubt that such
56 v i ole nc e and i mpe ri al i sm

paranoia prompted the assault on Gondar—and other acts considered ‘ille-

gitimate’ in the eyes of contemporary beholders—but it was also indicative
of the fact that Tewodros’ state was restless and uncertain of itself, uncom-
fortable with the political inheritance it claimed for itself. There was no
clearer sign of that instability—at least in the European mind—than the
imprisonment of a number of Europeans, including missionaries, following
the apparent lack of interest in Tewodros’ struggles on the part of the British
Government. The chief resistance to him came from Gojjam, Wollo, and
across the north, including Semien and parts of Tigray.64 Repeated peasant
uprisings overthrew his authority and that of his soldiers, and no amount of
vengeful violence—of which there was a great deal—could restore his
power. By 1867, he was largely confined to a rump state in the eastern high-
lands with perhaps 15,000 men, around the great rocky outcrop of Magdala,
surrounded by enemies.65 In one of the most expensive overseas military
operations mounted by the Victorians, the British Government appointed
General Napier to lead a force of 42,000 men, from the Gulf of Zula on the
Eritrean coast up into the highlands, to confront Tewodros and free the cap-
tives. It was a remarkable expedition, although its progress was greatly eased
by the fact that the north was in arms against Tewodros, and there were
several local leaders there—including Ras Kassa of Tigray, the future
Yohannes—who were willing to assist the British.66 Napier’s job would
have been rather more difficult, perhaps impossible, had this not been the
case—a lesson forgotten by the Italians a generation later. The British
reached Magdala, where Tewodros, besieged and outgunned, committed
suicide rather than fall into the enemy’s hands.67
A standard view is that Tewodros was both reformer and upholder of
tradition, harbinger of change as well as embodiment of continuity. There
is evidence that he attempted to introduce certain changes in military
organization, and improvements in technology capacity (again usually of a
military nature) and political administration. But too many historians tend
to mistake this for something called ‘modernization’, which denotes some
kind of inexorable forward movement, largely inspired by expanding
Europe and the models of organization it supposedly offered. Tewodros,
rather, needs to be seen as the beginning of an era of high violence, his
reign marking the beginning of an era—not yet at an end—defined by the
hegemony of armed force in political affairs.Tewodros may indeed at times
have dreamt of the equipment enjoyed by European armies; he was also a
remarkable man, possessed of considerable talent, and vision and energy—
 state s of v i ole nc e, to C. 18 7 0 57

the past is, surely, marked by them. But he was a product of the peculiar
political, cultural, and intellectual environment that was the northern
Ethiopian Highlands in the late eighteenth and early nineteenth centuries;
he was at least as concerned with looking backwards, into some half-
imagined, dimly perceived past, as he was about the future. Yet in many
respects the debate about the opportunities and capacities for ‘moderniza-
tion’ dominated foreign discourse about Ethiopia for the next century—
and indeed came to concern the thoughts of many Ethiopians, once some
of the political and intellectual elite had glimpsed this other world for
themselves. As for Tewodros himself, he was concerned first and foremost
with acquiring the equipment with which to combat Ethiopia’s sworn
enemy—Islam, whose representatives had stolen piece by piece the ancient
kingdom’s rightful inheritance.
No doubt external models provided some impetus for so-called ‘mod-
ernization’; modern Ethiopia has indeed been seen as an exemplar of the
defensive modernization which characterized a number of states in north-
ern Africa, along with the Ottoman Empire and Japan (in the course of
the nineteenth century), dominated as this period was by European impe-
rial expansion. No doubt, too, the combination of the attractions of neo-
Solomonism along with the perceived threats on external frontiers
prompted several provinces toward a greater degree of unity and cohe-
sion—the overcoming of internal frontiers, in other words, in order to pay
closer attention to the external. Yet there is something rather too neat in
this model of political development. In the quest for the reassuringly solid
notion of modernization, this paradigm has tended to downplay the vio-
lence of the era, and the role played by armed force—and its dramatically
enlarged scale in the course of the nineteenth century—in bringing about
a coerced unity. The rough edges are largely removed; and in any case, the
Ethiopia that was born of these developments remained an intrinsically
unstable political, territorial, and indeed ideological entity.The nineteenth-
century Ethiopian region witnessed something of a violent revolution, and,
again, a dramatic expansion in the scale of warfare, its organization and
objectives. There can be little doubt that, at least in part, this was itself
driven by external stimuli in the form of new territorial and religious
threats. It was a process which began during the zemene mesafint; and one
of the chief legacies of the zemene mesafint was not simply a widely held
belief in the primacy of armed force and the utility of political violence,
but in the means to put this into effect—namely, a greatly enlarged soldiery.
58 v i ole nc e and i mpe ri al i sm

The conflicts of the zemene mesafint had produced cultures of violence and
had swelled communities of armed men, better organized and more effec-
tively channelled than previously, creating a more professional military
ethos ably supported by the militias which could be conjured up according
to need. It was these which ensured the success of the programme of polit-
ical and territorial enlargement initiated by Tewodros. The enlargement of
armed force enabled a succession of rulers—temporarily, at least—to over-
come the ‘internal’ borderlands (political, ethnic, and physical) which had
long fragmented the political environment of the Ethiopian Highlands.
But a swollen militarism also led to a dangerously restless violence across
the region which would periodically destabilize the very polity that it had
brought into being.
Tewodros has iconic status in the region’s history. He was, and has been,
many things to many people: shifta, restorer of order and unity, nation-
builder and modernizer, bloody tyrant, tragic hero. As with the era to which
he supposedly put an end, the zemene mesafint, we know a great deal about
Tewodros, and yet we also know very little—certainly in terms of what he
represented. He embodied the essential violence at the heart of political
discourse in the Ethiopian Highlands, even if he was the extreme manifesta-
tion of it; he also foreshadowed the supposedly ‘modern’ phenomenon of
the margins seizing the centre—whether shifta or guerrilla—and becoming
the ‘established’ order, although in fact this is a recurrent theme in the his-
tory of African political development more broadly, and certainly of African
violence. Above all, he should not be seen as some essential ‘stage’ in a model
of linear political development, a link between the bloodletting savagery of
the zemene mesafint and the emergence of the modern nation-state: he was
the past, and he was the future. Certainly, he contributed to the reinvention
of Solomonism, or if we prefer the invention of neo-Solomonism, which
encapsulated something of a ‘new’ vision of Ethiopia. But his actual history
reflects the realities of violent statehood, insurgency, and counter-insurgency:
he was unable (or unwilling) to move beyond these realities, and forge ‘new’
ones. Critically, for our purposes, his was shifta statehood, often brilliant in
the context of the politics of armed force, but restless and unstable; it was a
form of government which would re-emerge in the course of the later
twentieth century across the region, manifest in the modern leadership of
the TPLF in Ethiopia and the EPLF in Eritrea. In more than one sense, shifta
statehood embodied the violent frontier which has defined political culture
in the region.
 state s of v i ole nc e, to C. 18 7 0 59

Frontiers of faith
The frontier between Christianity and Islam across north-east Africa was, over
the longer term, a shifting one; and while it would be misleading to depict the
religious history of the region as characterized by conflict alone, it is clear that
the positioning of that frontier, and the relative importance of the events and
interactions along it, periodically had an enormous impact on the political
and cultural evolution of states and societies along a broad arc linking Somalia,
Ethiopia, and Eritrea.Violence erupted along and around the frontier episodi-
cally, and sometimes with long-term consequences. Islam and Christianity
have co-existed uneasily in the southern Red Sea and lower Nile valley region
for nearly a millennium and a half, and the sanguine perspective might be that
it is remarkable the extent to which religious conflict has generally been man-
aged and controlled. Although at least one scholar of Islam has seen the
Eritrean conflicts of the later twentieth century in primarily religious terms,68
and there is no doubt that the faith-driven component has often been played
down here, the fact remains that Christians and Muslims have often fought on
the same side, or against one another, for reasons other than matters of the
spirit. Arguably, for example, recurrent violence between the Amhara and the
Somali has often been economic and political rather than religious per se,
notwithstanding the fact that, as Ernest Gellner put it, each group possessed a
rival edition of The Book.69 Faith, in other words, often lubricated the
machines of war, but it was not necessarily the Faithful who built them.
African Islam, at least south of the Sahara, has been strongly influenced
by Sufism.This has made it much more eclectic, flexible, and less vulnerable,
if not wholly immune, to external stridency than might otherwise have
been the case. Although there has been and continues to be disagreement
about the precise nature of Sufi influence in Africa, the emphasis placed by
Sufism historically on personal piety and exemplary behaviour, in the words
of Knut Vikor, has been rather more important than ‘its external functions
as a focus for political combat and jihad’.70 In other words, African Muslims
have been historically less responsive to the call to arms than others of the
Faith. Second, and more directly pertinent to north-east Africa, it has been
suggested that Somali ‘xenophobia’ has likewise rendered Islam in that area
comparatively immune to external influence.This goes some way to explain-
ing what Iqbal Jhazbhay terms ‘the relative inter-faith détente that has
existed between Christian and Islamic spheres of influence in the Horn of
60 v i ole nc e and i mpe ri al i sm

Africa’. Somali Islam ‘appears to be solidly located within a tradition of

regional, geo-cultural, peaceful co-existence between Christianity, Islam
and indigenous animistic tendencies’.71
That said, there was a great degree of religious conflict. The collapse of
the Gondarine state coincided with a rejuvenation of Islam in north-east
Africa and the Red Sea zone more generally, and there was a significant
religious element to the zemene mesafint.The nineteenth century witnessed
a marked increase in intra-faith violence, both because of more aggressive
Islamic presences across the region, in various guises, and because the
‘restored’ Ethiopian state of Tewodros and Yohannes was deeply rooted in
a sense of Christian mission. In the early nineteenth century, Muslim influ-
ence in the Ethiopian region was rejuvenated because of the expansion in
Red Sea trade, and the strengthening links this brought with the Arabian
peninsula and the Middle East. North-east Africa became much more
closely connected to the Islamic world in the course of the nineteenth
century; there was also something of a revival of the hajj from Ethiopia to
Mecca. Importantly, these new connections were developing at a time
when the Muslim world itself was witnessing the emergence of a series of
revivalist movements which sought the restoration of ‘pure’ or ‘fundamen-
tal’ Islam, and a more rigorous enforcement of the shari’a or Islamic legal
code. Often, these revivalist movements were messianic, predicated on the
imminent arrival of the ‘Mahdi’, or saviour, who would oversee the resto-
ration of a purified Islam on earth; revivalist movements asserted that inno-
vation within, and deviation from, pure shari’a was wrong, and they often
advocated jihad, violent confrontation sanctified by God, in order to bring
about purification and orthodoxy. The result was that often their enemies
were not simply Christians or other ‘infidels’, but Muslims who were held
to have lapsed and wandered from the righteous path. While some of the
most important such movements emerged far beyond the African land-
mass—Shah Wali Allah’s challenge to fading Mughal power in India in the
eighteenth century, for example—it was the appearance of the alliance
between zealous reformer Muhammad ibn ‘Abd al-Wahhab and local
chieftain Muhammad ibn Saud in the Arabian peninsula in the eighteenth
century which had the most direct impact on the Red Sea world and
north-east Africa.
The rapid development of commercial enterprise across north-east
Africa—in the form of long-distance caravans which penetrated across the
highlands and into the west and south—was largely under the direct control
 state s of v i ole nc e, to C. 18 7 0 61

of Muslim merchants.72 The result was the further Islamicization of the

Oromo in those areas being brought into the Red Sea and Gulf of Aden
commercial orbit. Meanwhile, the continued spread of those Oromo who
were already Muslim in the late eighteenth and early nineteenth centuries
provided Muslim merchants with a base within which and from which to
establish ever greater influence. The founder of the Enarea kingdom, a
merchant-adventurer named Bofo Abba Gomol, for example, quickly
adopted Islam and employed Muslim scribes, traders, accountants, and
counsellors in the process of consolidating his power.The capital of Enarea,
Sakka, boasted a community of busy Muslim merchants, but also hundreds
of ulama, religious scholars.73 Sufi brotherhoods, again, were instrumental
in the spread and establishment of Islam across the region. In the east and
south-east, Islamic political culture was increasingly centred on the strate-
gically positioned city-state of Harar, and although the Somali to the south
were politically fragmented, the Afar had developed a sultanate in the cen-
tral Danakil area. The Qadiriya, a prominent Sufi order, had been active in
the Harar area for some time prior to c.1800, but in the course of the nine-
teenth century the brotherhood spread further among the Somali and also
into parts of present-day Eritrea. Reformist orders such as the Salihiya
were active in the Ogaden from the 1850s, attracting the displaced and
those seeking refuge from conflict to their farming communities. In the
north, the Eritrean region saw the expansion of the Mirghaniya from the
1860s. Often, Muslim settlers married into chiefly families—much as they
had on the east African coast—and gradually brought those lineages and in
time entire communities over to Islam.
In Wollo in the eighteenth century, converts were brought to the faith
through the activities of Shaikh Muhammad Shafi, the major Sufi leader of
the region.74 A student of theology, law, and the Arabic language early in life,
he was initiated into the Qadiriya brotherhood and became one of its key
figures, a successful recruiter with an increasing personal following; as he
grew in stature, he acquired a reputation for learning, saintliness, and the
performance of miracles. Notably, he had uneasy but peaceable relations
with local Wollo chiefs, many of whom were themselves Muslim; he avoided
their courts, and in fact in his later career preached increasingly of the need
for jihad in order to both purify existing Islamic practice and to extend
the faith itself. Indeed, he had even attempted to solicit support for jihad
in Wollo while on pilgrimage to Mecca, but the Meccan authorities
were largely unmoved. But he did succeed in establishing a permanent
62 v i ole nc e and i mpe ri al i sm

settlement—Jama Negus, the community of the king—inhabited by a large,

devout, and militant following. From among their number—in much the
way that medieval Christianity had spread in the northern mountains—
came clerics in a similar mould to Muhammad Shafi, consolidating and
spreading Islam across the region.
And yet this was an eclectic, flexible Islam: new Muslim converts contin-
ued to venerate the shrines of local saints, and the pilgrimage to Mecca and
the worship of Muslim saints were frequently combined with pre-Islamic
practices and cults, notably among the Oromo, such as sacrifice and posses-
sion. Indeed it was because of such practices that reformers such as
Muhammad Shafi in Wollo had urged the need for purification. Precise
numbers are impossible to come by, but clearly conversion to Islam in the
course of the nineteenth century was indeed considerable. And Islam went
beyond communities of clerics and merchants: Ethiopian Islam could be
found within the political establishment, and among farmers and herders.
Mixed communities were common, and it is indeed the case that Christians
and Muslims could and did live entirely harmoniously. Many ruling families
themselves were mixed, with coexistent Christian and Muslim branches.
However, while such harmony might exist at the local level, larger ten-
sions—more entrenched frontiers of faith—were developing as the nine-
teenth century went on. It was against this spread of Islam—among the
southern Oromo, in the eastern Somali region, and to the north in Tigray
and parts of modern Eritrea—that the Christian neo-Solomonic monarchs
began to carve their own ‘new’ states in the central and northern highlands
from the mid-nineteenth century. Tensions within the Ethiopian region, as
well as conflict with external Muslim enemies, came about not least because
at the ideological core of the highland state was the linking of Christianity
and habesha identity. In other words, Muslims could not be ‘true’ Ethiopians.
They were the enemy within, as well as the enemy beyond, for Solomonic
rulers who were deeply concerned with the threat from both.Tewodros and
Yohannes frequently spoke of Ethiopia’s deeper history as well as their con-
temporary concerns using the language of religious struggle. Muslims were
traders, and therefore involved in grubby work, unlike the Christian warrior
nobility; one contemporary source in the mid-nineteenth century sug-
gested that ‘[f]ew [Muslims] are soldiers, they being esteemed by the
Christians as cowardly and effeminate’.75
Tewodros’ spiritual mission was effectively his foreign policy, his view of
Islam shaping his view of the region, and the world. He was deeply spiritual
 state s of v i ole nc e, to C. 18 7 0 63

as a youth, receiving a religious education,76 and apparently from a young

age he saw it as his mission to ‘restore’ an imagined glorious past, doubtless
in large part inspired by the Kebre Negast itself, and to wrest back from
Muslim antagonists, who had gradually encroached on the former Ethiopian
empire, all those territories once ruled by the King of Kings. His ambition
was remarkable, and was indeed duly remarked upon by most Europeans
who encountered him: Tewodros wanted the restoration of Jerusalem to
Ethiopian control, and the destruction of Mecca; his territorial vision
(which he considered his birthright through the union of Solomon and
Makeda of Sheba) encompassed everything from Egypt to Zanzibar, from
Arabia to the Holy Lands and Mesopotamia.77 Dufton asserted that Tewodros’
ambition, simply put, was that ‘Mohammedanism must disappear from his
country;and not only so,he would wipe it off the face of the earth . . . Theodore
is the first and only patriot Abyssinia ever saw’.78 Tewodros was the enraged
incarnation of the chauvinism and mythology contained within the Kebre
Negast; he was the product both of the violent political culture brought
about by the zemene mesafint and of a fertile religious imagination inspired
by the myths surrounding Ethiopia’s antiquity—a deadly combination. In
some respects he was an oddity, too, in that his ire was also frequently
directed at the Church itself, and in that he belonged rather more to the
tradition of militant monasticism than to that of monarchical pomp.
In the end, of course, Tewodros had believed that a natural alliance lay
with the Christian powers of Europe, and could not understand why the
latter appeared less than interested. Considering the great struggle to which
he supposedly dedicated his life—the destruction of Islam—it was a terrible
irony indeed that he would end his own life as a British force advanced on
Magdala, determined to destroy him. Yohannes may have been cleverer, in
some respects, at balancing the various issues confronting him, but he cer-
tainly shared Tewodros’ hatred of Islam, and in that regard the two men were
very much exponents of the Christian nationalism that lay at the heart of
the neo-Solomonic restoration. In the early 1880s, with the rise of the
Mahdist state on his north-west border, Yohannes was anxious about the
loyalty of his own Muslim subjects, and the possibility of a general upris-
ing.79 He, too, had a highly developed sense of the past and its political
utility: his wars—both actual and verbal—with Islam were repeatedly his-
toricized, as he reminded would-be European allies that Ethiopia had always
had to fight the Muslim aggressor in the past. In particular, his reference
point was the early sixteenth-century invasion by Ahmed Gran, around
64 v i ole nc e and i mpe ri al i sm

which much memorial verbiage was constructed;80 in the mid-1890s, the

Russian Bulatovich would note that Gran ‘was an outstanding personality
and to this day still lives in the memory of the people, who ascribe to him
supernatural qualities’.81 Yet apparently somewhat less fanatical than
Tewodros, at least in terms of political vision, Yohannes’ main concern was
rather more specific than that of his predecessor—namely, the Egyptian
presence on the Red Sea coast. And even in this he was prepared to be
rather more placatory than Tewodros, clearly setting considerable store by
diplomacy rather than simply force of arms.
Yet a willingness to negotiate over the contested northern frontier was
only politic, and was in some respects indicative of feelings of insecurity—
after all, Yohannes had seen at first hand what might happen if the ire of a
foreign power was aroused. Nor should Yohannes’ diplomatic overtures dis-
guise his own very real intolerance of Islam within his realm. Muslims in
Tigray, Wollo, or Amhara might be allied to the hated ‘Turk’, he thought,
and used as enemies behind the frontline; the fear of a Muslim uprising
against Christian rule within the provinces was genuine, and grounded in
some historical experience. In order to address this threat,Yohannes moved
aggressively against the Muslim community in Wollo, and forced the con-
version to Christianity of many Muslim leaders and their followers. But
here, professions of faith and loyalty to Yohannes were both insincere and
perceived to be insincere, and indeed many Wollo Muslims fled south, thus
providing further impetus to Islamic missionary activity in the southern
marches in the later nineteenth century. Ultimately,Yohannes’ more imme-
diate religious antagonists were not within his borders but on them, in the
form of both Egypt and the Sudanese Mahdists.
Menelik was cautious enough to grant freedom of religious practice,
apparently in an attempt to assuage Islamic movements across his realm. But
the royal court and the ‘new’ Ethiopian state were to remain resolutely
Christian institutions. The events of the long nineteenth century had con-
firmed in the minds of the Ethiopian political establishment the absolute
need to be watchful of potential Muslim enemies. It was a ‘lesson’ deeply
ingrained in habesha political culture, and in any case was of some consider-
able antiquity. The story of Menelik’s own heir, Lij Iyasu, makes the point
clearly enough: while probably not an actual Muslim convert, he nonethe-
less sought a greater degree of integration of Muslims into government,
including the members of some prominent Muslim dynasties. He showed
some sympathy to the Somali Islamic resistance against the British and the
 state s of v i ole nc e, to C. 18 7 0 65

Italians, and, with the ‘Great War’ in Europe already underway, made over-
tures to the Ottoman Empire, apparently seeking an alliance. It was all too
much: he was deposed and excommunicated in 1916, though he escaped
and remained at large—a rallying point for future opposition—for a number
of years.82 Yet in the course of the twentieth century, his successor—Ras
Tafari, later Haile Selassie—would assiduously position Ethiopia as a
Christian ally of the West in a predominantly Muslim and increasingly hos-
tile part of the world. This was no mere rhetoric, for the Christian state was
itself a product of external as well as internal borderlands.
 4
Borderlands, Militarism, and
the Making of Empire

Resurgent Tigray
The rise of Tigray had been signposted several decades before the emergence
of Yohannes. Modern Tigrayan nationalist writing—scholarly and other-
wise—has described a gradual but inexorable erosion of Tigray’s rightful place
as both the birthplace and the guardian of ‘Ethiopian’ civilization following
the decline of Axum, andYohannes as its (temporary) saviour. Indeed,Yohannes
himself is often credited with giving birth to modern Tigrayan nationalism
itself, while he is also, in many respects, the single most important reference
point in the modern Tigrayan struggle for self-fulfilment and ‘repositioning’
in the Ethiopian political order.1 Tigray’s resurgence was in large part con-
nected to the expansion in trade to which it had better access than many poli-
ties of the southern interior. As Valentia declared in 1808:

Abyssinia is at present . . . torn by civil dissension, as it has been for many gen-
erations, owing to the great power vested in the governors of the
Provinces . . . The Province of Tigre, being the only one that has any commu-
nications with the Arabians, is considered as the most important.2

Yet even within Tigray itself there were significant regional tensions and politi-
cal cleavages, and despite the geopolitical importance of the province there was
no real Tigrayan ‘unity’ until at least the 1830s.3 Ras Welde Selassie’s death in
1817 was followed by several years of conflict until Sebagadis, previously a close
ally of the late ras, seized control in 1822; throughout this period, Tigrayan
political leaders simultaneously aspired to the reassertion of Tigrayan-Amhara
power over the old Ethiopian polity, and viewed with distaste the prevalence
of Oromo leaders in the old Gondarine power structures, a situation, of course,
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 67

which they also ascribed to Amhara weakness. A broad coalition led by the
Yajju noble Ras Marye Gugsa, however, invaded Tigray in 1831 and defeated
Sebagadis—ironically, perhaps, using Oromo cavalry whose predations across
Tigray lingered long in the memory—whereupon Ras Wube was placed in
charge.4 Wube succeeded in achieving some degree of unity, not least by
co-opting the family of the executed Sebagadis, and dominated northern poli-
tics for two decades. As well as arguably creating modern Tigray, Wube posi-
tioned the region at the forefront of both external contacts, with regular access
to the Red Sea coast, and the struggles of the zemene mesafint; he might have
laid claim to the Solomonic inheritance himself, but for his defeat at the hands
of Tewodros in 1855.Tigray’s increased geopolitical importance exacerbated its
conflicts with rivals further south; at the same time, Tigrayan rulers had prob-
lems of their own with the frontier districts of the coast and the lowland plains,
over which they had little direct control but in which Tigrayan forces fre-
quently conducted military operations—not least to secure commerce.
Ultimately, Tigray occupied the somewhat anomalous position of jostling for
position at the heart of the Solomonic state, while simultaneously becoming
something of a political, economic, and indeed ethnic frontier zone. It is a
curious role which in many respects Tigray’s political leadership has yet to
address.
The future Yohannes IV, Kassa Mercha, was born in 1831 into impeccable
aristocratic stock,5 able to claim both Solomonic blood through the line of
his paternal grandmother, and Tigrayan nobility with blood links to the
prominent eighteenth-century family of Ras Mikael Sehul—the man whose
actions had effectively ended the Gondarine monarchy and heralded the
zemene mesafint. In the course of the 1860s, Kassa had consolidated his posi-
tion as one of the major leaders of the north, a key shifta in opposition to
Tewodros’ gradually imploding state, and when the British passed through
his territory in 1868 Kassa proved himself an eager ally. He succeeded in
winning Napier’s support in the form of a significant batch of modern fire-
arms which placed him in a strong position in the forthcoming succession
conflict. The short-lived reign of Takla Giyorgis II—formerly Gobeze of
Lasta, and an erstwhile ally of Kassa, who defeated him near Adwa in 1871—
was followed by Kassa’s proclamation as Yohannes IV in early 1872. As with
Tewodros, a former shifta had become negus negast, albeit one with a rather
stronger claim to the Solomonic inheritance. Yohannes, faced with rivals
both within and outside Tigray (among the latter, Menelik of Shoa), made
strenuous efforts to demonstrate his bloodlines to the Solomonic dynasty,
68 v i ole nc e and i mpe ri al i sm

and had himself crowned at Axum; his coronation ceremony drew on

‘ancient’ tradition and practice, and thus was emphasis also duly placed
on Tigray as the root of habesha culture and civilization.6
Yohannes steeped himself self-consciously in tradition, but his base in
Tigray also brought him into contact with an external world at a time when
European commerce and influence were becoming significant. The world
beyond was encroaching, but Yohannes’ immediate task was to consolidate
his position within both Tigray and Ethiopia.To do so required a deft com-
bination of tact and force. In contrast to his predecessor, he was willing to
allow local notables to accumulate a certain amount of authority, and was
not so swift to enforce submission as Tewodros had been. He proved himself
prepared to use armed force to bring recalcitrant districts to heel—Yohan-
nes’ polity remained a fundamentally military one, and he presided over a
marked degree of militarization and war-readiness—and he did so in both
Gojjam and Shoa, notably; but he nevertheless tended to confirm local
nobilities as rulers and, rather more cunningly than Tewodros, he built intri-
cate webs of alliance by playing off regional overlords against one another.
Thus were Gojjam, under Takla Haymanot, and Shoa, under Menelik, set
against one another, unable to build the kind of broad front against Yohannes
that might have seen the latter rendered impotent.
It was a precarious business, and arguably Yohannes was never truly secure.
Many in the regions continued to undermine and defy him; despite offer-
ing submission in the mid-1870s, for example, Menelik was at the very least
ambivalent in his professed ‘loyalty’ to the Tigrayan negus negast, and spent
much of Yohannes’ reign biding his time and pondering a move against his
supposed master. Nonetheless,Yohannes was able to enforce proclamations
of loyalty from Menelik in Shoa, and from Takle Haymanot in Gojjam, so
that by the late 1870s he was regarded as having ‘pacified’ the region.7 In
1884, it was supposedly boasted that ‘a child could pass through his domin-
ions unharmed’, following his successful bringing to heel of such ‘turbulent
tribes’ as the Azebo and Wollo Oromo, among others; he had increased trade
through his domains, while ‘[t]he peasant and cultivator were also better off
and less molested by the soldiery’, as he was apparently more successful than
Tewodros at disciplining the armed forces at his disposal.8 He also managed
to stabilize his position somewhat by adopting a more conciliatory stance
toward the Church, reversing Tewodros’ policy by restoring some land to it.
He took the title Negus Tsion, King of Zion, and was ruthless in enforcing
religious unity, or at least suppressing open dissent; he was no less brutal in
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 69

his persecution of Muslims than his predecessor. Even so, it seems that
Yohannes—and those close to him—were careful to use the language of
magnanimity and mercy in dealing with his enemies (Christian and Muslim
alike), in stark contrast to Tewodros’ Old Testament smiting and righteous
rage.9 In this way did Yohannes manage to head a rough coalition which was
the foundation of the neo-Solomonic state. Yet it was his external border-
lands which dominated his reign. He perished on one of these, the north-
west frontier against Mahdist Sudan; but perhaps more importantly in the
longer term, while Yohannes was relatively successful in managing his rest-
less polity, he singularly failed to resolve the issue of his external frontiers.
Yohannes’ legacy was two-fold: first, a Tigrayan renaissance which rested on
the idea of rightful inheritance and self-realization; and, second, a failure to
bring stability and security to the Eritrean frontier zone, a failure in which
several other actors—Egyptians, Italians, British—were also complicit. Both
legacies would mature into full-blown crises for the Ethiopian state in the
century after Yohannes’ death.

Frontiers (1): the Mereb River zone

In the middle decades of the nineteenth century, the idea that ‘Ethiopia’ had
the right of direct access to the coast first emerged in its modern form. It
was an idea which would loom increasingly large in the highland imagina-
tion; it would become extraordinarily destructive in the twentieth century,
and to all intents and purposes remains unresolved at the time of writing.
But that is for later in our story: suffice to say here that the basic idea is
rather more novel than is sometimes assumed, and only takes on the proper-
ties of a political issue in the course of the nineteenth century. Prior to the
nineteenth century, and certainly the zemene mesafint, Christian polities in
the highlands had neither the ability nor the desire, on available evidence, to
in any way administer the escarpment and coastal lowlands beyond periodic,
tenuous, and frequently contested suzerainty over the kebessa. Conversely,
there is much anecdotal evidence that the average highlander heartily
despised the coastal zone as an alien and ghastly environment, and the peo-
ples there as ‘Arabs’, shifta, wild and lawless ‘tribes’. However, it became an
article of faith from the late nineteenth century onwards among foreign
scholars and observers as well as the Ethiopian political elite that some
70 v i ole nc e and i mpe ri al i sm

amorphous, ancient Ethiopian state had ‘always’ held the coast, until being
kicked out by the ‘Turks’. It has become one of the most curiously stubborn
misapprehensions in the modern history of the region—and, again, one of
the most dangerous.10 In many respects, to be sure, it is a key tenet of the
‘Greater Ethiopia’ thesis. Yet this is not to adhere to some wrongheaded
endorsement of an extrapolative Eritrean nationalism. The situation in the
nineteenth century was much more fluid and ambiguous than this. Due
recognition needs to be given to the very distinct political ecologies between
the central and northern Ethiopian Highlands and the shores around
Massawa port,11 including those between the kebessa, the escarpment, and
the hot coastal plains. For Erlich, Yohannes’ commander in the north, Ras
Alula,‘fortified Eritrea as Ethiopia’s gate’;12 Alula himself reportedly declared
with appropriate bombast that ‘Ethiopia goes up to the sea; Egypt begins
there’.13 Yet the reality was that ‘the north’—the Mereb River zone, the
kebessa, and the adjacent coast—was seen in the course of the nineteenth
century, and beyond, to present no end of trouble to the highland interior.
Eritrea was Mereb Melash to Tigrayans, Kemerab Wedya to the Amhara—the
land beyond the Mereb, a zone of conflict in which recalcitrant locals and
external enemies alike were to be confronted and overcome. If Ethiopia was
indeed, as Consul Portal had it, ‘a Christian island set in the midst of a
stormy Moslem sea’,14 then Eritrea was the place where the waves broke on
the rocks. Simply put, it was an extraordinarily dangerous fault line, arguably
the single most important zone of conflict in the region.
The Egyptian presence in Massawa was an increasing affront to the
Solomonic state-builders,Tewodros and Yohannes, who were only too aware
of the increasing value of the Red Sea—and fearful of the growing power
of Islam around it. An annual lease was granted by Ottoman Sultan Abd
al-Majid to Muhammad Ali on the Red Sea ports of Suakin and Massawa
in 1846, although it was not until 1865 that Khedive Ismail permanently
annexed them to the Egyptian Sudan.Tewodros was watchful on the north-
ern frontier in the 1850s and early 1860s, although ultimately he lost control
in the north in the face of repeated rebellion. Under Ismail from the mid-
1860s, Egypt embarked on an altogether more aggressive approach to
Ethiopia, and during the reigns of he and Yohannes the Mereb River zone
became one of the key frontiers of violence, with a number of minor skir-
mishes and two major clashes with Egyptian forces in the mid-1870s.
Contemporary European accounts, again, must always be treated with
caution, not least in terms of their crude racial stereotyping, but the
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 71

commentary they contain on the Mereb frontier are nonetheless important.

In the early years of the nineteenth century, Henry Salt considered that
Tigray ‘proper’ was ‘bounded on the north by the river Mareb’, and reported
that the inhabitants of Hamasien ‘are said to bear a very distinct character
from the rest of the Abyssinians’; the Danakil peoples, meanwhile, were
‘completely ungovernable’.15 In 1811, Ras Welde Selassie of Tigray lamented
in a letter to King George III of Great Britain that ‘before me are infidels
and behind me are infidels, on my right hand and on my left are infidels, and
in the midst of those that are all infidels; and all that is on the shore of the
[Red Sea] is infidel’.16 In the 1830s, Hamasien province—on this occasion
rather misleadingly described as stretching to ‘the territory of the wild and
uncultivated Shankallas’—was judged to be inhabited by people ‘barbarous
in their manners, cruel in their tempers’.17 The region was divided into
numerous chieftaincies, each ‘entirely independent of the Ras of Tigre’.18
Sebagadis had succeeded in temporarily imposing tribute on the Saho, but
they remained troublesome: the missionary Gobat recorded that on hearing
of Sebagadis’ death in 1831, the Saho revolted and refused to pay the trib-
ute.19 Likewise Plowden, in the 1840s, reported that the people of the kebessa
were ‘a fierce and turbulent race’, paying only periodic tribute to their
Tigrayan neighbours, while the Saho and Afar constituted ‘a wicked and
treacherous race . . . at enmity with all men’.20 While the communities of
Hamasien and Serae clearly spoke Tigrinya, they were ‘scarcely considered
by the people of Teegray as a portion of that country, whose governors since
[the time of Ras Mikael], have made war on them to enforce payments of
an irregular tribute’.21 Violence, indeed, defined trans-Mereb relations. ‘The
frontier provinces of Teegray toward the sea, Kalagooza [=Akele Guzay] and
Hamazain’, wrote Plowden,

are now disorderly republics, save a tribute forced on them by the arms and
the fortune of Oobeay, and will probably soon entirely detach themselves
from the shaking fabric of Abyssinian society; and here there is no law or
protection for the trader, save such moderation as self-interest may teach the
villagers on the road . . . 22

This gives some indication of the extent to which the violence of the zemene
mesafint was reflected north of the Mereb. Certainly, swathes of the kebessa,
and Tigray too, were devastated by cyclical violence: one account refers to
the ‘devastation and ruin’ in Hamasien in the mid-1840s, and also to the
‘deserted, uncultivated’ district of Shire in Tigray, formerly heavily peopled
72 v i ole nc e and i mpe ri al i sm

and agriculturally productive.23 Ras Wube periodically raided the Bogos

area, too: ‘[b]eing plains, and richly cultivated with Indian corn, they are the
favourite field for the incursions of Oobeay’s troops, affording the most
spoil with the least fighting’.24 This was the familiar pattern of large-scale
organized military expeditions against small agricultural settlements located
in vulnerable and contested borderlands.
This region may indeed have been a happy hunting ground for Tigrayan
soldiers, but it was the violent insecurity of the frontier that was both the
cause and the symptom of the militarized polity further south. Notably, for
all the later claims of ownership of the coast, the littoral itself was an alien,
hostile environment, containing diseases, according to Salt, ‘which produce
in the minds of the Abyssinians that great dread and horror of the coast
which they generally entertain’.25 Just south of Massawa, Tigrayan expedi-
tions mined salt under heavy armed guard, and relations between the
Tigrinya-speaking uplands and the communities of the coast were episodi-
cally violent.26 The same was true in the Tigrayan lowlands, where salt-
mining operations took the form of military expeditions owing to the fact
that the area was ‘infested by a cruel race of Galla, who make it a practice to
lie in wait for the individuals engaged in cutting it’.27 Welde Selassie himself
highlighted the fundamental tension at the heart of Tigrayan territorial
aspirations when he told Salt that ‘he could easily drive [the Muslims] from
Arkeeko or any other point . . . but of what advantage could it be to possess
a barren coast, which at all times the Abyssinians are averse to visit’.28 On a
separate occasion, the Ras declared to Salt: ‘ “You will know that my people
are very averse to go down to the coast for a day’’’.29 Wube’s later claim to
be ‘king from Gondar to the sea of Massawa’ was thus hyperbole indeed;30
nonetheless, he wrote to Louis-Napoleon Bonaparte in 1849: ‘All the land
on the coast is mine’.31
While Tigray’s geopolitical importance was growing in the late eight-
eenth and early nineteenth centuries, the province was also wracked by
internal strife, conflict which spilled over into the kebessa and which further
rendered the plateau volatile. When Welde Selassie was confronted with
Sebagadis’ revolt in 1809, his forces penetrated as far as Debarwa in pursuit of
the rebels but were at length compelled to withdraw, leaving the region of
Akele Guzay and Serae in some upheaval.32 Similarly, the violence with
which Tewodros had to contend in parts of Tigray, as well as north of the
Mereb, prevented him from even coming close to his declared ambition of a
permanent presence on the coast.33 In the late 1850s, notably, the rebel leader
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 73

Neguse was active in the Agame district in Tigray, and in Akele Guzay, where
he managed to establish some form of popular support; from there he raided
across Hamasien, forcing Tewodros’ representative there, Hailu Tewelde
Medhin, to flee into Serae.34 Neguse’s assaults on political strongholds in the
kebessa reflected his own territorial claims across the region, including the
coast.35 One source from 1860 refers to ‘the anarchy which again reigns in all
Tigray’,36 and there can be little doubt that this had repercussions north of
the Mereb, too. In the course of the 1860s, Tewodros was continually defied
in Tigray and the kebessa—despite his description of the na’ib of Massawa as
his ‘vassal’37—and the northern region was a patchwork of garrison-based
military occupation and zones of active hostility.38 One of Tewodros’ key gar-
risons in the north in the mid-1860s was at Keren, comprising ‘villainous-
looking scoundrels’, according to Blanc, evidence that Tewodros ‘was not
very particular as to whom he selected for such distant outposts’.39 In 1866,
as Tewodros’ reign entered its final bloody phase, the north was in turmoil,
with the effects of famine across Tigray and the kebessa exacerbated by the
armed incursions of Gobaze (the future Tekle Giyorgis) of Lasta, who like-
wise sought to impose order in the north, and whose troops clashed with
those of Tewodros in Akele Guzay.40 Detailed evidence is sometimes lacking,
but it seems clear that food shortages drove a great deal of the violence of this
period, as armies and beleaguered farming communities alike sought to
secure new agricultural supplies or protect what they already had. It is clear,
moreover, that the culture of violent confrontation which characterized
Ethiopian politics in the nineteenth century had a dramatic impact on the
Eritrean region; but the troubled frontier in turn served to destabilize habesha
politics and heighten levels of violent insecurity south of the Mereb.
With the fall of Tewodros and the rise to pre-eminence of Yohannes, the
history of the northern zone entered a new phase, with the northward shift
in the balance of power to Tigray having major implications for politics
north of the Mereb. Propagandist rhetoric for Dejjazmach Kassa in 1869 had
it that he was ruler from ‘Tekkeze to Massawa . . . All the land is pacified and
well provided for’;41 but the reality was that the convulsions which followed
Tewodros’ death rendered Yohannes’ power base even in Tigray unstable.Yet
his gaze was indeed fixed on the Eritrean frontier. From the early 1870s, his
chief tormentor was Egypt, which not only sought ‘to prevent me from
having any outlet on the Red Sea’,42 but actively pushed the frontier south-
ward and challenged his jurisdiction at every opportunity. Like Tewodros
before him, Yohannes interpreted the struggle unfolding before him in
74 v i ole nc e and i mpe ri al i sm

religious terms, perceiving a single pan-Islamic threat across a broad arc of

territory from Massawa, through Keren and Kassala, to the modern
Ethiopian–Sudanese borderland.
Egypt, wroteYohannes in 1872, had ‘seized all the country on the other side
of the Mereb where are white (lit. red) and black Barea (slaves), Bogos, Hebarb
[?], Mensa, Ailat, Asgade Bakla, Zoulla,Tora, Sanhali, Amfela, all the soil below
the Afaf [?], which properly belongs to Hamasien’. Moreover, Yohannes was
the lawful master of the ‘Shankelas’.43 Yohannes’ counsellor and confidante
General Kirkham suggested that while the ‘Turks’ might possess Massawa, ‘all
the other Bays, Inlets and Ports’ belonged to Yohannes.44 This broad zone of
conflict would come to define Yohannes’ reign, and his state, with long-term
implications.Yohannes asserted that Egypt had aggressively entered Abyssinian
territory, seizing ‘the whole of the [coastal] lowlands’ and the Bogos region in
the north-central highlands.45 He complained to Ismail in mid-1872 that an
Egyptian army had occupied Mensa, Halhal, and Hamasien: ‘Hamasen’, he
asserted, ‘is the original capital of Abyssinia . . . [and] from these countries my
kingdom extends as far as the coasts of the Red Sea’.46 Yohannes’ distress at his
inability to control this territory is evident in his anger at Munzinger’s appoint-
ment as ruler ‘over Mereb Melash—half of Tigray—[and] all the border-
lands’.47 Similarly, his assertion in 1873 that the lands of the Danakil and
Arafali south of Massawa belonged to him, as they did to his predecessors, has
a ring of desperation to it, for the Turks had recently pillaged the area, and
with apparent impunity.48 The Egyptians had occupied Bogos, seized Ailet,
and built a sawmill at Sabaguma; Metemma—which Yohannes acknowledged
as being ‘half to Egypt & half to Ethiopia’—had been fully occupied by the
Egyptians; in sum, ‘[b]etween Massawa and Metemma, the whole of the low
lands have been invaded, and on the other side of Massawa, Mennsa, Ailet,
Zula, Semhari, Dankali, and the province of Amphilea, have been occupied’.
In so doing,‘the Egyptians have completely isolated Abyssinia from the rest of
the world’.49 Yohannes begged the British to acquire for him ‘an outlet on the
seacoast’, for ‘Anfila and Zula have been my districts since early times’.50
Much of this, of course, was couched in religious terms: Egyptian advances
represented the encroachment of Islam on Christian Ethiopia, and thus was
Eritrea a cultural and religious frontier, as well as a political one, where
Ethiopia’s very existence was at stake.51
It was not simply that Yohannes believed in the historic right of Ethiopia
to the coast, however important this was; it was also the fact that Ethiopia’s
image in the world, its structures, cohesion, and prestige, were being tested
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 75

and defined in Hamasien, Akele Guzay, and Serae. Thus did the land across
the Mereb loom large in Yohannes’ imagination. By September 1875,
Yohannes had dispatched up to 30,000 troops down the escarpment onto
the coastal plain, occupying the key roads which linked Massawa to the
highlands.52 The Egyptians complained that this force was pressing in on
their territory, sending parties of marauders to within sight of Massawa
itself.53 And yet, when in November the Egyptians responded, advancing up
the escarpment and onto the plateau, little resistance was encountered.They
reached the Mereb unmolested.54 In a brilliant tactical manoeuvre,Yohannes’
army had drawn the Egyptians ‘into a most difficult and intricate country’,
and it now pounced at Gundet, attacking from several directions en masse
and wiped out the bulk of the Egyptian force.55 Only weeks later, the
Egyptians suffered a second major defeat, at nearby Gura.56 Not for the last
time in the nineteenth century, however, a habesha victory in the Mereb
zone was not followed up by any significant advance north of it, although
Gundet and Gura facilitated Alula’s temporary occupation of swathes of
Hamasien and the escarpment.
One of Yohannes’ major difficulties—as it was for his predecessors—was
the ambiguity of loyalties and identities across the central Eritrean plateau,
an ambivalence which further contributed to the instability of the frontier.
Tewodros had attempted to use sporadic influence over Hamasien to launch
attacks on Egypt by co-opting local elites,57 but by 1867,Tewodros’ formerly
‘loyal’ appointee in Hamasien, Hailu Tewelde Medhin, had begun to make
overtures to Egypt, no doubt realizing that political realities were changing
and that conflict in Tigray threatened to engulf the area. ‘The frontiers of
Egypt are near ours’, he wrote to Khedive Ismail in November 1867.
‘Therefore friendship and treaty terms ought to exist between us and the
ruler of Egypt’.58 Munzinger, moreover, described a complex array of shift-
ing alliances across the kebessa, with communities either under the sway of
the ‘King’s Governor’ in Hamasien, or owing allegiance to Egypt, or seeking
succour from alternative sources of power—notably in Tigray—in opposi-
tion to central authority further south.59 Yohannes found himself having to
deal with exactly the same problem. Welde Mikael of Hamasien was quali-
fied in his assessment of his own relationship with Kassa in 1869: ‘Although
I am under [the authority of] Dejjazmach Kassa, I am of old the legitimate
ruler of my country Hamasen from my father and my grandfather and my
great-grandfather’.60 Kassa himself was compelled to intervene in a plot to
overthrow his authority, imprisoning Welde Mikael who aspired to ‘rule
76 v i ole nc e and i mpe ri al i sm

beyond the Mereb’.61 The frontier zone was fractured and complex, witness
the shifting policy of the Saho, for example. At the end of the 1860s it was
reported that several Saho communities were ‘banded together in order to
hold their own against the Tigre people of the plateau, who are Christians,
and consequently their bitter foes. For this frontier service they receive the
protection of the Egyptian Government’.62 Yet in 1873 a group of Saho
chiefs complained bitterly that the Egyptians were now attempting to
impose taxes on them, prompting them to declare: ‘Our ruler and our lord
is the king of Ethiopia . . . We protect the caravans when they descend from
Ethiopia, and we escort them to the capital of the kingdom’.63
Such ambiguity and/or outright rebellion continued to undermine
Yohannes’ imperial project through the 1870s; confronted with violence
from different directions, Hamasien elites frequently opted for a strategy of
playing larger powers off one another in order to maintain some degree of
autonomy.64 More ‘popular’ feelings are clearly difficult to assess, though
one source asserted at the beginning of 1876 that the people of Hamasien,
which had been evacuated by Yohannes’ forces, ‘are reported to be friendly
to the Egyptians’.65 There was certainly much Egyptian propaganda to this
effect.The Moniteur Egyptien of December 1875, while omitting to mention
the defeat at Gundet, emphasized that ‘the Egyptian soldiers were welcomed
with great joy by the inhabitants of Hamasien’. At the Mereb itself, the peo-
ple of Gundet had warned the Egyptian force of the Ethiopian advance and
begged protection from the ‘invaders’.66 Whatever the case, it is clear that
Yohannes’ largely unprofitable victories over the Egyptians at Gundet and
Gura in 1875–6 exposed the fragility of his position in the kebessa. Welde
Mikael fled to Massawa to join the Egyptians, and the khedive proposed
placing him in charge of Hamasien as a vassal, creating a territory which
‘would be a good barrier between Abyssinia and our territory’.67
Yohannes himself surely recognized the futility of his victories: he imme-
diately sought a reconciliation, though not before reminding the Egyptians
that the war started because ‘the frontiers have not been respected and
because you have listened to the rebels in the kingdom of Abyssinia’,68 an
admission that the frontiers of the kebessa were in fact riddled with ‘rebels’.
Yohannes lamented that Welde Mikael, whom he had appointed as gover-
nor of Hamasien, ‘has betrayed us and joined you’, while ‘[y]ou have given
the reins of government to the inhabitants of Akkele Guzay in order to
rouse them against us’.69 Matters were further complicated, of course, by the
fact that the great rivals of Hamasien,Tsazegga and Hazzega, were themselves
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 77

engaged in escalating conflict through the mid-1870s.70 This aspect of nine-

teenth-century kebessa history requires much more research; but it is cer-
tainly clear that Welde Mikael himself, having sided with the Egyptians,
found the latter less than enthusiastic about responding to his requests for
assistance against Yohannes’ ‘pillaging’.71 Across the Eritrean plateau, vio-
lence escalated from the late 1870s onward, between rival groups whose
loyalties—or at least requests for protection—were directed toward either
the Egyptians or the Tigrayans.72
Much of Eritrea in the nineteenth century was defined in terms of crimi-
nality, banditry, and ‘illegitimate’ violence. A great deal of this was racial or
‘tribal’ stereotyping, of course; but it is indeed the case that war spawns new
forms and cultures of violence and definitions of ‘criminality’, particularly in
the context of the frontier. ‘Banditry’—shifta activity—went hand-
in-hand with social displacement and the mobile military frontier. From
early in his career, for example, Yohannes was keen to project the image of
legitimacy in his use of violence: as Kassa, he declared in 1869: ‘all brigands
[and] predators (lit. lions) have been subdued in the country I rule’.73 He was
the legitimate state, and thus his use of force was righteous; his enemies were
criminals and, indeed, little better than wild animals. At the level of the state,
it was crucial to project ideas about order and rectitude, and there can be
little doubt that the situation on the troubled frontier contributed in no small
way to Yohannes’ need to create such imagery. In the last two years of
Tewodros’ reign, as his state descended into chaos, so too did the northern
frontier, as we have noted: in mid-1866 it was reported that the route to the
coast through Akele Guzay was virtually closed because of bandits, with
merchants instead attempting to use the Serae road. Moreover, ‘[r]obbers and
bandits have become very numerous in Tigray’.74 Again, there can be little
doubt that much of this criminalized and paramilitary activity was driven in
large part by food shortage across the north. Such supposedly ‘illegitimate’
violence proliferated on the frontier, perpetuated by the excluded and the
disaffected, the criminal and the ‘ungovernable’. Around Massawa, for exam-
ple, southbound caravans were plundered with impunity throughout the
nineteenth century,75 and references abound to an endemic predatory war-
fare on the roads from Massawa into the highlands, with travellers depicting
a no-man’s-land where ‘wild’ groups preyed on unsuspecting merchants.76
Littoral and escarpment Eritrea was a zone of ‘illegitimate’ and stateless
violence. The much-maligned Saho, again, had gained a reputation for rob-
bery by the 1860s,77 while according to Portal they inhabited a country
78 v i ole nc e and i mpe ri al i sm

‘notorious for its breed of robbers and professional brigands’.78 By the 1880s,
indeed, ‘murder and brigandage’ had become so common ‘as almost to put a
stop to all trade’. Between the Italian position at Sahati and Alula’s base in
Hamasien was ‘disputed territory . . . said to be infested by wandering bands
of brigands and evil-disposed Arab tribes’.79 While some bandits were thus
held to be genetically ‘evil’, others were exiles and renegades from the high-
lands, one source identifying a ringleader of attacks on Abyssinian caravans as
Debbeb, no less than a cousin of Yohannes himself. According to the source,
moreover, ‘[t]he Governor of [Massawa] and other officials winked at (if they
did not actively encourage) the raids on the Abyssinian caravans, and allowed
the proceeds of the robberies to be sold openly in the bazaar’. Debbeb,
indeed, recruited his ‘robber-band’ in Massawa itself.80 The authorities at
Massawa saw this, no doubt, as war by other means: it was clearly in their
interests to encourage such ‘lawlessness’, particularly in respect of caravans
carrying firearms to Yohannes.81
Above all, Yohannes undoubtedly articulated habesha territorial claims
more clearly and more powerfully than anyone before him. Gerald Portal
wrote in the late 1880s that Yohannes
consistently denied the right of the Italian or of any other foreign Government
to be at Massowah at all. He maintained that by right of descent Massowah
and all the south-western coast of the Red Sea had for centuries belonged to
Abyssinia.Tradition lives long in Abyssinia; as far back as the sixteenth century,
the superior armament and discipline of the Turks had driven the Abyssinians
from Zeyla, and later from Massowah . . . [D]uring all these 300 years, argued
King Johannis, Abyssinia had never given up its claim to the sea-coast; the
Turks, and subsequently the Egyptians, had only held these places as they had
acquired them—by the power of the sword.82

Yohannes himself, of course, understood very well the power of the sword:
his own advances north, across the plateau and down onto the coastal plain,
were exercises in military adventurism.

Frontiers (2): Sudan and the north-west

Egyptian expansion between the 1820s and the 1840s from their newly
established base at Khartoum encroached into lands either traditionally
claimed by Christian highlanders, or which at least had been regarded as
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 79

carefully maintained buffer zones along the northern and western escarp-
ments and lowlands.83 In particular, the Egyptians—having occupied much
of the Nile valley—moved into the area of what is now the western lowland
of Eritrea and north-east Sudan, abutting the north-west approaches of
Ethiopia. The Gash-Barka zone was thus emerging as another key border-
land in the making of the modern region—and would remain key for
Ethiopia, Sudan, and later Eritrea itself. Yet the advance of the Egyptians
into this area by the early 1840s was ambiguous. For while this development
unfolded as a territorial (and indeed religious) threat, at the same time the
revival of Red Sea commerce which it prompted, again, was of enormous
benefit to Ethiopian territories in the northern and western highlands.
Tigray in particular was well-positioned to take advantage of the trade; so
too were the territories abutting Sudan, especially Shoa, whose rulers were
increasingly able to enrich themselves to the detriment of the imperial
court at Gondar.
From the 1820s, increasingly aggressive Egyptian commercial, military,
and religious expansion proceeded apace into southern Sudan, including the
upper Nile basin and equatorial regions, and the area of Bahr al-Ghazal and
Darfur, which had been reached by the mid-1870s.84 The Ethiopian–Sudanese
borderlands had long been zones of conflict and ambiguity, and in many
respects Egyptian imperialism ‘merely’ breathed new life into these, much as
Mahdism would along the Christians’ north-west frontier. In particular, the
Ethiopian–Sudanese marches had long been a destination for the displaced
and the discontented, thus rendering this zone politically volatile. Groups in
political or religious opposition to successive regimes in the Ethiopian
Highlands and in the upper Nile basin took to these borderlands, where they
merged with refugees from the slaving expeditions so common throughout
the nineteenth century. With the expansion of Turco-Egyptian administra-
tion from Khartoum, a number of groups took to the marches, and Ali
Khurshid Agha, Muhammad Ali’s administrator in Sudan from 1826, spent
much of his lengthy tenure seeking both their return and the stabilization of
the frontier. In terms of the former, he was moderately successful from the
late 1820s—Sheikh Ahmad al-Rayyah al-Araki submitted, for example,
bringing with him thousands of his Arakiyyin followers—but his success on
the borderlands themselves was somewhat limited. Campaigns against the
Hadendowa in Taka in the early 1830s were at best inconclusive, while fron-
tier wars further south against Kanfu—governor of Kwara and the paternal
uncle of the future Tewodros—rumbled on for several years during the 1830s
80 v i ole nc e and i mpe ri al i sm

with mixed results. In the later part of the decade, fearful that Kanfu was
seeking the annexation of Gallabat—which would have rendered that prov-
ince a haven for anti-Egyptian resistance—Khurshid prepared a major
assault on the Ethiopian marches, only for Muhammad Ali to be warned by
the British not to attempt any serious incursions into habesha territory.
Khurshid’s successor was Abu Widn, and he too was confronted with the
ongoing problem of rebels fleeing to the marches. He was rather more suc-
cessful than his predecessors in campaigning against the Hadendowa and into
the western Eritrean lowlands: under him, what would become the province
of Taka became somewhat more firmly established in the 1840s, and Kassala
expanded as a key strategic settlement on the site of a military camp.
Between the 1840s and the 1860s, an increasingly aggressive Egyptian pres-
ence was established in southern Sudan and the western lowlands of present-
day Eritrea, and was creeping into the Christian highlands around Agordat
and Keren. The latter area became, in the course of the nineteenth century, a
decidedly ambiguous frontier, and would evolve into an even more explosive
one in the twentieth; it was inhabited by a range of pastoral and semi-nomadic
groups which at best had shifting and multiple relationships with surrounding
expanding statehoods, and which would take advantage of the eventual ‘for-
malization’ of international boundaries in the area by using them as frontlines
in ongoing localized animosities, and as markers in the evolution of larger
(and politically expedient) regional identities. The creation of the Egyptian
administrative province of Taka—to the east of Khartoum, encompassing
Kassala and the Gash-Barka plains, and stretching into the foothills of the
kebessa—with a fortified outpost at Keren constituted a direct threat to Amhara
and Tigrayan control to the south. The same was true of the consolidation of
Egyptian control of Massawa on the Eritrean coast, leased from the Ottoman
authorities. Thus was a northern arc formed which effectively constituted a
new and aggressive front blocking the highlanders’ view of the Red Sea and
Mediterranean worlds which would become ever more important to them.
This was the habesha dilemma: those who were necessary as trading partners
for the highlanders were also, ultimately, political and religious antagonists. It
contributed powerfully to the habesha sense of encirclement, while Ethiopian
highlanders were increasingly frustrated by the seeming lack of interest of the
European powers who were supposed to be their natural spiritual allies.
New political and indeed religious communities were emerging else-
where, too, in the course of the nineteenth century, notably on the
Ethiopian–Sudanese borderlands.85 On the Ethiopian side of the frontier,86
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 81

the sheikhdoms of Aqoldi (or Asosa), Beni Shangul, Gubba, and Khomosha
were the result of a larger aristocratic network—that of the Watawit, of
Sudanese origin—over the Berta and Gumuz. These would be increasingly
squeezed between two expanding imperialisms, much like the peoples of
the Gash-Barka region, namely the Egyptian (later Anglo-Egyptian) Sudan
and Solomonic Ethiopia—although neither would be in a position to exert
control over the lands between them, which indeed came to define the
fractures and fissures of those empires themselves. If it is broadly accepted
that across north-east Africa states and societies have only ever been as
‘strong’ as their weakest and most volatile borderlands, and were indeed
ultimately defined by the dynamics of those borderlands, then what we do
know about the Ethiopian-Sudanese frontier territories tells us a great deal
about the nature of the political missions on either side of them throughout
the nineteenth and twentieth centuries. For the habesha, of course, these
were the lands of the shangalla or baria further north—savages and slaves,
usually with some derogatory inference concerning darker skin colour—
and these territories were indeed key hunting grounds for the predatory
highland state. This was politically motivated, and racially justified, slaving
violence. In many respects the western frontier was the one by which high-
landers judged their own cultural and genetic supremacy.
We have already observed that Yohannes perceived one continuous
‘frontier’ from the coast to the interior lowland plains, but for our purposes
it is important to look briefly at the region now known as the western
lowlands of Eritrea, as in some ways it constituted a distinct geopolitical
sphere. As with the central Eritrean highlands, this was a frontier which
reflected anxieties at the very heart of the habesha state, and which had
troubled highland rulers throughout the nineteenth century, particularly
since the conquest of the area by Muhammad Ali. The advance of the
‘Turks’ into the vicinity of Sennar and Metemma in the late 1830s led Sahle
Dingil to reflect ruefully that ‘a conqueror recognizes no boundaries’, and
that they had seized territories which ‘were the dominions of our king’.87
The lowland zone immediately north of Tigray was an ambiguous no-
man’s-land, known in Tigrinya, according to one source from the mid-
1840s, as ‘Addy Barea’, country of the slaves.88 To the Amhara, this formed
part of the vast shangalla frontier district between habesha and Arabic civi-
lization.89 The area was another of Wube’s hunting grounds, if not always as
fruitful as the Bogos region. In one prolonged campaign in 1844, Wube’s
troops were repeatedly harassed by the ‘Baria’—presumably Nara or related
82 v i ole nc e and i mpe ri al i sm

groups—and at length became lost and ran out of water, owing to ‘igno-
rance of the country’. Some did, however, succeed in returning with slaves
and cattle, despite fierce resistance.90 So wide ranging was Wube’s campaign
on this occasion that his troops inadvertently or otherwise attacked several
groups on the Egyptian side of the frontier zone.91 In the then thickly for-
ested Takkaze River valley, moreover, renegade ‘Arabs’ established military
camps away from the encroaching Egyptian authorities in Taka.92 The low-
land area swiftly gives way to the rugged spur of mountains north of
Agordat and Keren, the Halhal district, which reached Nakfa further north
still—the area, as we see in Part IV, of such significance in the history of the
Eritrean nationalist struggle. In the mid-nineteenth century, Halhal was
likewise a contested zone, prone, as Plowden observed, to attacks ‘alter-
nately by the troops of Oobeay, and the Turks of Sennaar, from the quarter
of Taka or Gasch’. The pastoral and semi-pastoral communities of these
hills periodically paid tribute to whoever was perceived as the greater
threat, whereupon, as Plowden memorably described it,
these districts are ...left to govern themselves as they please, and to replace
their losses by plundering their neighbours. As may be supposed, the only law
is that of the strongest; and wars and blood-feuds, betwixt man and man, vil-
lage and village, tribe and tribe, as each may find that exciting pastime agree-
able, are the only occupations that vary the monotony of driving their flocks
and herds to pasture.93

The Egyptian position in the mid-1870s was that ‘Abyssinian’ raids across
the plains of present-day Gash-Barka province had compelled them to
secure their frontier with reinforcements. But the alleged raiders were no
‘Abyssinians’, and this misreading—whether deliberate or otherwise—of
particular groups’ loyalties and motives only served to underline the ambi-
guity of the frontier zone. One source suggested that
the marauders belonged to the Bari tribe, who inhabit the country immedi-
ately south of Kassala, and are notorious for their predatory and lawless habits,
and who have probably a very indistinct idea of the nature of a frontier
between two States, and have not been as yet brought under subjection to
Prince Kassai’s Government.94

But the Egyptians held Yohannes responsible for any ‘incursions’ into Taka,
and certainly rejected the idea that Bogos belonged in any way to ‘Abyssinia’.
The district was supposedly non-negotiable as a part of Egyptian-Sudanese
administration, although one Egyptian source conceded that ‘the peoples
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 83

and barbarous tribes placed in this condition have naturally very little stabil-
ity; they are very restless and very mobile’. Moreover, certain acts by the
government—the punishment of insurrection, or the reinforcement of the
border against potential enemies by moving troops into the area—inevitably
caused such peoples to flee towards ‘Abyssinia’. However, this, asserted the
source, was no indication of sovereignty: the people of Bogos, in any case,
had come to accept Egyptian overlordship. Now Yohannes had destabilized
the region by positioning some 15,000 troops, deployed under five com-
manders, in a broad front stretching from Hamasien to north-west Tigray.95
Yohannes needed to maintain relations with whatever power occupied
the north-west frontier. In early 1884, Yohannes received ‘with great distinc-
tion’ a Beni Amer embassy which had come ‘to ask his aid against the
Mahdi’.96 The emperor’s response is unrecorded—although in July 1884 he
moved with a large force into the Barka area, effecting a rendezvous with
Alula’s army moving down from Hamasien, ostensibly to collect his annual
tribute97—but it is clear that he had wider concerns when dealing with the
communities of the north-west frontier. Yohannes was confronted with a
substantial Muslim population in his domains, and his ongoing conflict
with Egypt heightened tensions in this regard, even more so with the rise of
the Mahdist state. Yohannes had reason to fear that a direct clash with the
Mahdists would prompt a general Muslim uprising across the region.98 The
Mahdist state, of course, had its own ambitions in the region, with a bloody
clash between the Mahdists and Ras Alula’s forces taking place near Kassala
in 1885;99 and when the first major Mahdist invasion came, in 1887, habesha
military capacity was extremely stretched, as Wylde explained:
. . . some [forces] were in the north watching the northern frontier both against
the Italians in the east and the Dervishes in the north-west, others were at their
homes cultivating, and some in the south-east watching the Danakils and
Gallas, who had also been incited to attack Abyssinia. King [Takle Haymanot]
of Godjam had only his badly armed population and few soldiery . . . The
Dervishes gained, after severe fighting, a complete victory . . . 100

A few years later, the Italians themselves halted a Mahdist expedition into
Eritrea at Agordat101—although that is a different stage in the story of this
frontier, and one which we will pick up in due course.
Yohannes had in fact sent placatory messages to Khartoum in this period,
suggesting among other things that they had a common enemy to fight—
namely, European imperialism. Such offers of peace were haughtily rebuffed,
however, and the Ethiopians prepared for war.When in 1889 Yohannes went
84 v i ole nc e and i mpe ri al i sm

forth to give battle, near Metemma, his forces were initially successful, until
an apparently stray bullet ended the emperor’s life and the Ethiopian force
fell into disarray. Never the most loyal of Yohannes’ servants, Menelik none-
theless held off from outright rebellion until the emperor had been thus
felled. Throughout the 1890s, there was something of an uneasy truce
between Abdullahi’s state in Sudan and Menelik’s Ethiopia; to all intents and
purposes, the Mahdist jihad against the Christian empire was abandoned.
The Mahdists would, in any case, soon have their own wars to fight: the
Italian advance into the Eritrean highlands, and then down into the low-
lands, in the course of the 1890s was achieved by a military victory at
Agordat and then the seizure of Kassala. Ultimately the Mahdist state would
be conquered by the British, from the north, and between them—on
paper—Britain, Italy, and Ethiopia would now delineate their territories
with clearly marked boundaries, even if, in Rome, it was imagined that
Ethiopia was but a temporary actor and Eritrea only a springboard for much
more dramatic advances. In reality, those borderlands remained insecure, and
would become the breeding grounds for ever more militant, if moveable,
identities.

Pax Solomonia? The restless militarism

of Menelik’s empire
While Menelik may have briefly considered making his grab for the rein-
vigorated imperial throne in the aftermath of Tewodros’ destruction in the
late 1860s, when he was still only in his mid-20s, he was dissuaded from this;
never wholly reconciled to Yohannes’ accession, however, his was, again, an
ambivalent loyalty. In the late 1860s, he commanded between 40,000 and
50,000 men, of whom only a relatively small number were armed with
muskets; this number would increase dramatically over the next twenty
years, and thus was Shoa able to expand through both firepower, manpower,
and clever political agreements at the local level.102 A skilled political strate-
gist, Menelik spent much of the 1870s and 1880s creating the political and
material groundwork for an Amhara imperial renaissance, consolidating
his own position, expanding the borders of Shoa, and doing a brisk trade
with European traders via Zeila and Djibouti. In that way was he able to
stockpile large amounts of modern firearms, one of the largest such arsenals
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 85

anywhere in nineteenth-century Africa, at least under a single ruler.

Yohannes’ mounting problems with the Egyptians, then with the Italians,
provided Menelik with his opportunity; and even as Yohannes had been
preparing to make war on the Mahdists, the issue of internal disunity was
once again rearing its head—as the emperor sought to assemble an army on
the north-west frontier, rumours abounded that both Takla Haymanot in
Gojjam and Menelik in Shoa were plotting revolt. It was ironic indeed that
Yohannes was undermined in his dealings with both the Italians and the
Mahdists—representing two of the gravest threats to the habesha polity for
some considerable time—by his own internal enemies;103 but Menelik’s
achievement in subsequently stabilizing his fractured inheritance, if only
temporarily, was nonetheless remarkable.
When Yohannes was killed in the fighting in 1889, Menelik swiftly pro-
claimed himself negus negast, and in truth there was little significant opposi-
tion.104 There were few serious alternatives, especially following Menelik’s
military success in the south; the imperial army under Yohannes was in
some disarray following the debacle at Metemma, and—perhaps most criti-
cally of all—Menelik’s aggressive action came against the backdrop of one
of the region’s worst famines in living memory. Famine, or at least severe
food shortage, had had a major impact on the intensity and flow of violence
earlier, notably in the late 1860s.While those shortages may have been man-
made in the mid-nineteenth century, at least in large part, the ‘Great Famine’
of the late 1880s and early 1890s was caused first and foremost by a failure
of the rains. Nonetheless the catastrophe intersected with political and mili-
tary developments. The effects of famine were exacerbated by the cumula-
tive militarism of the nineteenth century which placed massive strains on
northern resources; in many respects, this junction of phenomena provided
a newly aggressive and increasingly urgent impetus to southward expansion
on the part of the habesha state from the early 1890s onward.
It was to Menelik’s advantage, particularly across swathes of the north and
north-east where potential opposition was dissipated by hunger and dis-
ease—the cattle epidemic, rinderpest, would make its appearance at the same
time—and his preparedness to act decisively in the event of Yohannes’ death
was crucial. However, he did find it necessary to tour the north in some force
to ensure obedience and acceptance; he was well aware that there were many
irreconcilables in Tigray who still waved the flag of Yohannes, not least the
redoubtable Ras Alula and Yohannes’ son Mangasha. He was met there with
a great deal of hostility—to him personally, as a Shoan and an Amhara, and
86 v i ole nc e and i mpe ri al i sm

also to the state structure he represented. Nonetheless it was not sufficient to

divert him from his path; he also enjoyed the support of the Italians, who
accordingly supplied him (somewhat myopically) with arms and ammuni-
tion.105 In late 1889, Menelik was formally crowned at Entoto, on a hill
which today lies on the outskirts of Addis Ababa—the capital which Menelik
and his wife Taitu (whose idea it largely was) would establish shortly after-
wards. Once again, the centre of political gravity had swung decisively away
from Tigray and back south to the Amhara.
Ethiopia was also fanning out below the central plateau along an expand-
ing economic frontier, deep into the south toward Lake Turkana, and to either
side, south-west toward the foothills of the Blue Nile and the river Akobo, and
into the hot plains of the Afar and the Somali to the east. Again, to a very real
extent this process of southern expansion was driven by the need for food and
a range of other exportable commodities to both offset the effects of chronic
famine in the north, and also to pay for the greatly swollen and massively
expensive militarism which was now the defining feature of Ethiopian politi-
cal culture. The Amhara genius was the ability to achieve these remarkable
imperial feats in the age of European hegemony, and while the Amhara them-
selves were simultaneously the target of European aggression. Shoan con-
quests in the south and west, and also to the east into Somali territory, were as
dynamic and ultimately as successful as any European imperial partition in the
region. In fact, to a very real extent Menelik was merely continuing the work
begun by his grandfather, Sahle Selassie, who was already expanding the
southern and western borders of Shoa into Oromo territory in the 1830s.106
Yet initially, Menelik was confronted with the dominance of Gojjam, both in
the Gibe River area and among the western Oromo groups: by the beginning
of the 1880s, Takle Haymanot had established himself as negus of Kaffa in the
south-west, and had had his overlordship recognized among the Oromo states
of the Blue Nile region. A brief but intense Shoan–Gojjami rivalry ensued,
culminating in Menelik’s triumph at the battle of Embabo in early 1882,
which essentially broke Gojjam’s hegemony in the west and south-west. In
the course of the 1880s, Menelik extended his control over a swathe of Oromo
territory—including Leqa Nekemte and Jimma—over which he instituted a
new form of control in the conquered areas, not merely extracting but creating
a form of indirect rule through tributary relationships with Gurage and
Oromo chieftaincies.There was economic benefit to be had for elites on both
sides of the arrangement, and indeed Menelik was also able to recruit Oromo
into his own forces for further expansion.107
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 87

Yet there could be little doubting the extreme violence that had accom-
panied this new dispensation of power in what would become southern
Ethiopia. First-hand testimony provided by the Russian military attaché
Alexander Bulatovich is worth quoting:
The Abyssinians pursue two goals in the governing of the region: fiscal and
political security of the region and prevention of an uprising. All families are
assessed a tax. This is very small . . . In addition, families are attached to the
land. Part of the population is obliged to cultivate land for the main ruler of
the country, and part is divided among the soldiers and military leaders. The
whole region is divided among separate military leaders who live off their
district and feed their soldiers.

At the same time, this had been a destructive and violent imperialism:
The dreadful annihilation of more than half of the population during the
conquest took away from the Galla all possibility of thinking about any sort
of uprising. And the freedom-loving Galla who didn’t recognise any authority
other than the speed of his horse, the strength of his hand, and the accuracy
of his spear, now goes through the hard school of obedience.108

These were clearly districts which had been conquered directly, having
originally resisted; others submitted willingly, and Bulatovich names
Jimma, Wellaga, and Leqa Nekemte, where ‘the former order has been
preserved . . . The Abyssinians obtain taxes from them and do not interfere
in their self-government. Aside from the payment of taxes, they also feed
the troops stationed there’.109 In part, then, the military hegemony of the
mid- and late 1890s rested partly on sheer force of conquest, partly on
cultural assimilation, partly on loose tributary and tax-based levels of
suzerainty. Garrisons of troops, sometimes several thousand strong, were
stationed across the conquered zone; these often became substantial urban
settlements, such as Bareilu, described by Bulatovich, which also had the
advantage of lying on the main commercial route between Shoa and
Wellega.110 Meanwhile, Amhara settlers were increasingly to be seen across
the south.111 In sum, a more stable, permanent, and economically driven
administrative system had been created on the expanding frontier.
Bulatovich wrote approvingly that in participating so decisively in the
scramble for the region, Menelik was ‘only carrying out the traditional
mission of Ethiopia as the propagator of culture and the unifier of all the
inhabitants of the Ethiopian Mountains and of the related tribes in the
neighbourhood’.112
88 v i ole nc e and i mpe ri al i sm

Considerable bloodshed also attended the subjugation of the Arsi, south-

east of Shoa, who put up several years of dogged resistance until their inter-
nal divisions and Shoa’s superior firepower put it to an end, in 1886. Harar
was Menelik’s next eastward target: the emirate had been evacuated by the
Egyptian garrison in 1885, and under Abdullahi was threatening something
of a resurgence—here, Menelik was able to argue to his British and Italian
counterparts in Somali territory further east that this was a dangerously
unstable borderland that he was best placed to subdue. They agreed, and
after initial setbacks the Shoans seized Harar in early 1887. Shoan expansion
entered a hiatus owing in large part to Menelik’s concern with domestic
politics, and also to the awful famine which now raged; but from the early
1890s there were prolonged wars of conquest against the key southern king-
doms of Welayta and Kaffa, while several chiefdoms on the Ethiopian-
Sudanese borders—including Illubabor, Aqoldi, and Bela Shangul—were
incorporated.113 Some of this imperial expansion was piecemeal, however.
Famine and disease in the late 1880s had caused much hardship across a
wide area, prompting local habesha administrations to raid into neighbour-
ing areas less severely hit, in search of resources. For example, Ras Mekonnen,
newly appointed governor of Harar and father of Haile Selassie, organized
a series of raids into the Ogaden and seized livestock from the Somali; such
limited-objective raiding swiftly became conquest. Ethiopian commanders
sought local allies, and easily prized open fissures between Somali clans, thus
establishing another form of indirect rule, like that in the south-west, in
which more permanent governance was achieved through particular Somali
clan heads. Ethiopian military garrisons were founded across the Ogaden,
as they were across the south and west, a heavily armed presence which
ensured that—certainly following Menelik’s victory over Italy at Adwa,
examined in the next chapter—Ethiopia needed to be treated with respect
as a ‘player’ in the region’s international politics. Menelik was surrounded by
some gifted commanders—among them Ras Mekonnen and Ras Gobana—
and that military establishment had proven itself easily the equal of the
European colonial militaries in adjacent territories. This was a military
expansion which also brought about a major economic shift from north to
south, as Shoa became the treasury into which the riches of the southern
territories were poured: the profits of gold, ivory, coffee, and indeed slaves
were reaped by the Shoan elite, who also embarked on a process of land
alienation across the south, with estates of land and commercial concessions
distributed by Menelik to a loyal soldier-administrator-settler class which
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 89

would form the socio-political backbone of the Ethiopian state for decades
to come. Local communities were reduced to what amounted to serfdom, a
state of servitude on the land now owned by armed settlers from the north.
Gabbar, the term for farmer with serf-like connotations in southern Ethiopia,
now laboured under a class of neftennya, which literally means ‘one who
owns a gun’, referring to the soldier-settlers across the south to whom Addis
Ababa granted rights over both people and land.114
Menelik’s kingship was a curious amalgam of the old and the new,
founded on violence but in many respects maintained through negotia-
tion—and spectacular Italian failure. In the course of the campaigns of
military and political expansion of the 1880s and 1890s, the new Ethiopia
was born, and a series of new frontiers—in physical terms, as well as in the
contexts of ethnicity, culture, belief—came into being, this time enclosed
within a shared sovereign space. Again, however, perhaps the most remark-
able characteristic about the empire founded by Menelik in the years either
side of 1900—a characteristic which would be inherited by Haile
Selassie—is that it was founded upon both violence and co-option. It was
not his post office, his banking system, his railways, and the various other
attempts at the stuff of modernity, but his ability—in the short to medium-
term, at least—to manage the vast and fractious territory he had helped
bring into being. Habesha culture was both hegemonic and assimilation-
ist—perhaps all successful cultures are—and the reasons behind the attrac-
tiveness of the Amhara–Tigrayan cultural complex to a host of incorporated
peoples are doubtless manifold. Many Oromo had been similarly drawn to
it since the sixteenth and seventeenth centuries. Christianity held its own
attractions, and certainly its monotheism explains much; above all, it was
the richly historic, articulate, and literate expression of habesha culture
which no doubt drew many to it, even grudgingly. All that said, however,
many more would come, in the course of the twentieth century, to reject
it violently. Although Menelik’s military machine had, for the time being,
imposed a pax on particularly hostile frontier zones, he had created deep-
ening pools of hostility for the longer term. Menelik would have been only
too aware of this, as his army was continually attacked by Tigrayan, Azebo,
and Oromo peasants alike on its march back from Adwa, the scene of the
emperor’s greatest triumph.115 The Azebo and Afar in the north-east, for
example, may have been temporarily becalmed in the face of modern
weaponry;116 but their resentment at habesha cattle-raiding would only
increase over time. Likewise, Tigray was subordinate but simmered with
90 v i ole nc e and i mpe ri al i sm

bitterness at the impositions of ‘southern Abyssinians’ who had impover-

ished vast districts in the run-up to the battle of Adwa, and at the apparent
greed of Menelik who kept the compensation paid by the Italians for him-
self, even though Tigrayan leaders ‘were the chief sufferers by the war, and
bore the brunt of the fighting’. As Wylde observed, ‘it is not at all unlikely,
that it will bear fruit in the future, and make the northerners more eager
to improve their present condition’.117 As another contemporary source
noted, ‘everywhere [Menelik] found his Shoans received with hatred and
contempt by the inhabitants of Tigre. The latter had not yet forgotten that
they were the true leaders of the Ethiopian race’.118
The ‘modern’ Ethiopia that came into being in the last twenty years of the
nineteenth century was the product of a restless, cumulative militarism. The
political system rested heavily on the effective deployment of armed force;
this was a political culture, and a social system more broadly, that was the
outcome of prolonged violence. It was, in turn, a fragile, volatile polity in
which arms had come to play a dangerously inflated role in political and
indeed economic development. In terms of the latter, notably, the military
burden would prove huge. This was a system open to abuse by ambitious
state-building elites, and which clearly facilitated authoritarianism; yet it was
also characterized by a degree of social mobility, and an access to the means
of socio-political change, which meant the empowerment of a military class
capable of inducing such change, should the opportunity and the will arise.
A large number and range of sources through the nineteenth century
point toward the increasing size and significance of both soldiery and mili-
tarism in the habesha socio-political order.119 Young men—and not a few
older men—were drawn increasingly to military service, and a vastly swollen
military establishment was the cumulative effect of cyclical violence as well
as economic insecurity. From the village level upwards, men had long been
eligible to follow their local leaders to war for a defined period of time;120
but the nineteenth century saw a widening and a ‘professionalizing’ of such
service, and an apparent increase in the numbers of camp followers.121
Moreover, firearms—and modern ones, at that—were common, and a veri-
table gun culture had emerged across the highlands by the 1880s and 1890s.
It was a reasonably recent creation, for while during Tewodros’ reign only
imperial troops were generally able to acquire guns,122 the expansion of
trade and a dramatic increase in supply meant, in Wylde’s words, that:
The peasant is no longer miserably armed with spear and shield, or sword and
shield, but is generally the owner of a fairly modern breech-loading rifle, and
 borde rland s, m i l i tari sm , and th e mak i ng of e m p i re 91

has a good store of cartridges, and can always procure more on next local
market day, where they are openly sold or bartered and count as coin.123

Firearms became the instruments of political power from the 1870s onward,
although always in conjunction with cavalry, down to the 1930s. New tech-
nologies were one thing, but cultures of violence predated these. Beginning
during the zemene mesafint and continuing under Tewodros, Yohannes, and
Menelik, large armies and their attendant cultures would prove difficult if
not impossible to demobilize once brought into being: military cultures
require regular feeding. As Portal observed at the time of Yohannes’ mobili-
zation against the Italians,
[t]he great probability is that, if the present object, i.e. war with the Italians,
were to be take away, these . . . armies, some of them from the country of the
Gallas, from Shoa, and from the extreme outskirts of Abyssinian dependencies,
would refuse to return empty-handed to their own countries, and Abyssinia
would soon be torn by a series of internecine struggles between the different
Chiefs and Kings . . . 124

A typically glum European prognosis, no doubt, but containing more than

an element of truth. Wylde was equally gloomy:
[The nobility’s] only chance of employment is if war breaks out, or they are
sent on an expedition to annex further territory or punish some border
tribe . . . The final settlement of the southern portion of the Abyssinian king-
dom will leave King Menelek face to face with the question of what he will
have to do with his fighting feudal barons and his large army, as he will have
no enemies to conquer . . . The military may settle down and turn their arms
into reaping hooks and ploughshares, but most likely civil war will break
out . . . The soldiery were called into existence by Abyssinia being surrounded
by their Mahomedan enemies, and little by little they increased and multi-
plied till they have got out of all proportion to the wants of a peaceful coun-
try. To keep these soldiers quiet they either have to be paid or allowed to
loot . . . Paying all of them . . . is out of the question . . . so looting has to be
allowed or expeditions started into country that never belonged to
Abyssinia.125

This was misleadingly stark; the empire would diversify sufficiently to pro-
vide alternative, or at least additional, activities for socio-political elites.
Nonetheless the central point holds true, namely that a dangerously potent
militarism had been created which would not easily be undone. It seems
possible to suggest that the region has hardly demobilized since the early
nineteenth century.
92 v i ole nc e and i mpe ri al i sm

Enormous levies were possible from the mid-1870s onward in the defence
of the region against outside aggression—and indeed such mass mobilization
became the mainstay of Ethiopia’s military success in the last quarter of the
nineteenth century. In many ways the epitome and apex of that success was
against the Italians at Adwa in 1896, examined in the next chapter. Suffice to
observe here that victory at Adwa was as much as anything else the culmina-
tion of a restless militarism which Menelik succeeded in harnessing, if briefly.
The common assertion, increasingly made in European sources as the nine-
teenth century progressed, that every Ethiopian man was born a ‘warrior’
was doubtless racial cliché; but it did reflect the reality of the well-armed and
readily mobilized community as the basic building-block of the habesha pol-
ity. At the same time, in the half-century or so between Tewodros’ accession
and Menelik’s apogee c.1900, the perfect fusion of political and military
establishments had been achieved; the key offices of state were held by sol-
diers, and no separation existed—although some would open up in the early
decades of the twentieth century—between political and military authori-
ty.126 Menelik’s Ethiopia was the product of armed force and a culture of
violence a century or more in the making; this was a century in which war-
fare had been honed as an effective tool for bringing about political change.
Further, the polity was defined by its conquered and volatile frontiers—po-
litical, ethnic, faith-based—whose evolution can only be appreciated over la
longue durée.The modern state was now precariously balanced, stilt-like, on a
series of frontier zones both internal and external, zones to which Western
boundary-making legality was now being applied but which were in differ-
ent ways of considerable antiquity. The last word, perhaps, should go to
Bulatovich. ‘[T]he history of Ethiopia’, he wrote, ‘is one of continual war
with both internal and external enemies. The basis of imperial power can
only be actual military strength, and on the army as on a foundation, has
been built all the rest of the edifice of the Ethiopian Empire’.127 The milita-
rization of political culture was the experience of many other states and
societies in Africa in the nineteenth century; but whereas it was generally
arrested, temporarily at least, by the onset of colonial rule, this would not be
the case in Ethiopia. The implications of this massive military complex and
the resultant primacy of armed force, for Ethiopia and for the region as a
whole, would become clear in the decades that followed.
 PART
III
Colonialisms,
Old and New
This page intentionally left blank
 5
Demarcating Identity
The European Colonial Experience,
c.1890–c.1950

A great deal of transformative power is attributed to colonial modernity

across Africa, and so it is in our region which had the unique experi-
ence of both European and African imperialisms in operation over a remark-
ably large area. This chapter is concerned with the territories of European
administration. It will be suggested here that, of course, European colonial
rule was important in certain respects, as particular identities became rather
more sharply demarcated. And yet the tendency of historians for a genera-
tion or more has been to contribute to the concept—whether inadvertently
or explicitly—of a transformative colonial modernity which, in north-east
Africa at least, amounts to a distortion of historical realities. I argue here that
in many respects European colonialism was nowhere near as significant as
has been suggested, and that a range of conflicts and core aspects of political
culture far pre-dated the European administrative presence, which merely
provided new elements in struggles of antiquity. It might be suggested,
indeed, that in the case of Eritrea and Ethiopia, the Italians were inadvertent
actors in ancient borderlands; ultimately, in many respects it was the Italians
who were co-opted into a larger, more ancient struggle, while their contri-
bution was the introduction of superficial modernity.

Blood on the tracks: Italians on the Mereb

From the humblest of beginnings—a foothold at Assab at the southern end
of the Red Sea, purchased by the Rubattino Shipping Company in 1869
and taken over by the Italian government in 1882—Italy extended their
96 colon i al i sms, ol d and new

control to the more centrally located Massawa in 1885, and thence crept up
the escarpment and onto the kebessa. It was not a straightforward process: in
1887, a company of Italian soldiers was destroyed at Dogali in the coastal
foothills by Alula’s frontier force, and the Italians were compelled to remain
clustered around Massawa until 1889, when crises in the interior facilitated
their rapid push inland. Menelik, of course, had courted Rome for some
time when still negus of Shoa, with a view to access to modern weaponry,
international recognition, and support for his eventual claim to the
Solomonic inheritance. In the fragile aftermath of Yohannes’ death at
Metemma, Menelik sought Italian support as part of a larger strategy of
rapid consolidation, and in the middle of 1889 signed the Treaty of Wichale
with the Italians.1 Essentially a treaty of friendship and commerce, in the
Amharic version Article 17 stated that the Ethiopians might, if they so chose,
ask for the assistance of the Italian government in conducting foreign relations.
The Italian version, however, expressly stated that the Ethiopian govern-
ment was obliged to go through the Italian government in conducting foreign
relations. In effect, as far as Rome was concerned, Italy had secured a pro-
tectorate over Ethiopia.2 Menelik spent the early 1890s demanding that the
treaty be withdrawn, and preparing for war. Further clarifications in 1890
had established in greater detail the Italo-Ethiopian border of the infant
colony of Eritrea; but the Italians subsequently made incursions into Tigray,
which were themselves in violation of the treaty, and although they with-
drew, the first half of the 1890s witnessed a build-up of both troops and
tensions along the northern frontier. At the same time, Italian forces were in
action (with rather more success) against the Mahdists in the western low-
lands; the latter were defeated at a key engagement at Agordat in late 1893
by an Italian force that was greatly bolstered by local levies.3
Much of this took place against a backdrop of crisis on the Ethiopian
side.4 In the middle of 1888, the epidemic of rinderpest, carried by infected
cattle through Massawa, swept into the highlands, devastating the Hamasien
region and then moving into Tigray before proceeding south into Lasta,
Gojjam, and Shoa, and destroying the livelihoods of large numbers of Somali.
Livestock mortality was huge across a wide region, with herds exterminated
with extraordinary speed; it is no coincidence that the Italians were able to
move into the kebessa with some ease through 1889, establishing a base at
Asmara in the midst of a province devastated and weakened by the epi-
demic. In the same period, 1888–9 witnessed a severe and prolonged drought
across the Sahel belt, which crippled agriculture across north-east Africa
 de marcat i ng i de nt i ty 97

and compounded the effects of rinderpest; it led to a severe shortage of

meat and grain, and in their weakened state people succumbed to a range
of other epidemics in the early 1890s, including smallpox, dysentery, typhus,
and cholera. It took several years for Menelik’s fragile polity to approach
recovery, and it was only through the creation of extensive granaries and the
postponement of direct, full-scale military engagements with the Italians
that Menelik was able to manoeuvre himself into a position from which he
could address the growing threat in the northern borderland. Once again,
climatic instability had a profound influence on the nature of military
affairs—as it had earlier in the nineteenth century, and as it would in the
later twentieth.
In September 1895, however, Menelik was able to declare a general
mobilization, seeking to harness the restless militarism which had built up
across the Ethiopian region since the 1860s, and indeed throughout the
nineteenth century. From this perspective, the great clash between the
Italians and the Ethiopians represented, among other things, a critical
release of pressure for Menelik. In Eritrea, General Baratieri—commanding
some 35,000 men, mostly Eritrean ascari—made the fatal error of regarding
Menelik as a mere savage at the head of an unruly rabble that would surely
be unable to withstand a modern European army. Italian intelligence and
planning were deeply flawed. In the second half of the nineteenth century,
habesha armies had proven themselves remarkably adept at dealing with
external threats, usually through mass levies vastly outnumbering oppo-
nents.5 A military confederation of upwards of 100,000 men assembled in
central and eastern Tigray in late 1895, skilled in the use of both terrain and
reinforcements, and better armed than most contemporary African armies,
with perhaps four in five men possessing a firearm of one form or anoth-
er.6 The late nineteenth century, indeed, had heralded the coming of age of
the firearm in Ethiopia as an instrument of real political and military pow-
er.7 Yet the outcome could still have been very different—indeed, there
might have been no battle at all. In January 1896, one important advantage
lay with the Italians, in that they could afford to wait, whereas Menelik’s
army was on the point of starvation. With food supplies in Tigray dwin-
dling, sectors of the army becoming restless, and tensions developing
between troops and locals—tensions which would endure long after the
battle—Menelik was considering disbanding by mid-February. He was
saved, however, by the intervention of Italian domestic politics. With the
government pressuring Baratieri to resolve the standoff and conjure up
98 colon i al i sms, ol d and new

some much needed national glory, the Italians decided to march into Tigray,
which they did on 1 March. The details of the ensuing battle, fought in the
hills around Adwa, are well-known:8 the Italians’ comparatively tiny force
of 14,500 men was outflanked and outgunned by a huge Ethiopian force,
and was quickly in disarray. By noon, when the retreat was sounded, 4000
Italians and 2000 Eritreans had been killed, as against the much smaller
proportion of between 4000 and 7000 Ethiopians, and the Italian imperial
dream in north-east Africa—at least as it had first been conceived—was
dead. Rome sued for peace, offering the abrogation of the Treaty of Wichale,
and unconditional recognition of Ethiopian sovereignty. A new treaty in
October 1896 recognized the Eritrean frontier—although Eritrean prison-
ers of war were mutilated in the standard fashion on Menelik’s orders,
effectively as ‘traitors’ and rebels. Menelik’s victory at Adwa—not to men-
tion his own campaigns of territorial expansion in the south and east—
confirmed him and his state as a fully sovereign actor in the region’s affairs,
and indeed a bona fide participant in its partition. Adwa itself would become
a reference point of remarkable power.
Much has been made of Menelik’s decision to ‘sell off ’ Ethiopia’s sup-
posed northernmost province—i.e. Eritrea—to the Italians at this time. It
would become a shibboleth of much twentieth-century scholarship that
somehow ‘Eritrea’, in some curiously timeless form, belonged immutably to
‘Ethiopia’, in a similarly timeless, imagined manifestation, and that Menelik
had for one reason or another ‘given it away’. At best, it was an ugly distor-
tion of a much more complex set of realities; at worst, it was simply an
untruth.9 Certainly, Menelik was nowhere near strong enough, either in
economic or in purely military terms, to contemplate some kind of grand
campaign to drive the Italians into the Red Sea. Menelik was confronted
with serious unrest in Tigray, and had rather larger territorial issues to deal
with in the south and east. Tigrayan ill-feeling toward the restored Shoan
political establishment rendered Menelik vulnerable and diffident in the
north; an assault into the kebessa and beyond would have exposed his aching
flanks to all manner of hostility (never mind a presumably regrouped Italian
opponent).As it was, the army was subject to attacks by Tigrayan and Oromo
communities in the period after Adwa. Groups of Oromo and Tigrayan
shifta and armed peasants harassed Menelik’s troops as they wound their way
home along southward roads. More broadly, as an Amhara, his view of the
region was rather different from that of his Tigrayan predecessor. He had less
of an interest in the north, for the riches of the south now absorbed him, as
 de marcat i ng i de nt i ty 99

well as access to Djibouti, which port he had always used. Above all, the
long frontier that eventually would be drawn up between the new imperial
Ethiopia and Italian Eritrea was, in the main, a recognition and formal
delimitation, whether unwitting or otherwise, of a pre-existing set of bor-
derlands.The Danakil depression, and in particular the Mereb valley and the
Gash-Takkaze lowland zone, had all constituted contested borderlands since
long before Italy was even unified as a nation-state. The fact that several
historical fault lines were now contained within a territory named ‘Eritrea’
scarcely altered the fact of their antiquity, nor indeed the fact that they
remained active.

Raising the first born: Italian Eritrea, to 1941

It has become axiomatic that Italian colonial rule was crucial to the creation
of ‘modern Eritrea’—axiomatic, that is, among scholars of Eritrea, and abso-
lutely fundamental to the nationalist interpretation of the region’s history.The
Italian creation of Eritrea, it is argued, set the territory on its own distinctive
path, contributed significantly to a discrete sense of identity—separate, cru-
cially, from that of Ethiopia—and fundamentally laid the foundations for the
modern nation.10 Italian colonialism is the sine qua non of Eritrean national-
ism. No doubt; but it is important to note that much of the literature pro-
duced on Eritrea from the late 1980s and through the 1990s was concerned,
explicitly or otherwise, to explain (and quite often laud) the nationalist cause,
and so it was history written backwards, characterized by an earnest search for
the first shoots of the nationalist liberation movement, broadly defined.
However, if we remove the desire to account for an explicitly modern national-
ism from the equation, we might have a rather different view of the region’s
development over la longue durée.This interpretation suggests that the nation-
alist thesis in fact prevents us from seeing what is most interesting about the
history of the Eritrean region and its adjacent territories, and argues for the
‘de-modernization’ of the Eritrean liberation struggle, which is—for one rea-
son or another—viewed exclusively in modernist terms, both by students of
it and of course by the protagonists themselves. In fact, the nationalist move-
ment grew out of a conjunction of dynamics, and was the product of a series
of conflicts and tensions, which had their roots in the nineteenth century and
arguably earlier, not in any transformative European colonial experience.This
10 0 colon i al i sms, ol d and new

is not to dismiss the very real impacts resulting from Italian colonialism, but
to place them in context.
The Tigrinya of central Eritrea, over the course of several centuries, had
been gradually cut adrift from the Tigray-Amhara political and cultural
system—and the inheritance it implied—further south. They increasingly
inhabited a fault line, a frontier which was international as well as ‘internal’,
according to the Kopytoff thesis. Like Tigrayans themselves, they had a his-
tory of producing mavericks, defying orthodoxy, and struggling against
southern domination of a Christian tradition supposedly northern in prov-
enance; over the long term, they had come to see themselves as different, the
product of the volatile northern borderland. Italian colonial rule under-
pinned and indeed exacerbated that sense of difference, and opened up new
schisms in terms of group perception. Of course, throughout Eritrea’s story,
‘Abyssinia’ lurks in the background of the nationalist narrative, like an
elderly and sinister relative with suspect designs. But the actions of the
Italians through the 1890s were in pursuit of one major goal—namely, the
creation of an east African empire that included Ethiopia—and the aim was
clearly that Eritrea would only form part of a much larger imperial terri-
tory. Eritrea itself was born of the failure to carry this grand scheme to frui-
tion, created essentially out of military and political failure; and the
subsequent relationship with the first-born was as that of a parent hoping
for a larger family but thwarted by a certain infertility. With the aggressive
advances of the Italians repulsed at Adwa—a notably unsuccessful attempt
at penetration—Eritrea was briefly assigned the role of only child, and after
Somalia and Libya were added, Eritrea was the eldest child in a dysfunc-
tional family, separated from its siblings by considerable time and distance;
the significance of using Eritrean soldiers in campaigns in both Somalia and
Libya is clear enough. Above all, however, Italian failure compounded the
unstable frontier, and served to compound the kebessa’s historic insecurity
and sense of political ambiguity.
Italy had initially conceived of Eritrea as a settler colony: swathes of the
highlands were alienated for the purpose between 1893 and 1895, while the
first experiments in commercial agriculture were attempted in the early
1890s, at Asmara, Mendefera, and Gura.11 But with the rebellion of Bahta
Hagos in 1894—examined below—as well as smaller-scale incidents of
resistance, and the defeat at Adwa, the project was mostly abandoned in
favour of developing Eritrea as a commercial and industrial centre. Two
large farms were established at Asmara in 1899, and from 1901 further
 de marcat i ng i de nt i ty 101

concessions in the cultivation of coffee, cotton, and tobacco were made to

both Eritrean and European farmers.12 Alongside the ongoing refinement
and consolidation of a system of colonial governance, the Italian period
therefore witnessed the development of a light industrial economy, signifi-
cant transport infrastructure, and some measure of commercial agriculture.
Combined, these dynamics brought about socio-economic transformation
across the territory, whose peoples’ relations with one another were thus
fundamentally recast. Hamasien province was arguably the most industrial-
ized zone in north-east Africa by the 1930s.This fuelled urban growth, most
notably in Asmara and Massawa; in Asmara, manufacturing and a vibrant
service sector served the needs of an expanding Italian community.13 In
terms of agriculture, while the kebessa was intended to produce food for
local consumption, the 1920s and 1930s witnessed the creation of land con-
cessions in the western lowlands through which capital-backed farms would
produce for export, for example coffee. The territory was bound together
by an extensive road and rail network, connecting the central areas with the
northern highlands and western lowlands, and with the southern Afar
region. The growth of a wage labour economy absorbed many young men
from across the region;14 it would be the failure to sustain economic growth,
however, that would propel a large number of them back into the ‘bush’ to
become shifta.
The colony was initially under a military administration, but in 1898, fol-
lowing the relocation of the capital from Massawa to Asmara, the appoint-
ment of Martini as Governor marked the shift to a civil administration.15
Between 1897 and 1907, boundaries with Sudan, Ethiopia, and French
Somaliland were delimited, although it was not until mid-1908 that the
question of the Danakil boundary between Ethiopia and the Italian author-
ities was resolved. While Eritrea was divided into commissariati, or districts,
each headed by an Italian bureaucracy, administration was in other ways
conducted through fairly classical indirect rule, with particular chieftaincies
and local aristocracies being co-opted for the purposes of stable governance.
At the same time, the provision of education was extremely limited—no
Eritrean could advance beyond the fourth grade—and thus created some-
thing of an indigenous underclass of unskilled labourers and soldiers, similar
to socio-economic systems in Kenya and across southern Africa. A system of
racial segregation had been in place since the beginning, but it became
rather more rigid and codified with the introduction of Fascist laws in
the early 1930s.16 Nonetheless through mission education and the rare
10 2 colon i al i sms, ol d and new

appearance of opportunity, a small class of Eritrean functionaries did work

their way into the colonial administration as clerks and translators, while at
the same time, in Asmara, an Eritrean lower middle class of entrepreneurs
was also beginning to appear by the 1930s. In many respects these would be
critical in the political developments of the 1940s.17 Otherwise, a common,
and not unattractive, career for tens of thousands of Eritrean men was as
ascari, soldiers in the colonial army: on the eve of the Second World War,
there may have been a remarkable 300,000 Eritrean ascari in the colonial
army. Such troops were critical, indeed, to the Italian imperial mission, and
were deployed in Libya and Somalia down to the early 1930s, in the invasion
of Ethiopia in 1935–6, and against the British in 1940–1. The early 1930s
witnessed a dramatic build up of men and material in Eritrea: the impact
was mainly experienced in Asmara and its environs, long the major destina-
tion for Italian traders and craftsmen, and by the end of the decade the
population of the city was some 120,000, half of them Italian. Preparation
for war against Ethiopia fuelled a short-term economic boom. The huge
influx of settlers and soldiers drove the racial segregation policies noted
above, and a marked increase in the recruitment of ascari.
Eritrea’s total population in the mid-1930s was some 600,000. Around half
of these were Christian farmers in the highlands, and the standard historio-
graphical position has been that they retained ‘a sense of continuing attach-
ment to Ethiopia’ which was ‘kept alive both by Orthodox clergy and by
European missions’. There were also obviously close ties to Tigray, across the
Mereb.18 Meanwhile, in terms of local landownership and authority, there
was both continuity and discontinuity. While some nineteenth-century
chiefly families in the kebessa retained their positions throughout the period
of Italian rule, and indeed beyond, others were displaced by rivals who won
Italian support, and also by ascari who were rewarded with land for service.19
Caulk has argued, at any rate, that there was rather more disruption among
pastoralists, among whom patron–client relations were increasingly severed—
the colonial legal system increasingly favoured serfs who were ‘freed’ from
obligations to their feudal masters in the western lowlands.20 In the 1910s and
1920s, this opened up a new cash crop economy, and perhaps more socio-
economic upheaval as a result—and it might be possible to argue that such
growing commercial competitiveness, as well as inequities in wealth, laid the
foundations for the emergence of new forms of violence (shifta) in the west-
ern lowlands toward the mid-twentieth century. More broadly, this became
a new economic borderland in the course of the twentieth century, one
 de marcat i ng i de nt i ty 103

replete with opportunities for adventuring entrepreneurs but one fraught

with risk as a result of political volatility—as well as being vulnerable to
environmental and climatic instability.
Eritrea continued to struggle economically through most of the era of
Italian colonial rule—and in this sense one must be careful to critique the
later Eritrean nationalist conceptualization of the colony as some kind of
light-manufacturing idyll, its economic prosperity in stark contrast to the
feudal misery of anachronistic Ethiopia. Certainly the territory’s economic
importance in many ways lay in transit trade, with products from Sudan
and Ethiopia imported and sold for re-export; while this was officially up
to 25 per cent of total trade in 1915, for example, it was reckoned in reality
to be much higher.21 Through the early twentieth century, colonial Eritrea
depended on imported grain from northern Ethiopia. Although the rail-
way was extended impressively—toward Keren and Agordat in the course
of the 1920s—hopes that it would spark a dramatic expansion in the cash
economy proved unfounded, as did the hope that thousands of Italian set-
tlers would be attracted to the place. Eritrea continued to rely heavily on
Italian government subsidy;22 the Italian government itself, indeed, often
appeared uncertain as to what to do with Eritrea, or unsure as to what the
‘grand plan’ actually was. The permanent settlers did not arrive, although
again there was a considerable influx of Italians during the build-up to war
with Ethiopia. In many respects Eritrea, as a colonial entity, remained one
in curious limbo, in a more or less permanent state of anticipation of some-
thing which never came. Its modern history has been defined by this state
of expectation.
This has to be understood as the backdrop for one other major phe-
nomenon of the era, and that of a later era, too, namely the migration of
Eritreans southward, especially to Addis Ababa. This has frequently been
understood as the manifestation of some kind of primordial longing, with
the image of kebessa Eritreans drawn moth-like to the great and ancient
beacon of Ethiopia, fluttering hopefully around the Emperor like prodigal
moths. But the fact is that thousands of Eritreans migrated to Ethiopia in
the early decades of the twentieth century for socio-economic reasons.
In 1931, for example, the Italian government—enacting new Fascist legis-
lation—barred Africans from government secondary schools, while from
the following year Swedish and American missionaries were prevented
from entering Eritrea.23 Through the 1930s, therefore, Eritreans increas-
ingly had to move to Ethiopia in order to continue their education, or
104 colon i al i sms, ol d and new

achieve any meaningful level of education. The migration intensified dur-

ing the economic downturn in the north during the 1940s and 1950s—al-
though of course paradoxically there was also a brisk northward, partially
seasonal migration of Tigrayans into Eritrea in search of casual work. But
the important point here is that Eritrean highlanders moving to Addis
Ababa did not signify some great wave of unionists voting, as it were, with
their feet: it was a question of economic opportunity, even if many
Eritreans did indeed become more or less absorbed into Amhara society
and culture. Moreover, these demographic shifts opened up new frontiers,
uneasy borderlands, within the urban environment itself—both in Addis
Ababa and in Asmara.
Elsewhere, boundaries and frontiers were more remote from the concen-
trated urban centres, but no less important in shaping the lives of communi-
ties and in demonstrating the weakness of the colonial state—for these were
frontier communities which in many ways represented the unstable founda-
tions on which colonial systems (and later nation-states) rested. Nowhere
was this more in evidence than the northern Eritrean border with Anglo-
Egyptian Sudan, where the Rashaida frequently skipped across the frontier
to escape Italian taxes, or defied attempted Anglo-Italian policies of separa-
tion by crisscrossing the border to trade and visit kin with impunity. In that
sense colonial borders were used when it was convenient—tax avoidance,
notably, or using legal boundaries as springboards for raiding patterns which
actually pre-dated them—and ignored likewise, again when it suited local
communities to do so.24 The borders between Eritrea and Sudan in the
1920s and 1930s were effectively uncontrollable, or at least represented zones
where the power of the colonial state was much diminished, where the hum
of colonial authority was so faint as to be almost inaudible. In 1922, for
example, the British encountered fierce Rashaida opposition to their
attempts to register arms which were increasingly being used in local raids,
and the following year Sudanese Hadendowa and Beni Amer drove their
cattle into Eritrea in order to avoid having it registered by the British, who
ultimately had to abandon the project and settle for the payment of token
tribute. Moreover, as Young has pointed out, the Rashaida were able to sim-
ply take their camels away from watering holes for long periods at a time,
thus avoiding the police units who necessarily gravitated toward them.25
The British also found themselves unable to prevent the thriving cross-
border grain trade in the late 1930s and early 1940s, with Sudanese mer-
chants—including some Rashaida, in all probability—taking advantage of
 de marcat i ng i de nt i ty 105

Italian demand in the war years especially. Police units were easily avoided.26
Again, this was the archetypal trading-and-raiding zone, an economic bor-
derland which was unstable, but potentially lucrative.
Throughout the 1890s, the Italians were faced with resistance, particularly
against the early expropriation of land on the plateau, and from chiefs—‘elite
resistance’, as Tekeste Negash has referred to it—who discerned a loss of
status and privilege in the European occupation. Dejazmach Aberra, a chief
from Hamasien, clashed with the Italians in 1892, went into hiding, and ulti-
mately succeeded in fleeing to Ethiopia, where Menelik famously uttered,
‘Rather than a thousand Amhara, a single Aberra’. He even took part in the
battle of Adwa, and eventually lived out his days in Shoa.27 Bahta Hagos,
based primarily in Akele Guzay, had worked with the Italians since the late
1880s—indeed by all accounts was considered the model ‘collaborator’, hav-
ing also converted to Roman Catholicism—but in the course of the early
1890s he was increasingly restive, and communicated his fears regarding
Italian expansion to Menelik. In late 1894, he rebelled, leading some 1600
men against the Italian district headquarters at Segeneiti and promising to
liberate the people from Italian oppression. He was defeated after only three
days, by Italian reinforcements which attacked him in his rear after he had
abandoned Segeneiti and moved against a small Italian fort at Halay. But the
uprising arguably ushered in a period of chronic instability for the Italians,
culminating in the defeat at Adwa.28 In fact, in many respects, the story of
Bahta Hagos—formerly a shifta, from the mid-1870s through much of the
1880s, following a family dispute with an uncle of Emperor Yohannes—
exemplifies the vagaries of frontier life in the Mereb zone.29 Bahta’s son
escaped into Tigray after the defeat in 1894, and ultimately joined forces with
the shifta Muhammad Nuri in the area of Senafe, and fought a low-level
campaign against the Italians,30 mostly hit-and-run attacks on Italian garri-
sons, plantations, and isolated settlements. There was a shift in Italian policy,
as the naked expropriation of land was moderated and some degree of co-
habitation with the wary ‘natives’ was achieved in the years that followed,
certainly by the time of the First World War.31 Chiefly titles on the plateau
were protected, particularly in terms of land ownership, while a lowland
aristocracy was maintained.32 Arguably the Italians also feared the growth of
Islam in Eritrea, and were keen to nullify any potential threat from that
direction; a British report from 1920 suggested that ‘Italian authorities are
agreed that Mohammedanism is making rapid progress in Eritrea’.33 In some
areas, perhaps, Islam offered an appealing form of quiet defiance.
10 6 colon i al i sms, ol d and new

According to Tekeste Negash, while the peoples of the western lowlands

very broadly welcomed Italian overrule—which pacified the area and pro-
vided them with protection from the habesha raiding parties that periodi-
cally stole cattle and seized slaves, as they had throughout the nineteenth
century—the Eritrean highlanders eschewed direct military confrontation
with colonial authority, indeed often actively collaborated with it, but also in
the 1920s and 1930s developed irredentist resistance networks.34 This was not
Eritrean nationalism per se, and therefore not even organic Eritrean protest,
but was motivated by a desire to rejoin the Ethiopian motherland, actively
encouraged by Ethiopia itself. Both highland and lowland models seem
somewhat simplistic.The real capacity of the Italian administration to ‘pacify’
the west, especially the borders with both Ethiopia and Sudan, must be ques-
tioned; cross-border violence continued, largely immune to the periodic
patrols by both British and Italian police forces, even if the overt threat from
Ethiopia had been somewhat nullified.With regard to the highlands,Tekeste’s
self-appointed mission to denigrate any kind of ‘nationalist’ achievement, and
simultaneously to depict any kind of Eritrean political consciousness as
essentially Ethiopian is not difficult to spot. For sure, Eritreans were in some
ways caught in a standoff between Rome and Addis Ababa, to the point at
which, from the mid-1920s, the Italians had sought to break the links once
and for all that existed between the kebessa and the Ethiopian population.35
But it is wholly understandable that the authorities in Asmara, in the years
following humiliation at Adwa, would have been careful to prevent a wider
African resistance developing from across the border, into which Eritrean
protest might tap. It is also wholly credible that Eritreans would indeed have
tapped into it, not because of any grand design to ‘rejoin’ Ethiopia necessar-
ily—although clearly some did aspire to this, witness the various strands of
unionism which came to the fore in the 1940s—but because Ethiopia repre-
sented the most natural reservoir and facilitator of resistance in the area. A
trans-Mereb resistance network was the natural recourse of Eritreans who
would defy Italian rule. For their part, it is clearly the case that many
Ethiopians—beginning with the Amhara political establishment of Haile
Selassie—sought the incorporation of Eritrea and the acquisition of a coast-
line which this would entail, and that the encouragement of irredentism
inside Eritrea was part of a strategy aimed at the achievement of this goal. It
was also, of course, an anti-Italian strategy. Just as unionism was anti-colonial
resistance for some Eritreans, for many Ethiopians the encouragement of
irredentism and unionism was a matter of regional security.
 de marcat i ng i de nt i ty 107

Along the Mereb borderlands, existing patterns of shifta activity became

instrumental in resisting Italian administration from the 1900s onward,
although to describe this as explicitly anti-colonial resistance would be mis-
leading. Often, armed action against the Italians was the by-product of local
disputes and even organized criminal activity—banditry, in other words.
As Alemseged Tesfai has made clear, shiftanet was common throughout the
Italian period, but this was mostly motivated by local personal grievance,
and was not driven by any larger political programme.36 Again, typical tac-
tics included limited raids on neighbours’ villages and livestock, and the
ambushing of traffic on the highways. Often such activities were driven by
personal vendetta. Nonetheless it is clear—the example of Bahta Hagos’ son
illustrates the point—that shifta could be co-opted into larger causes, wit-
ness the emergence of the Ethiopian Patriots in the late 1930s during the
Italian occupation, examined in the next chapter. Shiftanet represented a
long-established culture of defiance against established authority, in the
nineteenth as in the early twentieth century, and was the product of the
volatile, fertile borderland. In the 1940s and 1950s, shifta would indeed be
increasingly co-opted into causes being championed beyond their immedi-
ate vicinities, first in terms of unionist militancy, and later in the context of
the nascent nationalist struggle. The roots of the modern phenomenon lie
in the late nineteenth and early twentieth centuries, although disgruntled
and disenchanted armed men had long been utilized for larger political
causes across Ethiopia, Eritrea, and Somalia; the borderland had long been
pressed into service in order to effect change at the centre. In a sense, con-
tested frontier zones represented reservoirs of disengagement and defiance
which might be harnessed—which in essence is the story of our region’s
twentieth century, or at least a very large part of it.
In the final analysis, Italian rule in Eritrea was demolished not by mass
resistance or revolt, but by a British-led Allied invasion force—although they
were aided by both Ethiopian ‘Patriots’ and Eritrean deserters from the Italian
colonial army. But the Allied ‘liberation’ in fact took the familiar form of
borderland advancing on metropolitan centre, albeit mechanized. In the
middle of 1940, a few weeks after they entered the war, the Italians had made
advances into the frontiers of neighbouring British territories, capturing
Kassala in Sudan and Moyale in Kenya. But by early 1941, they were obliged
to repair to more secure positions: in Eritrea, they pulled back toward Barentu
and Agordat, and were able to construct a fearsome defensive position around
Keren.37 There, the fortified scene of much nineteenth-century contest, was
10 8 colon i al i sms, ol d and new

fought one of the bitterest battles of the African campaign, in February and
March 1941. Only with the Italian defeat at Keren was the way open to
Asmara and, ultimately, Ethiopia. The British, meanwhile, sought to exploit
the flagging loyalty and confidence of the Eritrean ascari, who had begun to
desert, although numbers are difficult to establish.38 The British also launched
a concerted propaganda campaign, dropping leaflets on the colonial troops
containing promises of what would happen if they deserted and helped crush
the Fascists:
Eritrean soldiers, listen!
Desert from the Italians and join us. […]
We know the reason you would not fight against us was that you did not wish
to be ruled by the Italians; you will receive your full reward.
You people who wish to live under the flag of His Imperial Majesty, Haile
Selassie I, and to have your own flag, we give you our word that you shall be
allowed to choose what government you desire.

Haile Selassie had his own, rather less equivocal statement prepared, likewise
deposited from the air by the RAF, in which he appeals to both Eritreans and
Somali:
Eritrean people and people of the Benadir! You were separated from your
Mother, Ethiopia, and were put under the yoke of the enemy…
But now the day has come when you will be saved from all this ignominy and
hardship.
I have come to restore the independence of my country, including Eritrea
and the Benadir, whose people will henceforth dwell under the shade of the
Ethiopian flag ...39

The British, famously, did not liberate Eritrea for Eritreans, as the popu-
lar anecdote goes,40 but as part of a much larger regional effort to secure
East Africa, the Nile valley, and the Red Sea from Axis control. Nonetheless,
their arrival did usher in a turbulent new era in the territory’s history, at
least in part because they showed themselves unwilling or unable to live up
to the promises that had fluttered from the sky.

Somalia divided
The Somali had close linguistic and cultural links to a range of groups across
the eastern and southern zones of our region, notably the Afar on the
 de marcat i ng i de nt i ty 10 9

eastern side of the Rift Valley, and the Oromo, including the Boran.41 In part
these links were enhanced by the common profession of Sunni Islam; and
so by the late nineteenth century, there was a potential network of action
across a wide area, without the need for overarching centralized leadership.
As far as resistance to infidel colonial rule is concerned, Islam provided
coherence and a relative unity of purpose—as it had, sporadically, in con-
flicts with the Christian highlanders to the north and west in earlier periods.
However, as we shall see, Somali unity was continually undermined by clan-
based and territorial factionalism, which often took precedence over the
authority of common law—heer—or Islamic law, shari’a, both episodically
evoked to resolve conflict.
In many respects the renewed aggression on the part of the Solomonic
state in the late nineteenth century represented a renewal of the titanic
struggle between Christian habesha and Muslim Somali which had been
played out between the early fifteenth and mid-sixteenth centuries.Yet the
struggle was greatly complicated by the presence of foreign powers, nib-
bling at the edges of the Somali region. Zanzibari suzerainty over the south-
ern littoral, including Mogadishu, was rather less threatening than the
appearance of an Egyptian presence in the north in the 1870s. Khedive Ismail,
with British encouragement, took control of the Berbera coast, and went as
far as sacking Harar, but with the evacuation of the Egyptians in 1885 came
new and more dangerous developments. Menelik’s Shoan forces defeated
Harar in 1887, while the British established their Somaliland Protectorate
with a view, primarily, to securing food supplies for their base at Aden. The
British, meanwhile, also assisted the Italians in the attempt to fulfil their
ambitions further south: in the course of the 1890s, the Italians established a
presence along the Benaadir coast between Kismayu and Mogadishu, and
although they initially posed as ‘protectors’ of the Somali against increas-
ingly aggressive Ethiopian campaigns, they could do little to prevent these
in reality.42 Indeed, less than a year after Adwa, Italy suffered another disas-
trous defeat at Lafoole, just inland from Mogadishu, in which a group of
Somali clansmen crushed a small force comprising mostly Arab ascari.43 It
set the Italian project back several years, but this did proceed in time—and
when it did, massive Ethiopian territorial expansion was also taking place
from the north and west, into the Ogaden and the Haud plateau. Both the
Ogaden and the Haud were of fundamental importance to the seasonal
migrations of Somali pastoralists, and in the era of high imperialism they
became the focus of rising tensions. In the years before Adwa, the British
110 colon i al i sms, ol d and new

were prepared to concede the Ogaden and Harar regions to the Italians;44
but with Ethiopia soon establishing dominance in the area, it was necessary
to deal with Menelik instead, with whom Britain negotiated seasonal access
to the Haud for Somali pastoralists. However, they formally ceded the area
to Ethiopia in 1897 in a treaty which only came to light for most Somalis
in 1954, when it was fully implemented. Not unlike the Eritrean western
lowlands, this also became a volatile economic frontier zone, across which a
commerce in livestock thrived, and into which adjacent states attempted to
extend political as well as economic control. The meeting of local and
external dynamics made for an unstable environment; it was a land of
opportunity, but a dangerous one with it.
As Menelik’s armies thrust into, and built military garrisons across, the
lands of the Somali, Oromo, and Afar, the French asserted themselves over
their little stretch of Somali and Afar territory, in Djibouti; and the British
were advancing from yet another direction—the south, from their East
African Protectorate (Kenya). As they reached the hot grasslands east of
Lake Turkana, they encountered both Oromo (Boran) and Somali, as well as
a vaguely defined Ethiopian jurisdiction, in the late 1890s and early 1900s.
This was the zone which would come in time to be known as the Northern
Frontier District of Kenya, abutting the southernmost reaches of the Amhara
empire in the hills north of Turkana.45 It would become the archetypal
colonial backwater, of little value and even less interest to administrators in
Nairobi, and yet in time it would become one of the key zones of transition
in the development of a putative pan-Somali identity, and interstate rela-
tions between Kenya, Ethiopia, and Somalia. Thus, between the 1880s and
the First World War, were the far-flung Somali divided into no less than five
imperial territories. From Djibouti, through British and Italian Somaliland,
eastern Ethiopia, and into northern Kenya, was an enormous, crescent-
shaped, cultural, and ethnic frontier—arguably one of the largest on the
continent.This vast borderland would become a major zone of violent con-
flict in the course of the twentieth century, drawing neighbouring metropoles
into it and profoundly influencing regional ideologies, as Somali nationalists
came to embrace the notion of a pan-Somali identity, fired by the image of
the nation split asunder. Somali irredentism and the struggle for union
would become one of the rallying calls of the twentieth century history of
the region. Nevertheless, visions of Somali unity were habitually under-
mined by the fractiousness of the clan system, which was to prove rather
more robust than the nationalist project; there was often mistrust and not
 de marcat i ng i de nt i ty 111

infrequently violent clashes, for example, between the Ogadeni groups and
the Isaq in what would become British Somaliland. Clan rivalries—mani-
fest in clashes between mobile bands of armed men, engaged in conflict
over access to land and water—would become sharper with the rise of the
Somali nationalist project.46
Much earlier, meanwhile, Somali violence had erupted on a more local-
ized scale—though still spreading across a considerable area, on a scale similar
to the kind of resistance put up by the Sanusiyya against the Italians in Libya.
The Somali region had been devastated by the economic catastrophes of the
late nineteenth century, with rinderpest decimating herds and drought piling
misery on misery. But while these crises rendered many communities inca-
pable of sustained resistance, desperation pushed many Somali toward violent
struggle against the forces amassing on their horizons. The key figure was
Muhammad Abdille Hassan, born in 1864 in the area which would become
British Somaliland, and by the 1890s a leading religious scholar in the region
with several thousand followers.47 He spent much of the 1890s warning
about the perils of expanding European influence and in 1899 he declared a
jihad against the infidel, proclaiming himself to be the Mahdi, the chosen
one, who would unify all Muslims and expel the Europeans. In early 1900,
his army of ‘dervishes’ launched attacks across British Somaliland and into
the Ogaden; between 1901 and 1904, the British organized four major cam-
paigns against him.48 Most of his support came from the Ogaden and
Dhulbahante clans. Although an uneasy armistice was maintained between
1904 and 1908, periodic raids were carried out by Muhammad Abdille’s men
on British garrisons, until in September 1908 he launched a new and sweep-
ing offensive across both British and Italian territories, and against the
Ethiopians in the Ogaden. For a time, the insurgency represented one of the
most successful anti-colonial rebellions of its time, forcing the British to
evacuate the interior in late 1909 and retreat to their coastal bases; only the
Anglo-Italian strategy of developing local resistance to the dervish forces and
exploiting divisions in the Somali community—tried and tested elsewhere
in the colonial world—brought the uprising under control. A pact between
the Warsangeli and Majeerteen Somali of the Darod clan—in the British and
Italian territories respectively—led to joint military operations against
Muhammad Abdille, who was at length driven out of Italian Somaliland;
between 1913 and 1920 he remained active in the British territory.
Muhammad Abdille’s forces kept the British occupied throughout the
First World War, and when rumours circulated concerning a possible alliance
112 colon i al i sms, ol d and new

between him and the new emperor in Addis Ababa, Lij Iyasu, there was
genuine concern in British military and intelligence circles at the prospect
of a region-wide insurgency. Iyasu, as we see in the next chapter, was of
Muslim ancestry and had clear Islamic sympathies, even if he had not yet
actually converted; his Islamic dalliances, the success of the Somali insur-
gency, and proclamations coming out of Istanbul about the need for a holy
war against the British infidel all fuelled growing fears of a mass uprising of
east African Muslims.49 Such fears would gradually dissipate: Iyasu was sum-
marily removed from office in 1916, although he remained at large in the
Ogaden for several years, and the call to universal jihad was largely ignored.
But the potential for religious violence remained real, anticipating events in
the early twenty-first century. The British threw all available resources at
Muhammad Abdille in late 1919 and early 1920, and the ageing leader was
now decisively defeated—although importantly, he again evaded capture,
fleeing into the Ogaden. The British pursued him, but never caught him,
and he died in either late 1920 or early 1921. He was a remarkable personal-
ity, and even if his influence and standing had waned somewhat among
Somali in his last years, he became something of a nationalist figurehead in
the years after his death.50 Never mind that certain aspects of the story were
decidedly ambiguous in terms of his role as nationalist hero—many Somali
had indeed remained ‘loyal’ to the British, the Italians, and the Amhara and
had fought against him—he held a natural attraction for the next genera-
tion of Somali political activists, and in particular the emerging pan-Somali
movement of the 1940s, which deployed his memory to good effect in the
ill-fated struggle for Somali unity.
In truth, actual Italian administration in Somalia was limited; swathes of the
territory remained beyond colonial control, periodic armed patrols in rural
areas notwithstanding, down to the 1930s. Likewise, Ethiopia maintained little
more than a light presence in the eastern part of its empire in the 1930s, and
the Ogaden was frontier country through which habesha caravans and patrols
travelled nervously.51 When they swept up from Kismayu, through Mogadishu,
and into Ethiopia across its southern frontier in early 1941, the British found
a dangerously restless, and heavily armed, society: unrest was endemic in
Mogadishu itself—the British were confronted by something of a crime wave
in the early 1940s—while in the deeper interior and across the Ogaden ‘bor-
der’ clan feuds erupted in 1942–3, most clearly visible in the form of livestock
raids by armed militias, requiring the rushing into active service of a hastily-
assembled Somali gendarmerie. The British attempted to disarm the ‘lawless
 de marcat i ng i de nt i ty 113

tribes’ of the interior, with limited success.52 The situation was brittle and
tense at best, and beyond the control of the authorities at worst: as one official
wrote despairingly,
[i]n the areas within immediate reach of authority some slight abatement of
lawlessness is discernible, but elsewhere the country is in a state of wild and
uproarious disorder in which it is bound to remain till adequate numbers of
political officers and gendarmerie can arrive on the scene.53

It was against this turbulent background that Somali nationalism emerged

in its ‘modern’ form, initially through the Somali Youth Club, established in
Mogadishu in 1943. As it expanded its operations through the mid-1940s—
becoming the Somali Youth League (SYL) in the process—the movement
became the major vehicle for Somali nationalists, who became increasingly
fixated with the idea of a ‘greater Somali’ unity.The SYL’s central message was
the unification of all Somali within a single nation, including (perhaps even
especially) their brothers in the Ogaden and what were now known as the
Reserved Areas, the latter comprising the Haud and the key market town of
Jijiga, administered discretely by the British from the early 1940s.54 Doubtless
prompted by international discussions over the future of the former Italian
territories, and in particular over the status of the Ogaden and the Reserved
Areas and Ethiopia’s increasingly aggressive claims over those areas, the SYL
became ever more active, and the eastern Horn once more became a frontier
of political contest which would define the future of the wider region.
In the end, however, the goal of Somali ‘reunification’—in any case
something of a misnomer—was unrealized, or at least only partially
achieved. While the SYL swept the board in elections held in Italian
Somaliland in 1959, in the British territory in 1960 an alliance between
the Somali National League—whose stated aim had also been Somali
unity—and the United Somali Party dominated the new parliament.55
Within days of both territories becoming independent, they united to
become Somalia, but large numbers of Somali—many, though by no
means all, of irredentist persuasion—remained ‘stranded’ in northern
Kenya, and in eastern Ethiopia. Once again, international machinations
meant that Ethiopian imperial ambitions in the Somali lands of the
Ogaden were fulfilled; but the force of nationalism among sovereign
Somalis, combined with that of irredentism in the Ogaden, would reignite
old tensions and conflicts between the habesha state stretching into the
eastern plains, and the Somali state beyond.
114 colon i al i sms, ol d and new

Choices: Eritrea, 1941–52

The clearest chance of acquiring Eritrea, as well as the Somali Ogaden, and
thus completing the process of Ethiopianization of the region begun in the
1880s, arrived in the 1940s. It was a crucial decade for Eritrea, for many
different reasons, and also for Ethiopia. Many opinions were expressed, and
arguments made, and identities manifest; there were bitter machinations on
both sides of the contested border. ‘Nationalism’—or perhaps more appro-
priately ‘patriotism’—meant different things to different people, and in
many respects the divisions which opened up in terms of regionalism, eth-
nicity, and faith remain entrenched. Crucially, the question of Eritrea’s status
and future was one of the first major African issues to be handled by the
post-war international community at large and the UN in particular, and
the strategy adopted failed: this too was critical, and would not be forgotten
in Eritrea. Meanwhile, the political dynamics of the decade were played out
against the background of escalating violence in Eritrea and Tigray. Shifta
activity, a key element in central and northern Ethiopian political discourse
throughout the nineteenth and early twentieth centuries, increased dra-
matically in the 1940s and 1950s, and destabilized the Mereb zone; the
insurgency was a serious concern to the outgoing British administration by
the beginning of the 1950s, and many of the insurgents would later contrib-
ute their manpower to movements of armed ‘liberation’ which would trans-
form the politics of the region. In the western lowlands, too, a resurgence of
borderland conflict would later connect with the violence of the wider
region, as it had in the nineteenth century.
In contrast to the Italian era, according to the popular Tigrinya saying,
during which one could ‘eat but not think’, during the British period it was
possible to ‘think but not eat’—alluding to the relative political freedom but
economic recession associated with the British Military Administration
(BMA).56 The general view is that while Eritrea had become materially
more integrated by the beginning of the 1940s, it lacked a sense of common
identity; but during the 1940s, the BMA facilitated the emergence of a new
Eritrean political culture, with Tigrinya and Arabic coming to the fore as
literary languages, and several political parties contesting Eritrea’s future;
these were frequently led by members of an emergent urban middle class
who had served as functionaries of the Italian and now the British bureauc-
racies. The BMA promoted a vibrant press, facilitating in turn political
 de marcat i ng i de nt i ty 115

consciousness, at least in the urban centres. Labour unions flourished in

Asmara and Massawa. Primary and secondary education were introduced, in
contrast to the restrictions of the Italian period.57 Self-conscious national-
isms emerged, but there were deep cultural, ethnic, and religious divides
between these, and fiercely contested visions of the Eritrean future—con-
tests which frequently exploded into open violence. The escalating insur-
gency, as well as the failure of Eritrean leaders to unite, contributed in no
small way to the decision of the UN to federate the territory with Ethiopia,
which was implemented in 1952. Nonetheless, an optimistic interpretation
would be that by the early 1950s, a liberal, democratic culture had begun to
take shape in Eritrea—only to be crushed within a decade by Ethiopian, or
more specifically Amhara, autocracy. Much of this political activity took
place against a background of severe economic problems. The short-term
economic boom experienced during the late 1930s and early 1940s rapidly
dissipated from the middle of the decade, and serious decline, particularly in
the industrial sector, continued into the 1950s. The British had dismantled
industrial capital in the course of the 1940s, while the Ethiopians would
starve the territory of investment from the 1950s onward, and thus Eritrea
entered a spiral of economic depression that would span several decades.
That sense of economic disaster—combined with political and cultural sub-
jugation and marginalization—unquestionably fuelled an upsurge in the
politics of violence by the end of the 1950s and beginning of the 1960s.
The history of the political parties and the key debates of the 1940s are
relatively well-known, even if the interpretation of that decade will remain—
as it should—a matter of heated debate.58 In crude summary, the 1940s was,
depending on one’s ideological positioning, either the period in which
Eritreans finally won the right to rejoin the Ethiopian motherland, and in
which unionism was the natural expression of Eritrean identity; or the era
in which nationalist consciousness was finally awakened, manifest in the
coalescence of several pro-independence parties. Neither pole is correct,
but likewise neither position is wholly invalid; the reality, however, is rather
more interesting, and can only really be understood by taking a long-term
view of the Eritrean region as a contested borderland, the fertile frontier
where the spaces within which groups define themselves are continually
made and remade. Moreover, the period highlighted not only conflict
between the Tigrinya frontier and the Amhara heartland, but fierce contests
within the Eritrean borderland, and, notably, between Tigrinya with differ-
ent visions of the nature of the territory; Eritrea was a contested space
116 colon i al i sms, ol d and new

among its inhabitants, and this would have far-reaching consequences, with
which Eritreans are still living.
Political views ranged across a spectrum from unconditional union, to
conditional union, to conditional independence, to unconditional inde-
pendence.59 Some—many, indeed—shifted their positions over time, and
certainly a number of these belonged to the first umbrella organization,
known as Mahber Fikri Hager Ertra (MFHE), or Society for the Love of the
Land of Eritrea, set up in May 1941. The MFHE started out as little more
than a coalition of various opinions-in-the-making—its members shared a
basic anti-colonial sentiment, and the organization might be considered at
least proto-nationalist—but it was increasingly dominated by a faction
which favoured some form of union between Eritrea and Ethiopia. Another
key faction came to disagree, and this was the point of departure: for them,
union—at least unconditional union—was unacceptable, and independence
for the territory, perhaps following a European trusteeship, was the way
forward. The fundamental point is that with the split in the MFHE, the key
battle lines were drawn, however much the precise position of the trenches
might alter one way or the other. Both unionist and separatist wings—
broadly defined—can be considered ‘nationalist’, again loosely interpreted.
Resisting the temptations of presentism, it is clear that those who advocated
union with imperial Ethiopia regarded themselves as patriots no less than
did those who envisaged an independent Eritrea. While it is clearly tempt-
ing to see the independence factions as somehow ‘more nationalist’, this is
essentially to see the history of the 1940s from the perspective of the 1990s,
when the independence movement had seemingly—for the time being—
won the argument. Modern distaste for contemporary references to Mother
Ethiopia should not blind us to the fact that the unionist position often
reads as stridently nationalist and patriotic, and that ‘union’ meant liberation
from European domination or the threat of it.60 It should further be noted
that the idea of ‘union’ has been variously defined since the 1940s and 1950s,
and in fact has been referred to positively in Eritrean political discourse
down to fairly recently.
This was, then, the era of competing nationalisms. Out of the MFHE
sprang what would eventually become known as the Unionist Party (UP)
in 1944, complete with its aggressive and militant youth wing, known as
Andinet, and its wholehearted backing from Addis Ababa, chiefly from the
direction of the Society for the Unification of Ethiopia and Eritrea, founded
in Addis Ababa in early 1944. The UP would also enjoy the active support
 de marcat i ng i de nt i ty 117

of the Coptic Church, and the shrilly vocal support of a pro-Ethiopian

lobby back in Britain.61 The independence faction, meanwhile, broke off
from the MFHE to found the Ertra n’Ertrawian—Eritrea for Eritreans—
organization in 1945.The latter movement would eventually evolve into the
Eritrean Liberal Progressive Party (ELPP), formed in February 1947. One
notable idea with which some members briefly flirted in the mid-1940s was
that of ‘Greater Eritrea’, or Tigray-Tigrinya, as it was also known. As a politi-
cal principle it appears to have fizzled out fairly quickly, but behind it was
the central notion that all Tigrinya-speaking peoples should be united
against ‘Ethiopia’, defined in this context in terms of Amhara hegemony. In
part, it has been supposed that certain kebessa elites were sympathetic both
to the oppression and decay suffered by the Tigrayan aristocracy, and the
plight of their brothers during the Woyane revolt of 1943, examined in the
next chapter. Certainly some scholars, notably Alemseged Abbay, have
become rather excited by the presence of the idea; at the very least its exist-
ence does indeed suggest a certain ambiguity on the part of some highland
Eritrean elites regarding their trans-Mereb identity.62 The Mereb frontier
was cross-cut by shared Tigrayan-kebessa loyalties and ties, very largely in
opposition to Amhara power.63 But certainly antagonisms between the
ELPP and the UP increased sharply over 1946–7—especially after an attempt
at reconciliation between the two factions collapsed in acrimony—with key
unionist Tedla Bairu highlighting the fact that leading pro-independence
leader Woldeab Woldemariam was of Tigrayan parentage, and therefore
could not legitimately participate in any discussions on Eritrea’s future.
This was clear evidence of major difference between two Tigrinya-speaking
blocs, difference—whether real or imagined, or both—which would
become all the sharper and more violent in the later part of the century.
In the meantime, while the Tigrinya highlanders were organizing them-
selves, the Muslim community formed the Muslim League (ML) at the end
of 1946, like the ELPP opposed to the union and indeed to the partition of
Eritrea between Ethiopia and Sudan.64 This was followed, in early 1947, by
the National Muslim Party of Massawa, and in September of that year
by the New Eritrea Pro-Italy Party which was also supported by a number
of Muslims who considered that some form of European ‘protection’ or
trusteeship might just be preferable to Ethiopian rule. Indeed, in many
respects, Eritrean Muslims were in the forefront of the early growth of
anti-Ethiopian nationalism, and remained so through the 1950s; the Muslims
of the western lowlands, in particular, had distinct historical reasons for
118 colon i al i sms, ol d and new

rejecting the emperor’s authority, just as they had reasons to feel hostile
toward the idea of joining Sudan. There can be little doubt that Muslims
across the territory had much clearer cause to champion an independent
Eritrea than did Christian highlanders.65 While the Tigrinya might share
certain basic tenets of faith and culture with the Amhara, agro-pastoral peo-
ples in the west had lived along a contested fault line since the early nine-
teenth century and had been subjected to violence from the direction of
Sudan and Ethiopia alike. In simplified form, this was the sentiment to
which such organizations as the ML appealed.Yet faith was only one fron-
tier of competition, and one which interconnected with a network of other
conflicts—a scarified political landscape in which a set of intertwined bor-
ders were beginning to open up dramatically by the late 1940s and early
1950s. There were tensions between the urban and the rural, between high-
land and lowland, and between provinces even within those broad divisions;
there were economic and religious frontlines, some of ancient standing,
others more recent creations. In the course of the late 1940s, especially after
the official British ban on political parties was lifted in 1946, a stream of
organizations bubbled to the surface, out of the fissures in the political land-
scape; a network of overlapping concerns and loyalties was manifest in the
continual formation and reformation, splitting and sub-dividing, of political
organisms. While the UP remained fairly constant and continued to enjoy
the support of—indeed was in some respects an appendage of—the
Ethiopian government, a host of other groups espoused various versions of
oppositional stances. There were tensions and ultimately splits among
Muslim lowlanders, largely as the result of the fact that the leading figure in
the ML, Ibrahim Sultan, was increasingly associated with Italian interests.
The ML now spawned other broadly Islamic groupings, such as the
Independent Muslim League—which now became basically unionist—and
the Muslim League of the Western Province, which held that union with
Ethiopia might be acceptable until such times as Eritreans voted to have
independence.
A broad church organization, the Independence Bloc, had come into
being in 1948 under Ibrahim Sultan, in part in order to petition the UN
Commission of Investigation dispatched to find out ‘the wishes of the
Eritrean people’.66 But by the beginning of the 1950s, there were tensions
within this coalition of pro-independence movements over the precise
meaning of ‘independence’, and how this might best be achieved;67 the
Liberal Unionist Party, for example, involving some former members of the
 de marcat i ng i de nt i ty 119

ELPP, came to advocate conditional union, presumably involving a form of

the federation which was eventually decided upon. Another group, the
somewhat ambivalently titled ‘Independent Eritrea United to Ethiopia
Party’, essentially came to espouse union. Other parties still, advocating
various forms of autonomy and trusteeship, but resisting both wholesale
Ethiopian takeover and the partition of Eritrea between Ethiopia and Sudan,
represented the interests of former ascari, or of the sizeable Italian and Afro-
European community. As of 1950–1, the Independence Bloc—which would
reform as the Eritrean Democratic Front in the middle of 1951—comprised
the Muslim League, the Liberal Progressive Party, the New Eritrea Party, the
Independent Eritrea Party, the Veterans Association, the National Party of
Massawa, the Intellectual Association, and the Italo-Eritrean Association. It
was a amalgam of interests, faiths, and cultures, loosely ‘united’ by the notion
that full annexation to Ethiopia would not be in most Eritreans’ best inter-
ests. It was an idea that cut across regional and ethnic lines, shared by groups
of highlanders and lowlanders, Christians and Muslims, urban-dwellers and
rural communities. Nonetheless, there remained, among a number of parties
and associations, a belief that ‘union’ might be variously interpreted, even
within the Independence Bloc, and that the concept per se was not neces-
sarily ‘a bad thing’, as long as it could be negotiated and was both condi-
tional and contingent. This is crucial, and is oft forgotten in the smoke and
noise of the battle that followed between the 1950s and the 1990s: condi-
tional union might have worked, had it been nurtured appropriately, and it
certainly was not inevitably doomed to fail.
The unionist cause, meanwhile, gathered pace, fuelled by Ethiopian prop-
aganda and funds. Violence was deployed, both by the Andinet youth mili-
tants and in the form of the co-opting of shifta along the Eritrean–Tigrayan
borderlands. Such violence took the form of beatings administered to, and
assassinations of, known separatists and those sympathetic to the separatist
cause, as well as the kinds of banditry common throughout the colonial
period aimed at keeping the borderlands unstable.The Coptic Church played
an integral role, too, threatening excommunication for anyone who espoused
the separatist cause; this had a significant impact among God-fearing com-
munities across the kebessa, although in fact it is striking that so many high-
landers continued nonetheless to support the independence cause.Whatever
the case, the Church was part of the socio-political establishment seeking full
restoration in an Ethiopian future. Indeed, the social context of the political
debate needs to be noted. Many of the early leaders of the various political
12 0 colon i al i sms, ol d and new

groupings—Ibrahim Sultan, Woldeab Woldemariam, Tedla Bairu—were part

of a newly educated class, a group which may have been rural in provenance
but which had been exposed to ‘Western’ education and which was increas-
ingly urbanized. Many of them represented what Jordan Gebre-Medhin has
described as ‘the first-generation Eritrean intelligentsia’, part of ‘a restless
new class’68 which might be considered in some respects displaced from its
roots. In some respects, they had emerged to stand alongside an older, ‘tradi-
tional’ elite, which, in Ruth Iyob’s view, sought the restoration of pre-colonial
status in the early 1940s,69 and herein was an early fault line: while some of
the new elite perceived that independence was the only way in which to
sustain social development and mobility, others, along with a late nineteenth-
century elite, saw union with Ethiopia as the guarantee of cementing their
places in the socio-political order. The frequently Byzantine nature of the
political debate and the fragmentation of Eritrean politics, ultimately, reflected
two things. First, it reflected the network of political tectonics across the
northern region which could be dated to the eighteenth and nineteenth
centuries; it represented a perpetuation of extant conflicts, borderlands across
which were contested faith, ethnicity, culture, province. Second, it reflected a
deep level of uncertainty about what ‘Eritrea’ actually meant. The splits
within an otherwise confident, talented new elite—from Tedla Bairu to
Woldeab Woldemariam—were symptomatic of an inherent diffidence about
what it meant to be ‘Eritrean’, and therefore represented the updating of a
series of more ancient struggles across the territory. A rather more militant,
assertive sense of Eritreanness would only emerge at the end of the 1950s and
the beginning of the 1960s.
Against these continuing political debates was the violence of the period.
Through much of the 1940s and 1950s, Eritrea was a violent place—in
terms of low-level, though episodically explosive, insurgency, as well as more
common forms of opportunistic armed criminality. Too often the activities
of shifta in this period are subordinated in importance in scholarly analyses
to the political debates taking place in Asmara, Massawa, Keren, Tessenei; in
fact they are central to understanding the 1940s and 1950s, and certainly
emblematic of the frontier zone in la longue durée. As we have seen, shiftanet
was one of the most important socio-political mechanisms available to indi-
viduals and groups with grievances, facilitating personal vendetta, political
resistance, and organized criminality—sometimes all three simultaneously. It
was a constant in habesha life, stretching back into the nineteenth century
and further back still; certainly, as we have noted, shifta activity persisted,
 de marcat i ng i de nt i ty 121

albeit perhaps on a lower level, through much of the Italian period, even if
it may not be described as explicitly political.70 Nonetheless it seems safe to
suggest that the 1940s witnessed the rise of new forms of shiftanet, in addi-
tion to those older patterns of armed defiance: whereas in the nineteenth
century the term often had noble connotations, by the mid-twentieth cen-
tury shiftanet also embraced the rural dispossessed and economically and
politically marginalized, and came to be associated with common ‘banditry’
and criminality as well as larger ‘noble’ causes.
From the beginning of the 1940s, violence escalated dramatically in some
parts of the territory. The history of modern Eritrea—the history of the
northern borderland—is the history of the transition from shifta to sha’abiya,
from patterns of banditry played out on the frontiers of culture and polity
to the ‘revolutionary’ guerrilla organization which attempts to seize control
of the borderland. Further, the escalating violence in the territory by the
early 1950s—when a general amnesty in the twilight of the British admin-
istration brought about a temporary lull—hastened the transfer of Eritrea to
Ethiopian overlordship; while unionist violence forced the issue in the late
1940s, the insurgency dissuaded interested foreign parties from considering
any other option than the incorporation of Eritrea into Ethiopia, albeit
within a federal framework. It is arguably the first instance in the making of
modern Eritrea of violence being deployed—successfully—in the pursuit
of a particular political issue. At any rate, the lessons for later nationalists
were there for the learning.
With the defeat of the Italians, shifta activity increased, for several reasons.
Jordan Gebre-Medhin’s work is important in highlighting the importance
of a dispossessed nobility leading to an upsurge in politically motivated vio-
lence across the kebessa in the mid-1940s.71 In particular, a great deal of shifta
violence has been understood as the armed wing of unionism, with various
bands of malcontents utilized for the purpose of terrorizing known or sus-
pected pro-independence figures and their supporters.72 The campaign of
bombing, assassinations, and more general tactics of intimidation—as well as
the destabilization of rural highways, for example, which bore the hallmarks
of rather more straightforward criminality—apparently intensified during
the periods when the Four Powers Commission, and later the UN
Commission, was deliberating over Eritrea’s future.The aim, it seems, was to
persuade foreign observers that Eritrea would never be pacified unless its
rightful status—a province of the Ethiopian motherland—was achieved. In
fact, of course, there were several factors fuelling shifta violence in this
12 2 colon i al i sms, ol d and new

period. The last British governor of the territory, Sir Duncan Cumming,
considered banditry simply as something which ‘usually emerges in Eritrea
and Ethiopia in times of stress’73—as good a summation as any, perhaps—
but Trevaskis provided rather more in the way of analysis.74
At the general level, there was the idea—not without merit, even though
it conforms to a standard colonial stereotype about the prominence of the
military in African life—that across the Eritrean and Ethiopian highlands, a
life spent at arms was covered in glory and associated with both honour and
enhanced social status. Military activity, very loosely defined, was an impor-
tant way of achieving social mobility, as well as effecting political change—
and there is indeed much evidence to suggest that this was the case, in the
nineteenth century as in the twentieth. In the 1940s, Trevaskis opined, the
only means by which this ‘spirit of military adventure’ could be sated was to
become shifta. There was no other outlet for it, and again—although we
need to exercise caution here, as in dealing with European sources for the
nineteenth century—it might be argued that there was a basic truth in
Trevaskis’ assessment. Combined with the large scale demobilization of
thousands of Eritrean ascari from the Italian colonial army, and the availabil-
ity of weaponry scattered across the territory in the early 1940s, it went
some way to explaining the significant upsurge in shifta activity throughout
this period. Eritrea was awash, for example, with an array of machine guns,
rifles, pistols, and hand grenades in the years following the overthrow of
the Italians, and thousands of Eritreans knew how to use them. Two further
factors were important. One was the economic depression, noted earlier;
unemployment and increasingly desperate living conditions, especially from
1946 to 1947 (coinciding with international deliberations on the territory’s
future), drove many young (and old) Eritreans with access to a few rifles
into a life of shiftanet. Second, there was political uncertainty, and there is no
doubt that this intersected with the prevailing economic conditions and
fuelled the existing culture of political violence from the mid-1940s onward.
Trevaskis noted that shifta often acted with impunity, calculating that the
BMA was not destined to remain in place for long, and that there was a
general lack of belief that the British could do much about banditry with
the limited means at their disposal, and given their evident lack of long-
term commitment to Eritrea.
Certainly, much shifta activity was rooted in personal grievance and feud,
even if larger communities often took up arms for a particular local cause.
A century earlier, indeed, Walter Plowden appears to have been describing
 de marcat i ng i de nt i ty 123

much the same phenomenon across the northern mountains.75 But local dis-
turbances frequently became politicized, just as they did in the mid-and late
nineteenth century; ‘liberation struggle’ as it emerged in the 1960s and 1970s
was rooted in the complex intersection between local and ‘national’ issues in
the 1940s.Violent struggles between ruling families in the kebessa—notably in
Hamasien and Serae in the early and mid-1940s, for example—which stretched
back into the nineteenth century became complicated by the identification
of one or other group with larger political causes, namely unionism or sepa-
ratism, broadly understood.76 Local shifta were brought into these struggles
and deployed accordingly, although much of the overt violence—which
ranged from beatings to destruction of property—was associated with union-
ist-employed shifta against those who had identified themselves with the pro-
independence camp. A further issue involved the continued Italian presence.
In the minds of many unionists, pro-independence activists were preparing to
‘do business’ with the Italians, who might therefore return to power in some
shape or form; thus were Italian land-owners, who had been the beneficiaries
of land concessions at the expense of local communities, targeted relentlessly
through the 1940s. Many of the shifta involved saw themselves as ‘patriots’, in
this respect, attempting to drive out the hated colonial occupier.Thus we can
identify, at the most basic level, two broad strands of the ‘new’ shiftanet of the
period, namely that which was rooted in local causes and disputes, and that
which was tied to larger political causes affecting the territory as a whole, and
developing into a full-scale insurgency. Both strands escalated as the 1940s
progressed, notwithstanding a temporary lull of sorts in the 1944–6 period,
and each increasingly intersected with the other.
Immediately following the British occupation of Eritrea, there was a cer-
tain calm across the territory, but by 1942 this was beginning to be threat-
ened by low-level activity.77 Raiding across the Tigray–Serae frontier seems
to have been related to the deposition of the ruler of the Adi Quala district,
Dejazmach Haile Tesfamariam, who had links with Tigray; this violence con-
tinued into 1943. At the same time,Tigrayan shifta were beginning to resume
raids on the Kunama around Barentu. Again, the years 1944–6 were report-
edly relatively quiet—quiet enough, indeed, for the British ‘Frontier Striking
Force’ to be disbanded and the number of police to be slightly reduced—
while the British used the opportunity to remove a number of troublesome
Hamasien chiefs. But in 1947 shifta violence increased sharply across the
kebessa, and included attacks on pro-independence figures and Italians, while
Tigrayan bands increased the incidence of their attacks on the Barentu area.78
12 4 colon i al i sms, ol d and new

By 1949, the violence had amounted to something approaching a unionist

shifta insurgency, especially on the plateau, although this was compounded by
ongoing conflict between various groups around Barentu and across the
western lowlands. The British increased their military presence as the UN
Commission began their deliberations; it was no mere coincidence that shifta
activity increased so sharply as the UN delegates arrived. In the course of
1950 and 1951 the British security forces struggled to deal with the shifta
violence: their deployment of locally recruited armed counter-gangs along-
side regular forces (which included British troops as well as the Sudan
Defence Force) foreshadowed similar tactics in Kenya just a few years later,
but it was met with limited success in Eritrea, where by 1950 the conserva-
tive estimate was that some 2000 shifta were active across the territory. Even
the Eritrean Police needed to be watched carefully, as many of them were
suspected of being sympathetic to the unionist cause, and a number deserted
in the final years of the British administration.
As for the shifta themselves, these were fighters who might easily melt
into their local communities, or into the hills, or—particularly irksome for
the British—across the border into Tigray. In addition to patrolling the
highly porous border areas and vulnerable main roads, and creating mili-
tary ‘special areas’ where shifta were especially active, the British adopted
the soon-to-be-common tactic in counter-insurgency warfare of collec-
tive punishment, imposing fines, confiscating property, and arresting key
figures in communities suspected of harbouring or otherwise supporting
shifta.79 Ethiopian support—tacit or overt—presented a rather more deli-
cate problem, involving as it did the need to place pressure on an ostensibly
sovereign state. Just how much direct imperial involvement there was in
‘terrorism’ is perhaps debatable; but numerous shifta did operate out of
Tigray—many were indeed Tigrayan—and it was felt that the Ethiopian
government could have an impact by publicly condemning shiftanet and
naming those receiving succour in the north. Addis Ababa was initially
loath to do so; unionist shifta were not terrorists but patriots, cut from the
same cloth as those who had taken to the bush to fight the Italian occupi-
ers of Ethiopia in the late 1930s. In the ‘official mind’ of habesha imperial-
ism in the late 1940s and early 1950s, there was a direct line of continuity
between anti-Italian guerrillas of 1936–41, and the fighters now taking up
the cause in the ‘occupied north’. However, by 1951–2, the Ethiopians were
cooperating on the Tigrayan side of the Mereb frontier zone, dispatching
‘anti-shifta’ missions to consult with the British and attempting to patrol
 de marcat i ng i de nt i ty 125

the border with rather greater intent;80 a cynical interpretation, however,

might have it that this was—by the early 1950s—an empty gesture, as by
this time the insurgency had largely achieved its goal of some form of uni-
fication with Ethiopia.
At the same time, conflict of a somewhat different kind had been unfold-
ing in the western lowlands. The war between the Beni Amer and the
Hadendowa dated to the early 1940s, and stemmed from Hadendowa claims
over pastureland between the Eritrean frontier and the Barka river, although
in fact it appears to have been both instigated—and pursued with much
greater ferocity—by the Beni Amer.81 It quickly brought about major inse-
curity in the area of both Agordat and Barentu, and provided something of
a western lowland vortex within which other localized conflicts erupted—
similarly over access to land—involving the Kunama, other groups of Beni
Amer and the Nara. At the same time,Tigrayan shifta conducted raids on the
Kunama, and from the BMA’s viewpoint were harder to apprehend, owing
to their ability to shrink back into the Adi Abo area of north-west Tigray.
There were also ongoing skirmishes, over land and cattle-theft, between the
Beni Amer and groups of farmers on the western Hamasien plateau; and a
similarly land-related feud between the Tsenadegle in Akele Guzai and the
adjacent Teroa, associated with the Saho of the eastern escarpment. While
the Tsenadegle were Christian and the Teroa Muslim, this seems to have
mattered less than rights over particular tracts of grazing land; importantly,
it was a dispute which dated to the early Italian period, if not earlier, as
many such disputes did. The contest between the Beni Amer and the
Tigrinya of Hamasien, for example, was not a novel one in the late 1940s.
The Beni Amer-Hadendowa war was negotiated to an end by late 1945, but
it had created a great deal of upset across the area; it had also intersected
with an initially isolated peasant revolt among the Tigre in the north against
their ‘aristocratic’ rulers, which by the mid-1940s had spread into the Barka
area. The hard-pressed BMA had managed to bring about some form of
‘serf emancipation’ by the end of the decade, but it had been a traumatic
decade for the communities involved, characterized by a fair degree of rein-
vention and creativity on the part of the people themselves as well as the
British authorities.82 Importantly, the socio-economic disruption and dislo-
cation occasioned by the various conflicts in the west and north bred new
patterns of violence, enhanced existing ones, and produced an armed
‘class’—for example, Hamed Idris Awate, later the putative pioneer leader of
the Eritrean liberation struggle—willing to turn their guns on whatever
12 6 colon i al i sms, ol d and new

system they perceived as the enemy, and willing to be mobilized for larger
political causes. In this group were situated the progenitors of the nationalist
liberation struggle, the armed pioneers of the early Eritrean Liberation
Front. Importantly, there emerged a large, if far from coherent, pool of
Muslim consciousness across this swathe of Eritrea into which early Muslim
nationalists might tap—particularly in terms of fears of Christian domi-
nance, from whichever side of the Mereb.
In the end, although it had some success in resolving the more localized
disputes over land, the BMA was unable to bring the shifta under control by
military means. In 1951, they declared a general amnesty, and over several
months several hundred fighters took advantage. Others were continually
reported as being anxious to do so, including Hamed Idris Awate, although it
seems that many, including Awate, regarded the offer of an Amnesty merely as
an opportunity to ‘rest up’ and prolong negotiation. The larger picture sug-
gests that the British had been soundly beaten. With the passing of the UN
resolution federating Eritrea with Ethiopia at the end of 1950, the main aim
of the outgoing BMA was the most basic management of Eritrea’s affairs so
that the territory could be handed over, under the uncomfortable glare of the
UN itself, on time and in reasonably good order with Britain’s reputation for
governance more or less intact.83
Above all, the insurgency of the late 1940s and early 1950s presaged the
violence both of the liberation struggle, and of the civil war that would rage
at its centre. The shifta crisis was of course the outcome of the unstable,
contested borderland of some antiquity, as well as representing the tried and
tested mechanism of political and socio-economic protest in the region. In
that sense there was a great deal of continuity from the pre-1900 era; there
is certainly something in the pattern of events in the 1940s and 1950s which
is redolent of the 1860s–80s period across the kebessa, in terms of divided
loyalties, the ambiguities of the frontier zone, and opportunistic and low-level
but pervasive violence. At the same time, however, mid-twentieth-century
shiftanet was also the symptom of very particular grievances and circum-
stances: unemployment and economic anxiety more broadly were particu-
larly important. We have seen how in earlier periods such economic stresses
were crucial determinants of violence, driving forward the militarization of
communities and the propulsion into the bush of disaffected and heavily
armed young men. Militarization was what occurred when such communi-
ties were denied access to critical resources; groups took up arms in the
absence of industry, and political causes flourished in these fertile frontiers.
 de marcat i ng i de nt i ty 12 7

It was so at key points in the nineteenth century; and in the mid-twentieth

century, economic trauma provided a fertile seedbed indeed for political
radicalization—a process of radicalization which continues to reverberate
across the region. In the 1940s and 1950s, economic uncertainty intersected
with political uncertainty and upheaval, the proliferation of guns across the
region as well as ex-servicemen well-versed in their use in the wake of the
Second World War. At the same time, more specifically, many shifta were
heroes to some, insofar as they represented new forms of political con-
sciousness, pro-union or anti-Italian—either way, essentially, anti-colonial.
More generally, shifta themselves represented an early example of the armed
frontier marching on a contested centre: it would be repeated in the decades
that followed, and provided some early, but salutary, lessons for those who
would later take to the bush in ever more sophisticated ways in order to
engineer political change.
It was against this background, then, that decisions were taken by
foreign powers regarding Eritrea’s future.84 In December 1950, the UN
passed Resolution 390A(v) which proposed that Eritrea would be feder-
ated with Ethiopia, as a constitutionally autonomous unit, under the
Ethiopian crown. The Ethiopian government regarded it as a less than
total triumph—it was not, after all, full incorporation. But it was a start.
As provided for by the Federal Act, Eritrea was to have a Constitution
which would safeguard the rights and freedoms of its citizens, and which
would in turn, in theory, be guaranteed by the UN itself. That
Constitution,85 adopted by the first Eritrean Constituent Assembly in July
1952, was representative of an ill-disguised tension between the suppos-
edly budding democratic culture in Eritrea and the ‘feudal’ omnipotence
of the Emperor in Ethiopia. In political and legal terms, it was profoundly
problematic. Even more significant for the medium term future was the
fact that around the same time the Constitution was being adopted, the
US signed a mutual defence pact with the Ethiopian government. In
return for military assistance and development aid, Haile Selassie leased
Kagnew base in Asmara to the US for twenty-five years. It was to become,
for a time, the US’s largest communications base outside the US itself.
This secret agreement was to prove of rather greater significance—and
indeed permanence—than the 1952 Constitution. The Federation for-
mally came into being in September 1952, when the British withdrew
and the Eritrean Constitution came into force. At that point, there was, by
most accounts, a reasonable level of optimism that the system might work,
12 8 colon i al i sms, ol d and new

given the appropriate safeguards and degree of good faith on the part of
Ethiopians and Eritreans. Its swift failure—engendered by both interna-
tional duplicity and Ethiopian imperial aggression, according to a later
generation of radical Eritrean nationalists—would provide the latter with
a foundation stone around which the architecture of grievance would
come to be created. It would prove a robust structure indeed.
 6
The Empire of Haile Selassie,
c.1900–74

The anomalous empire

Haile Selassie inherited an intrinsically unstable, restlessly violent state. Yet
Haile Selassie himself occupied a curious position: a ‘modernizer’ to some,
arousing opposition accordingly, and the arch-conservative to others, with
fierce critics in close attendance. In truth, in the end, he was the latter rather
more profoundly than he was the former, but those who overthrew him
would prove themselves no more capable of revolutionary change than he
was; they themselves were products of the system, and ultimately proved
themselves not revolutionaries, but legatees with occasional flashes of crea-
tivity. That is for later in our story. For now, suffice to say that our concern
in this chapter is with how Haile Selassie inherited both the restless cen-
trifugal militarism of the nineteenth century and the unstable frontiers of
the Menelik state. Our concern is also how he and the Amhara political
establishment spectacularly misunderstood—wilfully and otherwise—what
it was they had inherited, and in particular its inherent flaws, and how this
spelt a level of destruction with which the region is still dealing. The out-
come of Haile Selassie’s reign was the exacerbation of a number of extant
wars, with some bitter new twists added.
Nonetheless, the young son of Ras Mekonnen was not the initial choice
to succeed the old emperor; rather, it was Iyasu—known as Lij Iyasu, owing
to his youth, the son of Ras Mikhail of Wollo and Menelik’s daughter—
who was his chosen heir.1 Menelik had designated his successor in 1909
while also exhorting his ministers to ensure a peaceful succession and to
avoid the political upheavals of the previous century. Yet the situation thus
created involved a fragile balance of power. Owing to Iyasu’s minority, a
130 colon i al i sms, ol d and new

regent was appointed in the person of the Shoan noble Ras Tesemma Nadaw;
meanwhile the young Iyasu had already been married to the daughter of
Ras Mangesha Yohannes of Tigray in an attempt to consolidate the alliance
between Shoa and that province. Hovering over all these proceedings, while
her husband’s life slowly ebbed away, was Empress Taitu, who belonged to a
long tradition of powerful women in habesha political life. Her continual
interference—and her Gondari background—drew fierce opposition from
the Shoan political establishment, which turned to the increasingly politi-
cally important mahal safari, the units of the imperial army attached to the
palace. They moved against Taitu in 1910, and she was sidelined to Entotto
on the outskirts of Addis Ababa—though in a final dignified act, she
appealed to her brother, Ras Wale of Yajju, and Mikhail, Iyasu’s father, to
avoid bloodshed as they squared up to one another on the contested bor-
derlands between Yajju and Wollo in the north. Iyasu’s regent Tesemma died
in early 1911, and thus began the young man’s de facto reign, even though
Menelik clung on to life until 1913.
Iyasu is deserving of much greater attention than for a long time he
received.2 He appears a complex character, capable of administrative reform
yet a slave-dealer whose reign was characterized by a system of personalized
rule. His much-discussed hedonism was in stark contrast to the abstemious-
ness of an older generation. But this was unimportant alongside the larger
issues of politics and faith which combined to be his undoing. He moved
increasingly against his grandfather’s generation of advisors and indeed
against the Shoan heartland, favouring his father’s province of Wollo instead;
indeed, his father Mikhail became negus of Wollo and Tigray in a remark-
able move which saw Wollo rival Shoa as the political centre of the empire.
Iyasu (who was never actually crowned, either as negus or as negus negast) was
contemptuous of the Shoan nobility, and spent much of his time outside
Addis Ababa.The political concerns of the Shoan elite were paramount, but
they were easily disguised in religious garb, for Iyasu’s tolerance toward
Islam and indeed abiding interest in the rehabilitation of Muslims in
Ethiopian society and polity led to (inaccurate) rumours that his own con-
version was imminent. His father had been a Muslim, converting to
Christianity in the late 1870s, and to the Shoan nobility it all added up: a
residual Islamic influence, both in the family and within his retinue, his
building of mosques, his marriages to the daughters of important Muslim
families, his wanderings in the (predominantly Muslim) Ogaden. Indeed his
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 131

interest in the Ogaden was another striking feature of his short reign; he
spent a great deal of time there, opened up a new level of engagement with
local Somali elites, and provided support to Muhammad Abdille Hassan in
his war against the British and Italians.
Of course, he had also married into prominent Christian families, and pro-
vided funds for the building or expansion of churches and monasteries—but
no matter. By 1916, the Shoan Christian elite, with the support of neigh-
bouring European administrations, were ready to move against him; European
colonial officials were persuaded by talk of Iyasu’s imminent apostasy, and
further fuelled the rumours. He was formally deposed in September 1916 by
an alliance headed by Zewditu, Menelik’s daughter by a previous marriage.
Iyasu himself was defeated in a pitched confrontation in the nineteenth-
century fashion some way east of Addis Ababa, and retreated to Afar country;
Negus Mikhail mobilized indignantly, assembling a force of some 100,000
men. The new regime in Addis Ababa was soon able to mobilize a sizeable
force of its own, from Shoa and from the south and west.While the coalition
troops assembled on the northern frontier of Shoa, Mikhail advanced, and
battle was joined at Sagale, a few miles to the northeast of Addis Ababa in
October 1916. Mikhail’s army was comprehensively defeated by that of
Fitaurari Habte Giorgis, a decisive moment in modern Ethiopian history, in
what would prove to be, in essence, a victory for the forces of conservatism.
Menelik’s order would be preserved. The whole episode was also a sharp
reminder of the deep-seated religious tensions at the heart of the Ethiopian
state, regardless of the lack of evidence for Iyasu’s imminent apostasy: Islam
was feared and despised, and the early-twentieth-century polity was founded
upon Christian intolerance which was itself rooted in a history of religious
antagonism. Ethiopia was to be a Christian state, and although Muslims
might serve even in high office, they were in effect co-opted by the Solomonic
polity. The Amhara political establishment, no less than the European colo-
nial authorities in neighbouring territories and indeed probably more so, was
keenly aware of its being surrounded by several million Muslims, and of the
perils of a radical, political Islam becoming active across northeast Africa.
Iyasu himself remained a fugitive until his capture and imprisonment in 1921,
although even in detention he remained a figure of considerable allure to
many, especially as pools of opposition emerged against Ras Tafari.
Tafari himself—formerly governor of Harar, and moved by Iyasu to
Kaffa—may have been an influential figure in the alliance that led the coup
132 colon i al i sms, ol d and new

against Iyasu, but he was very much a junior member of that alliance, and
subject to greater authorities than himself, including Empress Zawditu; his
elevation to Zawditu’s regent was precarious.3 Tafari found himself con-
ducting the day-to-day administrative business of the empire, but very much
under the shadows of the powerful Zewditu herself, the great general Habte
Giorgis, and various provincial governors. The soldiers at his own disposal
represented a paltry force, and his military force remained for many years
greatly overshadowed by the larger, more impressive provincial forces, under
much more experienced military commanders; Tafari himself had no mili-
tary experience to speak of. He slowly bolstered his position in a number of
ways. He concentrated on foreign affairs, effectively becoming Ethiopia’s
foreign minister, and sought to project a favourable image of Ethiopia
abroad. Recognizing the importance of Shoa at the centre of the empire, he
formed an alliance with Ras Kassa Hailu, the province’s most important
political figure. He thus focused on Shoa throughout his career, and on
Addis Ababa, which very much became his own creation and the arena in
which so many political battles would be fought. From the early 1920s
onward, Addis Ababa became an increasingly important socio-political
environment, a swelling and politicized urban jungle which represented
many of the stresses and strains in the Ethiopian polity. The restless city
would, in the longer term, become an incendiary device at the heart of the
imperial regime.
The early months of the regency were defined by a military crisis, devel-
oping in the aftermath of the defeat of Mikhail of Wollo. In the course of
1918, thousands of soldiers protested that they had failed to receive pay and
food, and blamed their venal officers, who in turn accused the council of
ministers (itself something of a recent innovation) of incompetence. A mil-
itary committee was formed and demanded the dismissal of the council,
and its replacement by a regency council comprising Tafari, Zewditu, and
Habte Giorgis. They had their way, and in the aftermath of the dramatic
reshuffle, Tafari began to appoint ‘new’ men to key government posts; this
modernizing reorganization of the Ethiopian government through the
1920s brought about enhanced centralized control and a much greater flow
of information from the provinces to Addis Ababa. It was a policy which
gave rise to a new generation of habesha officialdom, the Young Ethiopians,
who would be, in Marcus’ definition, ‘efficient, modern, and patriotic’—
but the transition to this new breed of government bureaucrat would be a
gradual one.4
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 133

Slavery had long been endemic in highland society, and with the increas-
ing attractions of commercial agriculture—for example coffee in the
south—landlords sought to strengthen their control over both land and
labour. Indeed Menelik’s expansion of the empire in the late nineteenth
century provided a fillip to the slave trade in the early decades of the twen-
tieth: these new frontier regions in the south and west were fertile hunting
grounds for the acquisition of slaves by the nobility, for whom the owner-
ship of people remained an indicator of social status. Slaving campaigns,
involving, as in the nineteenth century, the large-scale organization of men
and firepower targeting dispersed agro-pastoral communities, regularly
resulted in massive dislocation in the deep south, in newly acquired regions
such as Maji and Kaffa; Lij Iyasu himself led one such campaign, in 1912.
The slave trade continued to thrive on the western border with Sudan, too.5
In short, these new commercial frontiers in southern and western Ethiopia
would greatly alter the context within which the political battles of the
twentieth century would be fought. In the wake of Menelik’s imperialism,
a new set of provincial governorships became available—highly lucrative, in
terms of both trade and production, and politically weighty—while the
state’s organizational capacity for local exploitation was greatly enhanced.
The expansion of the highland ruling class to the south and west created,
over the longer term, reservoirs of resentment and collective experiences of
brutality at the hands of an Amhara ruling class. In 1918, Haile Selassie had
issued an edict abolishing the slave trade, but had added the strong caveat
that this was only the first step toward the abolition of slavery itself. He was,
in sum, having to address the whole system of violence which underpinned
Ethiopian political economy.6
His cunning, patience, and inscrutability became proverbial; but he was
also opportunistic, and needed to be, given the stubborn forces ranged
against him. Notably, his camp was associated (in the minds of Zewditu’s
followers) with the dilution of Ethiopia’s hard-won independence through
the increasing foreign involvement necessary to achieve political and
material progress; but in fact, for many of the western-educated reformers
attached to Tafari, his professed reformism was, in the end, limited in its
ambition and indeed at times disingenuous. These reformers—in many
cases inspired by the model offered by Japan—advocated ‘modernization’
in various spheres, including the economy, the constitution, the civil serv-
ice, and education;7 but although Tafari was indeed enthusiastic about
many of the accoutrements of European modernity, notably in the
134 colon i al i sms, ol d and new

technological realm, his primary concern was the consolidation of

Solomonic centralism and autocracy, and whatever ‘modernity’ could offer
him in the achievement of this. Following the death of Zewditu’s key ally
Habte Giorgis in 1926, Tafari went about undermining Zewditu’s author-
ity and enhancing his own power base. He dismissed Dejazmach Balcha,
influential governor of coffee-rich Sidamo and a key conservative oppo-
nent, in 1927; crushed an attempted mutiny by the formidable Dejazmach
Abba Weqaw Berru, head of the palace guards, in 1928; and utilized his
supporters’ unrest in Addis Ababa to forward his claims to the imperial
throne. Elevated to Negus in 1928, and crowned as Emperor Haile Selassie
I in 1930, his accession was not universally welcomed, and his position was
far from secure. Arguably his most dangerous challenge came from Ras
Gugsa, governor of Begemeder and Zewditu’s former husband. Impor-
tantly, Gugsa was supported by the Italians in Eritrea, and by his neigh-
bouring governors in Gojjam and Tigray, although they, in the end, held
back from outright rebellion. Gugsa’s refusal to acknowledge several orders
from Addis Ababa was tantamount to rebellion, and his defiance culmi-
nated in the battle of Anchem, in March 1930, in which Gugsa was killed
and the insurrection crushed.8
At Haile Selassie’s coronation, Wilfred Thesiger opined that the newly
crowned emperor was unpopular, and described how ‘[s]trange rumours are
afloat that Lij Yasu is coming back in the role of a saviour of the country’,
and how a number of the key governors had been biding their time during
Gugsa’s revolt to see the outcome.9 No doubt Thesiger was referring to
Gojjam and Tigray.Yet the 1931 constitution formalized absolutism, and laid
the framework for its expansion. By the mid-1930s—the eve of the Italian
war—it was widely acknowledged that Haile Selassie had gone some way
toward ‘modernizing’ his administration, which in effect meant a height-
ened level of monarchical authoritarianism, in particular through the place-
ment of ‘loyal’ governors over formerly troublesome provinces. Such
appointments, moreover, were increasingly based on merit rather than
inheritance. The ghosts of the mesafint—at least some of them—appeared,
for the time being, to have been laid to rest. In the Christian, Semitic heart-
lands, Shoa, Amhara, and Gojjam appeared pacified and stable—or at least
those who distrusted Haile Selassie were temporarily in abeyance; a greater
sense of stability was discernible even in the west and south-west, too, not-
withstanding the increasingly troubled autonomy of Jimma, granted by
Menelik following Abba Jifar II’s peaceful submission in 1882. The
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 135

incorporation of Jimma in 1932—doubtless motivated by economics but

justified by Abba Jifa’s grandson’s failure to cooperate with central govern-
ment10—arguably sparked a renewed habesha imperial land grab, interrupted
by the Italian occupation but culminating in the securing of the Ogaden
and Eritrea in the 1950s. There had also been the successful incorporation
of some southern and western aristocracies into the habesha ruling elite,
such as that of Wellega, for example; these often proved themselves adept at
imposing systems of taxation on the new territories, establishing more effi-
cient forms of economic and political dominance, and simultaneously rein-
forcing both their own positions and that of their Amhara overlords in the
process. Clearly, as Bahru Zewde points out, ‘the class basis of exploitation
and oppression was as important as the ethnic one’,11 and such systems of
co-option were particularly effective in the south and west.
In the north, and on the eastern side of the Rift Valley, the picture was
rather different from that presented in the central highlands. Haile Selassie
may have secured himself at the centre, but his authority diminished by
degrees further out.Tigray under Ras Kassa was equivocal, as would be evi-
dent during the Italian invasion of 1935–6, and in the early 1940s. In the
north-east, ‘Ethiopia’ scarcely existed, no more than did Italian Eritrea or
French Somaliland. Neither the Aussa sultanate—even more autonomous
than Jimma, based largely on ignorance that Haile Selassie even laid claim
to it—nor the Afar federation recognized the authority of Addis Ababa.
Thesiger’s lively descriptions of hunting in Danakil country in the early
1930s indicate the extent to which this was territory which was hardly
under the control of Addis Ababa—or even Harar—in any meaningful
way.12 These were the proverbial ‘wild tribes of the frontier areas’, according
to contemporary scholarship.13 In the late 1920s, there was serious unrest in
eastern Tigray and Wollo, regions struck by drought and locusts, as the high-
landers descended into the Afar lowlands on raiding expeditions; indeed it
was Ras Gugsa’s unenthusiastic response to an order to restore order here
that led to the denouement between him and Haile Selassie.14 The pro-Italian
Evelyn Waugh would write dismissively of ‘Ethiopian unity’ in the mid-
1930s, and of restive and resentful subject peoples groaning under ‘the dead
weight of the Abyssinian occupation’.15 These, he asserted in his bombasti-
cally inimical way, were only too glad to see the end of ‘Abyssinian’ rule—as
in Harar, where Italian ‘liberation’ was eagerly received. Haile Selassie’s state
collapsed easily enough, owing to sullen subject peoples, rampant disloyalty
within the composite Ethiopian army, and the ‘ungovernable’ places into
136 colon i al i sms, ol d and new

which Haile Selassie’s rule barely reached.16 Much the same was true of the
Ogaden, which was a vast, dangerous borderland in the 1930s; south of
Harar or Jijiga habesha power scarcely extended beyond the well-armed
caravans necessary to conduct trade with the European-governed Somali.17
Notably, ‘[t]he Ogaden—the chief Somali tribe of the south’, wrote Jones
and Monroe in the mid-1930s,
are equally primitive and lacking in consciousness of a nation to which, offi-
cially, they belong....Whatever the progress made in Addis Ababa, it will for
years to come be idle to mention the word modernisation in the same breath
as these wild lowland tribes. It is they who, bent on blood feuds and quarrels
over wells and grazing grounds, are responsible for the frequent raids into
British, French, and Italian territory. The modern frontier lines are as little
known to them as is a national feeling for Abyssinia.18

Moreover, borderlands to the south and west of the empire inherited by

Ras Tafari had been impossible to govern since the partition of the region
in the late nineteenth and early twentieth centuries.They represented zones
of alternate realities, alternate, that is, to the various exercises in modernity
and putative civilizing missions unfolding in distant metropoles—whether
British or Amhara or Italian. The complexities of local interaction are cap-
tured in Almagor’s studies of the Dassanetch, notably.19 The Northern
Frontier District (NFD) of Kenya abutting southern Ethiopia was the scene
of ongoing skirmishes in the years either side of the First World War between
the King’s African Rifles and Amhara, Oromo and Tigrayan ‘raiders’ and
‘poachers’; it was a volatile and violent zone of interaction, requiring a con-
tinual military presence on the part of the British, toward whom Amhara
forces and renegade shifta were apparently equally hostile. Sir Philip Mitchell,
senior British colonial official in east Africa, would write ruefully of ‘the
everlasting murderous raids from across the Abyssinian frontier’ which would
render life and service in the NFD so intolerable.20 As the KAR historian
Lt-Col Moyse-Bartlett put it, describing the situation between 1908 and
1914:
...the role of the KAR became increasingly that of a frontier force. The pres-
ence of unadministered tribes near Lake Rudolf, in country where illicit
ivory-hunting and gun-running still flourished, and of intractable Abyssinian
neighbours beyond the Northern Frontier District, who rendered scant obe-
dience to the control even of their own government, punctuated the monot-
ony of service in those barren and little-known areas with unexpected
incidents.21
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 137

Ethiopian raiding incursions, targeting scattered agro-pastoral settlements

across the region, were frequent, however, and the situation in this arche-
typal frontier zone was complicated by local patterns of resistance—from
the Turkana, notably—and by the influx of Somali herdsmen toward the
Moyale–Marsabit area in search of water, especially in times of drought, as
in 1909. Communities were caught in the crossfire of militarized societies,
for example the Samburu and Rendile, regularly squeezed between the
raids of the Turkana and armed habesha bands in search of ivory and live-
stock. For Ethiopians, these were natural hunting grounds; they were also
places of commercial opportunity, with a large belt either side of Lake
Turkana representing a burgeoning market for intrepid highland peddlers of
the tools of destruction. The reach of the armed Ethiopian frontier was felt
elsewhere, too: in northern Uganda, the first British units in 1911–12
encountered well-armed Acholi who had enthusiastically purchased guns
from Ethiopian traders as security in uncertain times. Early British opera-
tions were thus against both Acholi and Ethiopian gun-runners who were
operating in the territory west of Lake Turkana.22
Across northern Kenya, the British struggled to maintain order in a vast
area with relatively few resources, although in the years after the Anglo-
Ethiopian boundary delimitation of 1907 they expanded their military posts
from Moyale to Marsabit and other key points.23 Major operations in 1919
were followed by further incidents in the course of the early and mid-1920s.
In 1925, for example, a group of several hundred ‘raiders’, including Oromo
and Amhara, launched a major attack near Moite and took off with some
4000 camels.24 The success of this operation is noteworthy, given that the
frontier herdsmen who were the victims were by now long accustomed to
the sudden ambushes of highlanders from the north. They were, however,
vulnerable in their need to gravitate toward wells, and when they periodi-
cally clustered around markets. The situation in the area was exacerbated
further in the mid-1930s with the Italian invasion of Ethiopia, whereupon
thousands of refugees, armed opportunists, deserters, and assorted despera-
does flooded into the NFD from southern Ethiopia and Somalia.25 In this
zone, too—as on the Nilotic frontier—was group cohesion contingent on
local circumstance and external intrusion, and was political creativity height-
ened as a result.26 The dynamism of the frontier—often fuelled, at least in
the first instance, by survival strategies—in turn drove creativity at the cen-
tres of the state-building exercises, which (consciously or otherwise) defined
themselves increasingly by what was happening at their peripheries.
138 colon i al i sms, ol d and new

Gun-running was an expanding business in the early years of the twen-

tieth century in other areas, too, and highland Ethiopians were again the
pioneers. In the Upper Nile districts along the Sudanese-Ethiopian border,
a continual flow of firearms was facilitated by Ethiopian entrepreneurs into
the hands of the Nuer and Anuak, among others, in exchange for ivory.27
Indeed, as Douglas Johnson has demonstrated, Ethiopian–Nilotic relations
along the border between the 1890s and the 1930s were nuanced and finely
balanced, though also continually shifting; the Ethiopian ‘presence’ in the
area was mediated by local power structures, and indeed limited by the
‘highland distaste for the unruled and unruly lowlands, whether forest,
swamp or desert’.28 There was flexibility in terms of administration, and
continual ‘reinvention’ in terms of which side of the border particular
groups inhabited at a particular point in time, to whom they owed alle-
giance, and in terms of assimilating outsiders and outside influences. Cyclical
conflict between a range of groups in the area, moreover—not least between
Amhara and Oromo—meant that Nuer and Anuak could squeeze between
the cracks thus left in administration and carve out their own spaces along
the borderland.29 From the early 1900s, in other words, local groups along
the Ethiopian–Sudan border were already utilizing external intrusions and
adjacent state-building projects to their own advantage; thus were identities
continually made and remade in this especially fertile frontier.
At the same time, Ethiopian raids into Sudan remained a feature of fron-
tier life throughout the colonial period. In the northern border areas, around
Gedaref and Sennar, raids on Sudanese communities by Ethiopian renegades
and generically termed shifta west and north of Lake Tana were not infre-
quent, and reflected perhaps even older patterns in that area of essentially
privatised violence; and further south, on the Nilotic frontier, while some
Ethiopians traded guns, others attacked those same Sudanese communities
for slaves and ivory.30 In 1923–4, for example, Ethiopian attacks (as well as
those by Swahili poachers) along the Daga valley, across south-eastern Sudan,
and as far as northern Uganda from the direction of Maji and the Boma
plateau, prompted the convening of a conference in Kitgum to discuss
increased security in the region.31 Maji inside Ethiopia, and the adjacent
Mongalla province in the far south of Sudan, bordering Kenya and Uganda,
were regarded as beyond the reaches of any administration. Despite the
highlanders’ hatred for the lowland swamps and forests of the Nilotic fron-
tier, this was nonetheless a borderland of great opportunity for traders and
shifta alike.Violence and commerce went hand-in-hand in the places where
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 139

imperial projects, whether of the British or the Amhara, could scarcely

reach, or at least where hegemonies were transient. Further north, in the
Funj borderland west of Lake Tana, and in the Kassala district abutting the
Eritrean western lowlands, the activities of armed Ethiopian adventurers
were restricted by mobile border security units; but even here, throughout
the 1920s such activities persisted. Of particular concern to the British was
the highly organized slave trade which operated across the Funj–Lake-Tana
borderland, in which slaves were exchanged for firearms. Even in the early
1930s Haile Selassie’s officials were—or at least claimed to be—powerless to
prevent it.32 The victims included the Berta, part of a large swathe of popu-
lation in the eastern Sudan long known to be contemptible to the habesha
as shangalla. Along the entire stretch of the border, while there were many
local communities which were victims, others were able to take advantage
of frontier flux and grow wealthy on the proceeds of a commercial system
which was rather older than the colonial administrations which now tried
to suppress it.The western and southern economic borderlands were there-
fore lucrative but unstable, and have remained volatile frontier zones to the
present day. They have long sucked states into them: this has further desta-
bilized the frontier zones themselves, while the latter have served to pare
down state power to its essence, namely a restless, mobile militarism with
none of the accoutrements or even the rhetoric of civilizing modernity.
Vulnerable to predation and drought in equal measure, these were vortices
in which economic desperation and political opportunity interconnected.

From Wal Wal to Kagnew: war, reconstruction,

and reinvention
It was in this context, then, that the Italians moved to avenge Adwa; once
again the northern frontier was a source of national crisis. Ethiopia was
much weakened in military terms; opportunities for military development
had become rather fewer in the early twentieth century than in the second
half of the nineteenth, and the Ethiopian army was now scarcely a match for
Italian tanks and mustard gas. Western governments were prepared to do
precisely nothing, save for some half-hearted and ineffectual sanctions on
Italy imposed via the increasingly redundant League of Nations. Nonetheless,
it is the case that Haile Selassie and the Shoan political establishment were
14 0 colon i al i sms, ol d and new

able to use the painful experience of Fascist aggression to lend moral force
to their claims for recompense and protected sovereignty in the course of
the 1940s. Between the mid-1930s and the early 1950s, Ethiopia experi-
enced both trauma and triumph, however short-lived the latter would prove
to be; and after the nadir of foreign occupation came an aggressive reposi-
tioning and reinvention which constituted, in hindsight, something of a last
throw for the Solomonic state.The journey from Wal Wal—a remote water-
ing hole on the Ethiopian–Somali frontier, a name thrust into the notepads
of foreign correspondents as the point at which the Ethiopian–Italian war
began—to Kagnew, the US base on the northern plateau and thus a guar-
antee of Ethiopia’s place in the new world order, was a remarkable one
indeed.
It is not our purpose to describe in detail here the course of the war
itself.33 Suffice to say that tensions between Ethiopia and Italy had been
building since the late 1920s, notably over aggressive Italian plans for eco-
nomic concessions in Ethiopia; and in truth war was inevitable from the
beginning of the 1930s, when Il Duce set his sights on Ethiopia as vengeance
for Adwa and as part of his grandiose plans for the putative ‘new Roman
empire’. Predictably, perhaps, it was the undemarcated and volatile eastern
frontier zone which provided the Italians with the justification they required
for invasion: in late 1934, matters came to a head at Wal Wal, a small settle-
ment around some wells in the eastern stretches of the Ogaden, and an area
which was the point of heightening tension between Ethiopian and Italian
forces. The Italians believed firmly that Wal Wal was in Italian Somaliland;
they had been there several years, and the legality of their occupation of it
was—they said—beyond dispute. But the Ethiopians did indeed dispute it,
asserting that Wal Wal was very clearly in Ethiopia, in fact was some way
from the Italian Somaliland border. In December 1934, a clash between
Ethiopian and Italian troops there34 led inexorably toward a major confla-
gration—some futile attempts at negotiation notwithstanding—and in
October 1935 the Italians invaded Ethiopia in force.They crossed from both
Eritrea, where there had been a massive build-up of men and materiel, and
Somaliland.There would be no reprise of Adwa.With control of the air, and
massive superiority in terms of armour and firepower, and indeed often of
manpower, and despite several points of serious resistance by individual
Ethiopian commanders, the Italians swept through the north in late 1935
and early 1936. Campaigns in the south and in the Ogaden at times were
slower going and involved more Italian casualties, but the end result rarely
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 141

seemed in doubt.Winning key engagements, and then bombing and gassing

Ethiopian troops as they dispersed, the Italian forces exacted revenge for the
‘stain’ on Italy’s reputation inflicted forty years earlier. They entered Addis
Ababa in May 1936; but the emperor had already fled, heading to Europe to
admonish the League of Nations in a speech full of foreboding, and finally
settling in somewhat reduced circumstances in England. His decision to
leave Ethiopia was contentious; a number of senior figures had counselled
him to remain and lead the resistance burgeoning across parts of Ethiopia
beyond the lines of the Italian advance, resistance to which we return below.
Many would not forgive him for rejecting the advice, and for departing
with his entourage and his imperial dignity intact.
Various factors explain the Ethiopians’ swift defeat, weaponry and logis-
tics foremost among them. Menelik had enjoyed a great deal of parity with
the Italians in this respect; he was also the beneficiary of a restless and potent
militarism built up over several decades, and of armed forces with some
experience of repulsing foreign incursions. Haile Selassie’s Ethiopia, how-
ever, was a profoundly disunited polity, much weakened in political as well
as military terms; since Adwa, Ethiopian political culture had become no
less militarized, but it was a more insular, intensely localized militarism
which had grown up because of the Shoan regime’s aggressively centraliz-
ing tendencies. For all his ‘modernizing’ centralism—and indeed because of
it—the centrifugal forces which had long dogged highland state-building
exercises were once again rampant, and old fissures opened up across the
empire. The Italian invasion had exposed the fundamental fragility of
Menelik’s empire. The Italians were able to develop alliances or at least tacit
cooperation among a range of groups in habesha society. While the ‘patriot-
ism’ of many resisters to Italian rule has been much celebrated,35 the fact
remains that, as in other parts of Africa, foreign invasion and occupation was
characterized by complexity and ambiguity and above all pragmatism. The
Italians were assisted by political discontent within the unstable empire.
Corrado Zoli, former governor of Eritrea, wrote: ‘Some of [the feudal
chiefs]—e.g. the principal feudatory of eastern Tigre—not only spontane-
ously submitted to the Italians but faithfully fought throughout the cam-
paign under the Italian flag at the head of their troops’.36 It is certainly the
case that Zoli’s own policy from Asmara had been the winning of influence
among particular disgruntled figures within the habesha, and especially in
the Tigrayan, political establishment—a policy known as the politica periferica,
which clearly made some sense in the Tigrayan context at least.37 More
14 2 colon i al i sms, ol d and new

specifically, then, the Italian authorities in Eritrea pursued a politica tigrina—

encouraging provincialism in Tigray and luring key figures away from the
emperor—as opposed to the Italian Foreign Ministry’s prior politica scioana,
namely coexisting with Ethiopian centralization.38
On the face of it, at least, Haile Selassie’s increased centralized control
over key provinces had greatly strengthened the military capacity of the
centre; but in fact political and regional fissures swiftly re-emerged with the
Italian invasion, and the attendant military weakness of the state was exposed,
especially in the north. Chiefs and their troops in Begemeder, Gojjam,
Tigray, and Semien either jumped ship and joined the Italians—and were
often rewarded with political appointments—or fought half-heartedly for a
regime they scarcely trusted.39 Swathes of the Somali and Oromo popula-
tions had actively welcomed the Italian occupation.40 ‘The great majority of
the inhabitants in Western Tigre’, observed Zoli, ‘received the Italian troops
with sympathy, often with manifestations of joy, heralding them as liberators
from the dreaded yoke of Shoa’.41 Tigray is indeed an important case in
point.42 Politically marginalized since the 1890s, but still a crucial bulwark
between the Shoan state and the Eritrean frontier, Tigrayan elites—the rul-
ing families were essentially unchanged from the nineteenth century—were
always potentially problematic.Tigrayan leaders had been at least ambivalent
if not actively hostile to Tafari’s accession, watching the rebellion in
Begemeder in early 1930 with interest; following his enthronement, Haile
Selassie had sought to pacify the north by marrying his children to those of
the two key figures in Tigray, Seyoum and Gugsa, in western and eastern
Tigray respectively. The Tigrayan nobility remained bitterly opposed to
Shoa’s centralism, but the political marriages produced an uneasy balance of
power, both between Shoa and Tigray, and between Seyoum and Gugsa; but
the death in 1933 of both Gugsa and Haile Selassie’s daughter who had been
married to his son, Haile Selassie Gugsa, destroyed this delicately poised
arrangement. Haile Selassie Gugsa now sought to mount a challenge against
Seyoum by communicating with the Italians in Asmara, who were also in
contact with an array of more minor chiefs along the frontier. With the
invasion underway, Haile Selassie Gugsa crossed to the Italians in October
1935, following the fall of Adigrat. Seyoum opted for loyalty, and continued
fighting into the early months of 1936; but the potential role of Tigray as a
barrier against the Italian advance—both geographically and politically—
had been fatally undermined, and opened a crucial highway into the heart
of Ethiopia. In early 1936, moreover, in the wake of the repulsion of the
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 14 3

long-awaited Ethiopian counterattack, bedraggled and retreating Ethiopian

soldiers were attacked by the Raya and Azebo Oromo groups in the south-
east lowlands of Tigray, as revenge for punitive raids carried out on them a
few years earlier. As for Seyoum, after his surrender in July 1936, he spent
some time in prison, and in Italy, before returning in 1939 to collaborate
with the Italians. This forms the crucial prelude to the Tigrayan uprising
against the restored Haile Selassie in 1943, to which we return a little later.
The bulk of the studies of the Italian invasion and occupation of Ethiopia
have situated themselves very much in the context of interwar European
history—international relations, appeasement, and the rise of Fascism—and
have had little to say about the actual impact on Ethiopian society.Alessandro
Triulzi’s review article from almost thirty years ago, indeed, remains essen-
tially valid,43 although the work of Alberto Sbacchi and Haile Larebo’s
exhaustive study, for example, have gone some way to addressing the imbal-
ance.44 In terms of territorial reorganization, the Italians redrew the bound-
aries of the key ethnic regions, which reflected a massively simplified Italian
ethnographic wisdom, and which for a few short years (until 1941) provided
a glimpse of ‘what might have been’ had the battle of Adwa produced a dif-
ferent result. Eritrea was swollen dramatically to include Tigray and the bulk
of Afar territory; the Amhara and Harar zones formed the central districts,
along with Addis Ababa itself; the ‘Galla and Sidama’ territory broadly occu-
pied the present-day area of the Southern Peoples’ district; and ‘Somalia’ was
extended into the Ogaden. In a predictably pro-Italian assessment of the
new arrangement, Corrado Zoli described how the Italian victory had ‘lib-
erated’ the peoples of Ethiopia ‘not only from slavery but from all sorts of
servitudes and barbarous tyrannies’, and how a system of ‘direct rule’ was
necessary to avoid bolstering the feudal organization of the old Ethiopian
empire through the retention of ‘hordes of chieftains, great and small, who
used to attach themselves like leeches to the bodies and goods of the miser-
able native populations’.45 Indeed Zoli, after a fashion, paid tribute to the
modernizing and centralizing programme of the deposed Haile Selassie, and
to some extent his predecessor Menelik.The Italians, it seemed, were merely
finishing the job—albeit, of course, in a rather more efficient manner. But
Italy’s civilizing mission was otherwise attended by much the same rhetoric
as that of the age of high imperialism half a century earlier.The new admin-
istration was confronted by dramatic depopulation across swathes of for-
merly fertile land, the ‘direct result of the barbarous Shoan domination’.
Moreover, ‘[t]he abuses of the Shoan overlord—the continual wars, the
14 4 colon i al i sms, ol d and new

punitive expeditions and the large scale slave raids—have reduced this peo-
ple from prosperity to a primitive state of barbarism’.46
The violence of the new political order was inherent, of course; resist-
ance continued, as Zoli acknowledged, although he was predictably san-
guine about the ability of the Italian administration to crush it.47 In large
part, this was down to the new military restructuring put in place in the
months after Italian ‘victory’ in mid-1936, with the ‘colonial detachments’
composed of African soldiers under Italian officers (in addition to the
‘African detachments’, which were composed entirely of Italians). The sys-
tem was building on the success of earlier recruitment programmes among
Eritreans, Somali, and Libyans, native battalions of ascari who, ‘in addition to
their military value, form effective instruments of political penetration and
of civilization among the local populations’. Regional recruitment was
partly responsible for the evident success of these battalions, as was the
‘family-camp’ system,‘in which the soldiers live with their families while on
service’.48 The Italians undoubtedly reinforced the notion of soldiery as
representatives of the state, billeted on and among local populations, and
enjoyed a higher status as a result. It was the latest development in the con-
cept of military service as denoting a distinctive class—defined politically as
well as socio-economically and indeed culturally—and of its representing a
politico-military establishment which was ultimately about the provision of
authority and leadership and the inculcation of certain ‘values’. In the analy-
sis provided by Zoli, in fact, the numbers we are dealing with are actually
relatively modest, namely 40,000 ‘native’ troops in a total force of 65,000 for
‘imperial defence’.49 This would increase quite dramatically by the time
Italy entered the Second World War in June 1940.
The Italian presence unleashed a new wave of violence and unrest, a
significant proportion of which can be considered to have been continued
resistance in the wake of the emperor’s flight. Even Addis Ababa itself was
prone to sudden attack by marauding bands in the weeks that followed; it
was an edgy, nervous environment.50 Resistance was doubtless stimulated
over the longer term by the bloody reprisals carried out by the Italian
authorities following an assassination attempt on Viceroy Graziani in
February 1937; in Addis Ababa, the Fascist Blackshirts embarked on a frenzy
of killing, mutilation, and destruction.51 A great many intellectuals were
eliminated, and a number of nobles were also deported to Italy as a result.
In the provinces, shifta activity erupted in earnest, and very often, again, this
was opportunistic violence driven by largely criminal elements set loose by
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 145

the collapse of the political order. They were well-armed, highly mobile,
and an indication that—in many rural areas, at any rate—the Italians had
little prospect of establishing anything approaching military or administra-
tive control.52 Alongside this—and probably frequently overlapping with
it—there emerged the ‘Patriots’, groups of guerrilla fighters. Initial resist-
ance can basically be seen as an extension of the war itself, with conven-
tional tactics deployed under the leadership of the last few senior Shoan
figures still at large. One such was Ras Emeru Haile Selassie, who became
the focal point of resistance in 1936 in the west and south-west, inspiring
the newly formed Black Lion organization53 to follow his (increasingly dif-
fident) leadership. He surrendered to the Italians after failing to create a
stronghold in the south-west, and after his forces were confronted with
local Oromo hostility. But after the awful reprisals following the attempt on
Graziani’s life, resistance increasingly took the form of a guerrilla insur-
gency, especially in the northern provinces of Gojjam, Begemeder, and
Shoa, although groups of fighters were present across the empire. Shoa
boasted one of the most charismatic and significant resistance leaders in
Abeba Aregai, later confirmed as ras and Minister of War by the emperor in
the 1940s as a reward. The guerrillas attacked transport infrastructure and
government installations, developed an extensive network of intelligence
with regard to Italian strengths and weaknesses, and established cells in Addis
Ababa and the provincial towns, under the Italians’ very noses. In many
respects the Patriots’ resistance represented the harnessing of older patterns
of shiftanet, enlarged in scale and scope in order to confront the foreign
occupation, and across Ethiopia the Italian administration was both destabi-
lized and perpetually challenged.
Armed resistance, and the existence of such groups as the Black Lions in
exile, has contributed in some quarters to the notion that in times of national
crisis—the 1890s, the 1930s—Ethiopians were able to tap into an ancient
patriotism with which to confront external threats. It is, as we have seen in
the context of Adwa, a key component of national mythology and self-
image. The reality is rather more complicated. There were deep-seated and
frequently violent rivalries between guerrilla groups; conflict between them,
as well as with the Italians, was a defining characteristic of the armed resist-
ance of the mid- and late 1930s. A chronic lack of unity was reflected in the
absence of any overarching political programme or ideological direction,
moreover. This was parochial violence, concerned with local issues, and the
‘nation’ was notably absent from the discourse of the resistance. The fact
14 6 colon i al i sms, ol d and new

remains, too, that the Italians were able to create locally recruited counter-
insurgency forces—known as banda, a term which became synonymous
with collaboration—which were often more successful in their engage-
ments with the Patriots than the Italians themselves. Large numbers of
Ethiopians—Amhara and Tigrayan, as well as groups historically hostile to
the Shoan polity such as the Oromo—did indeed actively ‘cooperate’ with
the Italians whether politically or militarily, or at least reconciled themselves
to the new order as their emperor fled. Come ‘liberation’, there would be
room for both Patriots and banda in the restored political order, producing
a new set of tensions between those who had been on opposite sides during
the Italian occupation, but who now jostled for privilege and position in
the early 1940s. Most obviously, again, Ras Seyoum Mangesha of Tigray
worked closely with the Italians—in 1940 they made him ‘Prince of Shoa
and King of Tigray’—although he swiftly threw his support behind the
British in 1941; he was confirmed as governor of Tigray by Haile Selassie.
Blatta Ayela Gabre, for example, was placed in charge of the native courts
under the Italians; after 1941 he served as Minister of Justice. Tsahafi Taezaz
served the Italians in the administration of Addis Ababa from 1938 to 1941;
in 1942 he became vice-governor of Shoa. Other more minor figures served
both the Italians and then the restored emperor, Balachaw Yadate and Dawit
Ogbazghi (an Eritrean) among them.54
Haile Selassie and his entourage had been close behind the Allied advance
from Sudan, stepping onto Ethiopian soil in April and returning in great
pomp to Addis Ababa in May 1941. But it was a highly volatile situation,
with several dynamics interacting dangerously. The British were nervous
about the entry of a multitude of armed guerrillas into Addis Ababa, and
tried to limit numbers following the emperor (and, more importantly per-
haps, Abeba Aregai) into the city. Their fears were often well-grounded,—
witness the violent disorder which broke out when the Patriot leader Garasu
Duki was placed in charge of Jimma in 1941. Haile Selassie was only too
aware of the empire’s fragility. Nonetheless the emperor, recalled the British
colonial official Sir Philip Mitchell, ‘became suspicious of our intentions
and resentful of what he—sincerely, no doubt, but mistakenly—supposed to
be our designs on his full sovereignty’.55 While Haile Selassie was relieved to
have been restored, Ethiopia’s sovereignty was a compromised one, for the
British remained in control of key aspects of political administration and
military affairs.56 Eritrea could be treated straightforwardly as occupied
enemy territory, but Ethiopia was a rather odder case; nonetheless, one
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 147

European occupation had in effect been replaced by another one, and

British hegemony was a undeniable fact. Haile Selassie’s chagrin was palpa-
ble: for the second time in less than a century, British troops stood in the
central highlands. The emperor had cause to be suspicious not only of
British intentions, but also of his own nobility who had shown themselves
fragmented and in a great many cases less than faithful to him over the pre-
ceding five years. Many Ethiopians, indeed, believed he had abandoned
them by going into exile and that he should have remained to lead the
resistance. Meanwhile, the Amhara not only had the British to worry about;
the ‘empire’ itself was in trouble. Eritreans, Tigrayans, and Somali were
potentially problem peoples who resolutely did not share Haile Selassie’s
vision of the beneficent imperium—although once again the ambiguity of
some of these relationships was underlined by the fact that in the course of
the 1940s and 1950s thousands of Eritreans flooded to Addis Ababa in search
of socio-economic opportunities. The emperor would make good use of
some of them. The Oromo question would become ever more urgent,
meanwhile; but in fact it was Tigray which would be the first to explode
into open revolt, in the Woyane of 1943. It was the shape of things to come.
All in all, the Italian interlude had led to an escalation of the conflict between
centralizing autocracy, on the one hand, and restless centrifugal forces and
rebellious borderlands, on the other. Having said all this, the 1940s was also
a time of great opportunity for the restored Solomonic state. Haile Selassie’s
twin objectives in this period were the establishment of his authority over a
turbulent, fractious empire, and the pursuit of an aggressive foreign policy
and renewed programme of imperial expansion. Here was a chance to
embark on the first major imperial expansion since the late nineteenth cen-
tury: using a combination of wile, guile, threat, and manipulation, the
Ethiopian government went about laying claim to both Eritrea and the
Ogaden in what would amount to a second partition for the region.
The British initially made clear to Haile Selassie—in the first Anglo–
Ethiopian Agreement in 1942—that Ethiopia’s full sovereignty was very
much pending, and for the time being subordinate to the requirements of
the war effort. In mid-1941, notably, it was by no means certain—as Ethiopian
officials must have been well aware—that the Axis powers were never
returning to the region; after all, the battle for North Africa and the Suez
Canal was still in full flow, and Britain’s ability to hold Egypt was far from
clear. However, the second agreement, in 1944, gave Ethiopia rather more
independence, and the privileged position enjoyed by the British was largely
14 8 colon i al i sms, ol d and new

dismantled.The British Military Mission to Ethiopia was now answerable to

the Haile Selassie’s Ministry of War, and mandated to provide training for
the Ethiopian army.57 It was an awkward and ambiguous relationship, char-
acterized by mutual mistrust, but also mutual need.The success of the British
Mission in Ethiopia—which aimed at the stabilization of the region, and
resolutions for the issues of Eritrea, the Ogaden, and Tigray—very much
depended on Ethiopian cooperation; Haile Selassie still needed British
assistance in re-imposing himself on his fractious empire—not least in
putting down the revolt which erupted in Tigray in 1943.
The Tigrayan revolt, known as Woyane, had both deep roots and immedi-
ate causes—the former related to the province’s long-term neglect and
marginalization, and simmering discontent with Haile Selassie’s rule, the
latter connected to the turmoil of the late 1930s and early 1940s.58 It is, of
course, dangerous to search for the deeper roots of some form of distinct
Tigrayan nationalism, however tempting various actors find this exercise in
political modernity; it is clear, for example, that many early fighters in the
TPLF in the 1970s regarded themselves self-consciously as politically
descended from the ‘first’ Woyane.59 But the quest for nationalism and ulti-
mately nations, again, often distracts us from appreciating the real dynamism
of particular frontier zones—and this is as true of Tigray as it is, say, of
Eritrea or ‘Oromia’. The ‘first’ Woyane had its deeper roots in the shifting
political sands of the late nineteenth century—in Yohannes’ failure to resolve
the issue of the northern frontier, and in the emergence of Shoa as the
successor to Tigray at the apex of the remade Ethiopian empire. We have
noted the willingness of key leaders to work with the Italians in the late
1930s. The events of the early 1940s, therefore, must be seen in terms of
continuity from those years of fragility; the revolt represented, in a sense,
unfinished business. The Woyane of 1943 was the product—to borrow a
phrase once used of the French Revolution—of a ‘concatenation of events’,
a ‘relatively restricted, localized uprising with strong provincialist or ethnic
overtones’, as Gebru Tareke has summarized it.60 A disillusioned and embit-
tered local nobility, acutely aware of the gradual erosion of its local powers
at the hands of the Shoan-led monarchical state, was critical in providing a
leadership role to a peasantry weary of the exactions of the state, in terms of
brutal taxation and administrative inefficiency. Tigray had also long suffered
from environmental degradation and thus reduced resources in terms of
land and economic opportunity, further increasing the potential for popular
insurgency. As in Eritrea in the same period, once again environmental and
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 14 9

economic pressures drove violence—in different ways in different commu-

nities—while in Tigray the ghosts of remembered greatness and a supposed
imperial inheritance rendered those material traumas all the more intense.
A third element—the lowland groups in the south-east of Tigray, the Azebo
and Raya, Oromo in origin—likewise feared the encroachment of the state
on their egalitarian socio-economic system. As Gebru has pointed out,
although some common threads linking these groups are discernible, in
many respects the revolt was profoundly divided in terms of objective and
certainly lacked ideological coherence.61
The violent adhesive binding the disaffected together was the shifta activ-
ity which greatly heightened the potential for physical action. A number of
the key leaders of the rebellion were shifta, including Haile Mariam Redda;
and while these had long been active in Tigray, as elsewhere, in the early
1940s shifta activity increased dramatically—due to transitional weakness
between regimes, availability of weaponry in the wake of the Italian defeat,
and economic turmoil.As in Eritrea in the same period, shifta were co-opted
into political action and also took advantage of widespread turbulence.They
were from a range of backgrounds, again as in Eritrea: many were displaced
and landless peasants; others were demobilized banda or had deserted; others
still were a motley collection of socio-political malcontents. Haile Mariam
Redda provided inspirational leadership for a time, tapping into widespread
grievances using the language of political radicalism, and in May 1943 the
uprising began in earnest, with initial successes against small garrisons
around Mekele.Taken by surprise, it was several months before the Ethiopian
state reasserted itself—and only then by calling on the support of the British.
In the major counter-offensives of September and October, the rebel bases
were decimated by artillery fire and bombed from the air by British aircraft,62
with inevitably large numbers of civilian casualties.The uprising was crushed,
with heavy loss of life; in the end, only the nobility could claim much in the
way of success, as many were reinstated in the regional administration and
guaranteed certain local privileges. The Raya and Azebo arguably suffered
most in terms of the brutal suppression of the peasantry, enduring large-
scale land alienation and the quashing of their autonomy. Yet even the
Tigrayan nobility could hardly claim ‘victory’; in truth, the monarchical
state had—for the time being—asserted itself in the clearest possible man-
ner, and had in reality greatly curbed regional autonomy, that of both local
notables and the province of Tigray more generally. While the Woyane
was itself the reflection of several decades of accumulated and multifaceted
15 0 colon i al i sms, ol d and new

bitterness, the ‘second’ Woyane of the TPLF from the mid-1970s was made
possible by, and drew inspiration from, the widespread political violence and
its brutal suppression in the 1940s.
The involvement of shifta in the disturbances of 1943 was symptomatic
of the unstable borderland; bandit-rebels in Tigray were the product of
political disenchantment and anxiety, of economic marginality and upheaval,
just as they were in Eritrea. It is most likely, moreover, that in the wake of
the Woyane, many of the same shifta were involved in both periodic distur-
bances in Tigray, and in the insurgency in Eritrea; the continual complaint
from British officials in Asmara was that many of the shifta in Eritrea were
in fact Tigrayan, and frequently skipped back and forth across the border out
of reach of Ethiopian and British security personnel alike. The Ethiopian
government was either unable or unwilling to compel local notables and
their communities in Tigray to surrender shifta or refuse them succour.
Ethiopian forces lacked the reach to coerce in this way; but almost certainly
Ethiopian officials were unwilling to provoke local communities or the
shifta they were concealing in the wake of the Woyane of 1943, for fear of
igniting another. After all, Tigrayan shifta were pursuing an agenda broadly
in line with that of the Addis Ababa government—namely, the unification
of Eritrea with Ethiopia—which is of course striking, given the goals of
their own uprising; but for many Tigrayans it was never actually about the
illegitimacy of the emperor’s rule, but about the status and condition of Tigray
and its various constituent parts, and the unwelcome intrusions by the
monarchical state into Tigrayan political and economic life. It is not, there-
fore, wholly incompatible—indeed far from it—for many shifta to subse-
quently fight for the ‘return’ of their Tigrinya brothers across the Mereb,
thus both enlarging Tigrinya power in the north and removing the possibil-
ity of a hostile situation in a foreign-ruled or independent Eritrea. Besides
which, in rather more practical terms, Eritrea offered some easy pickings for
an opportunistic and well-armed band of warriors.
Having ‘pacified’ Tigray, Haile Selassie now moved ever more aggres-
sively to secure the Ogaden and Eritrea. In the meantime, he also manoeu-
vred Ethiopia increasingly toward the US at the expense of the British.This
was in part motivated by the fact that from an early stage in the war, the
latter were at least sceptical about Haile Selassie’s claim on Eritrea.63 He was
indeed adept at utilizing shifts in the global balance of power, sensing that
the US, not Britain, was his guarantor of future regional security.64 Thus he
projected himself a key ally of the West in terms of his hostility to both
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 151

communism and militant Islam, and offered his services—or more specifi-
cally, a crucial military base on the outskirts of Asmara, Kagnew station—in
return for military and other assistance. He sent a contingent of troops to
fight alongside the US in Korea to show his commitment to the cause.
Washington swiftly recognized that an independent Eritrea was a risky
unknown, a potentially destabilizing element in a volatile region rendered
vulnerable—in Cold War terms—by the imminent withdrawal of British
colonial rule and the rise of militant nationalisms across a broad arc of ter-
ritory from Kenya and Uganda to Egypt and the Middle East. No unneces-
sary gambles were to be taken with the Red Sea zone, and Ethiopia was the
perfect strategic ally. And so began a partnership—enduring until the
Marxist seizure of power in Ethiopia in the mid-1970s, and even that would
prove to be a mere interruption—which represented one of the US’s first
and enduring forays into African geopolitics. While Eritrea had been an
issue of international concern, however, the Ogaden was a question purely
for Addis Ababa and London, the British proving resistant to an Ethiopian
takeover of the Ogaden down to the late 1940s; however, after prolonged
discussion,65 and perceiving both long-term security issues and long-term
costs, the British handed over the bulk of the Ogaden to Ethiopian admin-
istration in 1948. The Haud and Reserved Areas followed in 1954. In many
respects, the process of expansion and partition begun by Menelik in the
1880s and 1890s was complete, or at least by the early 1950s, it had gone as
far as it was going to go, in territorial terms. Both Eritrean and Ogadeni
chalices were laced with toxin, however.
In the meantime, Haile Selassie created a veritable cult of personality and
built a coercive, repressive, and at times brutally violent regime as he sought
political consolidation, for the first time since the early 1930s. He created a
nominal ‘council’ of ministers—a continuation of the ministerial system
initiated in the early decades of the century—made up, by and large, of men
of low birth who were used to counterbalance the power of the provincial
nobility, the bane of centralizing monarchs over many centuries. As the
emperor became ever more concerned with foreign affairs, some ministers
did have authority in domestic matters, but never beyond the reach of the
royal prerogative.There was a creeping fear throughout the 1950s among the
educated elite and would-be reformers that Ethiopia was ossifying, and that
its political system appeared profoundly anachronistic alongside the momen-
tous political changes sweeping across Africa as decolonization proceeded
apace. The new Ethiopian constitution of 1955 certainly confirmed the
15 2 colon i al i sms, ol d and new

Emperor’s absolute power of decree, and appeared to sound the death knell
of liberal reform.66 Yet opposition continued to foment,67 and among its
early leaders, in fact, were implacable Patriot leaders from the late 1930s,
some of whom had long-standing grievances against Haile Selassie, not least
of which was the fact that the emperor had fled in 1936, and many of whom
resented the fact that banda were offered key positions after 1941. Among
these oppositional figures were Dejazmach Belay Zellaqa, in Gojjam, who
staged his own small-scale revolt in 1943; and Blatta Tekle Walda-Hawariat,
who made the journey from loyal servant to the emperor to one of his most
relentless antagonists between the late 1930s and the late 1960s. He was
involved in two plots against Haile Selassie in the early and mid-1940s;
released in 1954, he was brought into the administration, but rebelled again
many years later, in 1969, and was killed. More generally, across the empire
of the 1940s, shifta activity was widespread, and localized pools of anti-
Shoan opposition periodically appeared in regions hardly reconciled to the
‘new order’, such as in Wollo and Harar—disturbances in the latter involv-
ing Muslim activists perhaps taking inspiration from the activities of the
Somali Youth League further south.68 In the late 1940s, there was escalating
shifta violence in Tigray (as well as Eritrea), anti-taxation disturbances on
the part of Oromo communities north of Addis Ababa, and Somali raids on
Ethiopian military outposts in the Ogaden.69
More broadly, it is worth noting that between the major periods of
drought and famine in the 1880s–90s and 1970s–80s, environmental insecu-
rity also drove low-level violent instability, doubtless thrusting communities
into desperate straits and often outlawry. This was provincial volatility that
required a muscular state response; coercion, again, was critical to the func-
tioning of the modern Ethiopian empire-state owing to the episodic fragil-
ity of the political system’s environmental and economic foundations.
Famines appeared across the central and northern highlands in the middle
decades of the twentieth century—in 1927–8 (at a delicate point in the
internal balance of power in Ethiopia), 1934–5 (on the eve of the Italian
invasion), 1947–50 (as the Eritrean crisis was mounting), and in the late
1950s.70 In each of these periods, economic trauma was one of the key driv-
ers of entrepreneurial violence at the local level; and those entrepreneurs
either had political causes readily to hand, or did not have to search long to
find them.
Yet while storms of political violence were gathering on Ethiopia’s bor-
derlands, the main forces for change from within the country’s socio-political
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 153

system came from the ranks of the army and the young educated elite—
apparently disparate groups, but in the course of the 1960s coalescing into a
broad front for radical change. Ironically, as it turned out, one of the regime’s
‘modernizing’ policies concerned the army and the security forces from the
early 1940s onward. Although clearly one of the purposes of the reorganized
military apparatus—more so in Ethiopia than in many other African states—
was external defence, it is also clear that internal control was high on the
agenda of the military modernizers. The imperial bodyguard constituted an
elite force, while the British- and US-trained army formed the main fortifi-
cation around the state; the police force was organized along British lines,
while the potentially troublesome patriots and shifta were reformed into a
territorial militia.Yet tutored in the use of violence to achieve political goals,
and increasingly self-conscious in its role as guarantor of order and defender
of the state, the army in the 1950s and 1960s contained the seeds of the
regime’s own destruction. While rank-and-file soldiers inevitably shared
the concerns of the broad mass of the citizenry as socio-economic and polit-
ical conditions deteriorated in the 1960s, senior officers were drawn into
debates—increasingly radicalized—about Ethiopia’s political future.71 This
was parallel to developments in the higher education sector, where students
who were expected to benefit the imperial system in the long run came to
embrace radical change and challenge that same system.72 In this context, the
first clear cracks in the Solomonic edifice appeared in the latter half of 1960.73
While University students embarked on a series of public protests, in
December 1960 a small but fiercely motivated group of army officers in the
Imperial Guard organized a coup d’etat while Haile Selassie was in Brazil on
a state visit. It was disorganized, and lacked the full support of the military; in
subtle but important ways the US intervened to ensure the Emperor’s sur-
vival, refusing to endorse the coup and bringing pressure to bear on senior
Ethiopian officers. The attempted putsch was swiftly crushed, and many sen-
ior people were killed, including members of the large corps of middle-
ranking officials in government and a number of army officers; yet it was
only the beginning, and indeed it changed the very nature of opposition to
the imperial regime, ushering in a new era of broad-based and overt
protest.74
In terms of Haile Selassie’s twin objectives of domestic consolidation on
the one hand, and external expansionism and reinvention on the other, it
may be argued that in the short term he succeeded on both counts. He
overcame a series of internal challenges, including that of the reformers and
15 4 colon i al i sms, ol d and new

modernizers and an uprising in Tigray, and oversaw the 1955 constitution;

he acquired Eritrea and the Ogaden, and his alliance with the US brought
an unprecedented degree of international security. As we see below, he also
saw off the Somali threat during the 1963–4 war. Yet all these triumphs
were both hollow and short term. Over the longer term, forces of violent
change would prove too powerful for the Solomonic state to subdue, and
were already reasserting themselves in the course of the 1960s. There was
growing political radicalism in government circles, in the army, among stu-
dents and workers; there were ethnic, national, and class-based rebellions
breaking out across the empire, in Bale, Gojjam, and Eritrea, and the mak-
ings of revolt in Tigray and among the Oromo. But arguably the greatest
long-term force for change across the region—and Haile Selassie’s greatest
mistake—was in Eritrea, the frontier zone with transformative power on a
regional scale.

Fatal federation
A decade of scheming, cajoling and threatening had paid off, for now at
least. The UN-brokered compromise was a problematic one, however, as
was the entire unification project. The Orthodox Church had played a sig-
nificant role in effecting the ‘return’ of Eritrea to the ‘motherland’, yet there
were several hundred thousand Muslims in Eritrea who were at best ambiv-
alent about the empire to which they were now attached—not to mention
the hundreds of thousands of Christians who were likewise unsure about
just how much they shared in the way of language, culture, and outlook
with the Amhara and even Tigrayans to the south.This was the ambivalence
of the frontier writ large. Nonetheless, the Amhara acquisition of Eritrea in
the early 1950s was one of the great sleights of hand in modern African his-
tory, and one which would have long-term consequences. Decisions made
in the early 1950s regarding Eritrea would destabilize the region for the
next half-century, and continue to do so at the time of writing. The history
of the 1950s remains an area of considerable research potential, but the
basics of the story are well-known: Haile Selassie’s representatives in Eritrea
presided over the steady erosion of the territory’s federal autonomy, and
Eritrean rights and freedoms, in the course of the 1950s, and ever more mili-
tant movements swiftly emerged in response. Eritrean nationalism took a
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 155

decisive turn toward expounding unconditional independence, and many

who had tacitly or otherwise supported the federal compromise withdrew
their support for it once it became clear that Ethiopia had no intention of
honouring the arrangement. By the time the Eritrean Assembly was forcibly
abolished in 1962—under armed threat it voted itself out of existence—the
war had already begun. In the late 1950s, the Eritrean Liberation Movement
(ELM) had gone into exile and had begun the campaign for Eritrean inde-
pendence; it was soon supplanted by the Eritrean Liberation Front (ELF),
which in the early 1960s was beginning to organize violence against the
Ethiopian state. It did so initially by harnessing the shifta tradition in the
western lowlands; in the course of the 1960s shifta became guerrillas, and the
Eritrean war escalated rapidly.
From the point when Haile Selassie symbolically crossed the Mereb in
October 1952 on his first tour of the new province, it was clear that the
Eritrean federal apparatus was to be entirely subordinate to the imperial
executive.75 Eritrea had an elected assembly and a chief executive, the latter
position held initially by unionist leader Tedla Bairu. However, real power
lay in the person of the enderasse, the emperor’s representative, who was in
theory merely the link between the Eritrean assembly and the imperial
government but who by 1955 was making clear to the former that, as far as
Addis Ababa was concerned, there was to be no distinction between
Ethiopian and Eritrean affairs. Increasing interference led to Tedla’s resigna-
tion as chief executive in 1955, an important turning point in that it marked
the rapid disillusionment of those Christian highlanders who had been
‘unionist’ over the preceding decade but not, it seems, wholly uncondition-
ally. Tedla’s successor, notably, was Dejazmach Asfeha Woldemichael—no less
a person than the deputy to the emperor’s representative—and under him
the assembly became ever less ‘autonomous’ from the imperial government.
Political parties were abolished, and assembly members could only ‘petition’
the emperor as individuals; troublesome members were publicly vilified and
subject to the close scrutiny of state security. In the meantime, Ethiopian
army units were stationed across Eritrea, even if the police remained
Eritrean.
By 1955–6, organized, if often necessarily covert, opposition was begin-
ning to crystallize.76 Eritrean Muslims were already in the vanguard of nas-
cent nationalism, notably. In the course of the 1940s, an urban Muslim
middle class had tapped into the Tigre revolt in the rural west and north to
create a large umbrella of political activism which was, at the very least,
156 colon i al i sms, ol d and new

sceptical about the benefits of absorption into Ethiopia.77 Such scepticism

was also present among some Tigrinya highlanders, but it took repressive
imperial policies to galvanize a wider highland nationalism. As Ethiopia
dismantled political parties, outlawed independent associations, and sup-
pressed labour unions, it became clear, by the mid-1950s, that the Ethiopian
state actively sought the destruction of the autonomy supposedly guaran-
teed to Eritrea by the federal arrangement. Eritreans came to behold Haile
Selassie very differently from those outside the region who perceived a
benign, indeed heroic figure who was a champion of African unity and the
incarnation of ancient African civilization and dignity. Eritrea’s vibrant press
was soon muzzled; Arabic and Tigrinya were replaced by Amharic as the
language of government and public affairs. Eritrea’s flag was replaced by the
Ethiopian flag; its criminal code by the Ethiopian version; and imperial
signs and symbols appeared in Asmara and other towns across the territory,
while the influence of the Emperor’s representative over the Assembly
became pervasive and aggressive.78 Significantly, such influence led to the
resignation of the president of the assembly, Idris Muhammad Adam, in
1957; Idris Adam, an influential champion of autonomy and (significantly) a
key Beni Amer leader, had long clashed with chief executive Asfeha, and his
departure was as important a moment as Tedla Bairu’s resignation. Shortly
afterwards, he went into exile in Cairo.
Heightening Ethiopian repression was playing out against a backdrop of
severe economic problems across Eritrea—both urban and rural—which
had their roots in the slow economic decline of the 1940s. It is clearly no
coincidence that—just as in Tigray, where various forms of protest were
fuelled by economic marginalization—in Eritrea the foundational decades
of political activism were the 1940s to the 1970s, a period of chronic mate-
rial decline and seriously limited opportunities for the bulk of the popula-
tion, especially the young. This clearly connects with the broader point
already made that economic conditions frequently fuelled violence, in the
mid-twentieth century as at various junctures in the nineteenth. In 1958, for
example, the Eritrean highlands suffered a very poor harvest, and hardship
was acute in the months which followed.Violent, popular insurgency grew
up within a framework provided by economic deterioration. Nationalist
politics, arguably, had their roots in these economic challenges; certainly, as
Tom Killion has suggested, the Eritrean labour movement was in the van-
guard of protest in the mid- and late 1950s.79 Workers’ protests culminated
in the general strike of 1958 which was violently crushed by the Ethiopian
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 157

security forces; and although ostensibly the strike was concerned with
workers’ rights, in reality it represented the crystallization of political pro-
test against the undermining of Eritrean autonomy and civil rights more
broadly. As Markakis has pointed out, the crushing of the labour movement
effectively alienated a nascent Eritrean working class, the majority of whom
were Christian, thus pushing swathes of the latter into the anti-Ethiopia
camp.80 In the aftermath of the strike’s brutal suppression, all labour and
other associations were banned, and political protest began to take new
directions, under new forms of leadership. Although there was some overlap
between the labour and nationalist movements, political activism from the
late 1950s onward was led by nationalist figures in exile and members of the
student movement within Eritrea itself.
Eritrea itself was by now a dangerous place for open protest, and by
1957–8 Cairo was becoming the assembly point of Eritrean nationalism.81
In addition to Idris Adam, nationalist leaders such as Ibrahim Sultan and
Woldeab Woldemariam had fled into exile there, and were using radio
broadcasts to promote the cause. Eritrean leaders were able to take advan-
tage of the pan-Arabism sweeping the region, and Muslim leaders in par-
ticular were keen to tap the reservoir of sympathy for the Eritrean cause
across North Africa and the Middle East.82 This in itself would later cause
tensions between Christian and Muslim elements within the Eritrean strug-
gle; but in the absence of alternative audiences, certainly in Europe, Arab
nationalism represented a natural source of succour. This made sense, too,
given that the nationalist struggle was to begin in the Muslim western low-
lands of Eritrea, with the Christian highlands following later. Nonetheless,
there can be little doubt that Eritrean nationalism as a modern phenome-
non was the product of two converging streams, namely deep-seated Muslim
hostility to the Ethiopian state, now encouraged by pan-Arabist agendas
across the wider region, and Christian disengagement and disillusionment
from the Ethiopian state following the latter’s betrayal of the federal arrange-
ment. Meanwhile, in Asmara, the Ethiopian state moved ever more aggres-
sively toward total incorporation as a means to ‘resolving’ that instability. In
mid-1960 the Eritrean assembly was renamed the Eritrean ‘administration’,
explicitly under imperial authority; harassment of disruptive members was
intensified, and the full weight of the coercive security apparatus developed
in Ethiopia since the early 1940s was brought to bear on the territory.
Ultimately, under threat of armed intervention—the assembly buildings
were surrounded by Ethiopian army units—the assembly voted itself out of
158 colon i al i sms, ol d and new

existence in November 1962, whereupon Eritrea became the fourteenth

province of the Ethiopian empire. Federation was finally exposed as fiction;
and although, as Erlich makes clear, there is legalistic debate over the ‘right’
of the Eritrean assembly to vote itself out of existence,83 the fact remains
that Ethiopia’s abrogation of the Federal Act was an explicitly and unabash-
edly political manoeuvre, clearly signposted since 1941.
The thoughts of some nationalists had already begun to turn to armed
struggle in the late 1950s. In this context models and inspirations were on
hand in neighbouring Sudan, where a serious conflagration was in the mak-
ing which would have a major influence on events in Ethiopia and Eritrea.
Tensions had been building in southern Sudan in the first half of the 1950s;
a mutiny at Torit in 1955 was followed by some very low-level military
campaigns against those rebels still at large, but by the end of the decade the
oppressive and aggressive political and military strategies adopted by
Khartoum toward the south were pushing southern elements toward armed
struggle.84 Southern opposition either went into exile to form the Sudan
African Nationalist Union (SANU)—modelled on eastern and central
African nationalist movements—or joined the Anyanya, the name given to
the emerging guerrilla forces in the bush. Both SANU and Anyanya, and
events in southern Sudan more broadly between the mid-1950s and the
mid-1960s, were closely monitored by Eritrean nationalists.
Violence seemed now to present itself as the only alternative in combat-
ing Amhara incursions; but from the outset, the Eritrean struggle was char-
acterized by sectarian rivalries and divisions.85 In 1958, a group of young
Eritrean Muslims in Port Sudan formed the ELM. Inspired by Sudanese
independence a couple of years earlier, and by the organization and activity
of the Sudanese Communist Party, the ELM espoused religious unity as
well as armed struggle in the pursuit of independence from Ethiopia, and
began to organize clandestine cells across Eritrea. In the kebessa, the ELM
became known as the Mahber Shewate (‘association of seven’), as each cell
contained seven activists. Its appearance, however, was denounced by the
exiled leaders in Cairo, who had established the ELF; they rejected the ELM
as communist and (they claimed) a cover for the Ethiopian security forces.
As the ELM attempted political mobilization, it was outflanked by the
somewhat better-organized and better-funded ELF, which now tapped into
the growing unrest in the western lowlands—even if its claims to be leading
that unrest were far-fetched. By 1961, it had succeeded in recruiting a mot-
ley collection of shifta and former ascari in the western lowlands for the
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 15 9

purpose. Eritrean nationalist mythology has it that the first shots of the
thirty-year campaign were fired by the well-known Nara shifta Hamed
Idris Awate on 1 September 1961, when he and a small band of men attacked
a remote Ethiopian police post in the area of Haicota in Gash-Barka.While
it is clearly crucial for those involved in nationalist struggle to establish
clear chronological markers for the purpose of memorialization and
celebration—poignant punctuation in the narrative arc—the reality is that
the armed struggle had been fizzling into life since the late 1950s, and that
by 1961–2 two distinct streams of activity had converged, namely shiftanet
and nationalist ideology.
Like most armed insurgencies, it began inauspiciously, and almost imper-
ceptibly; initially, at least, the Ethiopian security forces would scarcely have
distinguished early attacks on remote outstations from the episodic actions
of ever-restless shifta in the Gash-Barka zone. But it was quickly clear that
there was more to this, and indeed the public squabbles between the ELM
and ELF drew attention to the emerging struggle itself. The ELF was soon
recruiting across a wide area, and the insurgency may even have prompted
the Ethiopian authorities to finally abolish the federation in late 1962 and
send forces into the territory in considerable strength. They were swiftly
emulating their British predecessors in patrolling the streets of Asmara,
targeting suspect rural communities across the territory, establishing intel-
ligence networks in zones of rebel activity, and chasing shifta wherever they
could be espied. Certainly, the Ethiopian authorities could hardly be in any
doubt that a serious revolt was brewing, even as the official line from Addis
Ababa was that these ruffians were mere shifta, perennial malcontents with-
out programme or higher aims. Most of them were indeed shifta but they
were also rather more than that: these bedraggled bandits on horse- and
camel-back with Second World War rifles, ranging across the hot plains in
the west and the turbulent mountains in the north, were the harbingers of
one of the bloodiest upheavals in the region’s modern history.

Crises impending
The empire that Menelik had built, and which Haile Selassie had striven to
maintain, was beset by a series of crises from 1960 onwards. While the
Emperor struggled to defend autocracy, notwithstanding some extremely
16 0 colon i al i sms, ol d and new

limited and half-hearted reforms, in the wake of the attempted coup d’etat in
1960, the decade saw increasing political radicalization among Amhara stu-
dents, workers, and army officers. Rebellion in the countryside—notably in
Bale, in 1968—and the escalating violence in Eritrea was sharply redolent
of the situation a century earlier, in the 1860s, although Haile Selassie was
no Tewodros. Meanwhile, while many Tigrinya continued to join the assault
on the imperial state—indeed some of them would aspire to lead the
attack—a complex relationship was suggested by the fact that many other
Tigrinya lived prosperous lives in Addis Ababa, and even served in the
imperial government. It is true that some of these would become politically
active in support of their Eritrean compatriots’ armed insurgency; but the
status of tens of thousands of Eritreans at the heart of the Ethiopian empire
was always ambiguous, and politically troubled, if economically secure.
There were heightened ethnic tensions elsewhere, too, in particular among
the Oromo whose political consciousness was beginning to stir in the early
and mid-1960s.The sum of these events was that by the early 1970s, Ethiopia
was crippled by ethnic, religious, and nationalist conflict—each of which
was driven by economic exclusion—and even if Eritreans (who were not,
in any case, an ‘ethnic’ group) and Oromo (who were) struggled to find
internal unity, there was enough in the way of militant identity to launch
serious assaults on the Solomonic state. Moreover, the war with Somalia in
1964 was a reminder of the violence that lay waiting just beyond the border
to the south; and all the while, radicalized Amhara in Addis Ababa and
beyond began to ponder whether Solomonism had not outlived its useful-
ness, and came to the conclusion that the problem lay not with provincial
violence per se, but in the structures of the Ethiopian empire itself, and in
the ideologies which underpinned them.
It was in Eritrea that the violent contradictions of the Ethiopian state
were writ large. Many of the early fighters in the ELF were recruited from
among the predominantly Muslim Tigre, seasonal farmers and herdsmen
inhabiting that vast arc of territory from the western lowlands, across the
northern mountains, and onto the northern coastal plains; in particular,
many were from the Beni Amer, Marya, and Mensa sub-groups of Tigre
speakers. Other early ELF fighters were Bilen, from the Bogos area around
Keren, and Nara, from the far western plains.86 Recruitment, or the lack of
it, elsewhere in Eritrea was often shaped by older, more local rivalries.
Among the Afar of the northern Danakil, for example, loyalties were divided
between Massawa and the Aussa sultanate in Ethiopia; only those of the
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 161

former persuasion provided recruits to both the ELF and, later, the EPLF.
The Kunama around Barentu had long been ambivalent, at the very least,
toward highland centres of power on both sides of the Mereb, and locally
had long been hostile toward the Nara and the Beni Amer; it meant that the
Kunama largely stayed aloof from the unfolding struggle, and some even
worked alongside Ethiopian authorities against the armed movements—
and, more recently, against the independent Eritrean state, which has conse-
quently come to regard them with deep suspicion.
Meanwhile the Eritrean nationalist cause was leant critical early support
from Khartoum, peeved at Ethiopia’s pro-Israeli stance and Addis Ababa’s
own tacit sympathy for the southern Sudanese rebels. In the course of the
1960s, strains between Sudan and Ethiopia created vital space—physically as
well as politically—within which Eritrean guerrillas could operate and con-
tinue to consolidate.87 Throughout the 1960s, the ELF expanded rapidly, its
recruits predominantly Muslim but increasingly from the kebessa, too; by the
mid-1960s it tapped into the radicalized and politicized youth of the
Christian highlands, mostly—though not exclusively—from the land- and
property-owning petit bourgeoisie of Asmara and the major towns, whose
children were politically active at school and even Haile Selassie I University
in Addis Ababa. The student movement, indeed, was increasingly important
in Asmara, especially, in terms of expressing dissent—more important, for a
time, than either the ELF or ELM themselves.88 Continued harsh economic
conditions in Eritrea throughout the 1960s pushed recruits into the ranks of
the guerrillas, although in truth the liberation forces themselves were
responsible for a fair amount of rural hardship as they not infrequently
preyed off rural communities’ food supplies. More common shifta strategies
of survival were adopted, too, including highway robbery and raids on plan-
tations and factories.89 Clearly, ‘civilians’—insofar as such a category had
ever existed in the region—were being caught up in this war, and certainly
the Ethiopian security forces’ response drew ever fewer distinctions between
combatant and non-combatant, especially in the rural areas.
Yet just as the armed liberation movement was growing, it was riddled
with internal tension and dissension. By 1965, notably, the ELF had effec-
tively crushed the attempts of the ELM to establish itself as an armed move-
ment in the north of Eritrea. Some inside the ELF ranks—especially young
Christian recruits—quickly came to despise its overly hierarchical, aristo-
cratic leadership structure, while Christians were frequently persecuted and
increasingly felt themselves at grave risk for questioning the direction and
16 2 colon i al i sms, ol d and new

organization of the movement.90 The ELF leadership also had a natural

leaning toward the Arab world: it sought succour from Egypt, Syria, and
Iraq, and took inspiration from—and to some extent modelled itself
on—the FLN in Algeria. This created a distinct unease among Christian
recruits who considered that they had not embraced revolutionary secular-
ism—or at least espoused it—only to see their leaders associating themselves
with Arabism or even Islamism. More specifically, there was disquiet among
Christian highlanders at the obvious connections being emphasized by the
ELF leadership between northern Eritrea and northern Sudan, connections
which again underpinned a set of Arabist, Islamic identities and loyalties.
The ELF was, for young radicals both Christian and Muslim, nowhere near
revolutionary enough for the demands of the age and in fact seemed to be
too much rooted in the feudal structures of northern and western Eritrea.
Many young ‘commissars’ returning from training in China or Syria, as well
as the young, educated, politically sophisticated class joining up from across
the kebessa, became highly critical of the supposedly ‘reactionary’ nature of
the organization—particularly as its showing against the first major Ethiopian
offensive in 1967 was poor.
The ELF has been much maligned in recent decades, the EPLF having
established—for the time being—a monopoly on ‘public truth’ with regard to
the story of the struggle. But the ELF was indeed a flawed and largely ineffec-
tive organization in the late 1960s.91 Its leadership was almost wholly in exile,
with the revolutionary command based in Kassala; it was structured according
to territorial zones, with initially four divisional commands treated as personal
fiefdoms by absentee commanders, with a fifth zone subsequently added cov-
ering the Christian highlands.The movement was thus fragmented and ill-led,
and fault lines swiftly opened up. Christian recruits were prone to be made
scapegoats for both military failure and internal dissent, as in the aftermath of
the 1967 offensive, when more than two dozen Christians were executed at
the orders of the Muslim deputy commander of the highland zone for failing
to adequately perform their duty. The commander himself was in Kassala at
the time. Other incidents—for example the killing of Christian fighters on
suspicion of their being Ethiopian agents—produced profound levels of dis-
enchantment and mistrust within the rank-and-file. A reform movement,
known as the eslah, gathered momentum, and culminated in 1968 in the abo-
lition of the zonal structure and the revolutionary command in Kassala. The
changes—notably the dismissal of the existing leadership structure—would
themselves engender long-standing rivalries and factionalism within the ELF,
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 16 3

even though the reforms represented the beginning of a wholesale restructur-

ing and indeed reinvention of the movement over the following decade. But
it was already too late. By 1970, the Eritrean struggle had splintered into sev-
eral groupings, a process of violent fragmentation that has come to define
Eritrean political discourse and the culture of ‘liberation’. Out of a cluster of
small bands emerged the EPLF, which would in time come to challenge the
ELF itself, as well as the Ethiopian state; the great northern borderland was
erupting in spectacular fashion, and it would wreak havoc on both Ethiopia
and its own inhabitants alike.
Nonetheless, despite these deep cleavages across the liberation move-
ment, Eritrean guerrilla forces had by the late 1960s and early 1970s begun
to carve out loosely defined ‘liberated’ zones. Operating largely on foot,
while using such supplementary transport as donkeys and camels, guerrilla
units were between twenty and sixty strong, according to contemporary
British reports,92 and thus highly mobile. They conducted hit-and-run
attacks on Ethiopian security targets and on selected non-military targets—
foreign-owned farms and factories, notably—and then sought to evade the
inevitable Ethiopian military response, all the while feeding off local com-
munities which were thus brought into the conflict whether they were
sympathetic or not. In many respects these were the classic tactics of the
long-established shifta. As the 1970s unfolded, they were increasingly effec-
tive tactics, too, not least because of the disarray and demoralization experi-
enced by the Ethiopian forces in light of political and economic upheavals
further south.
Of course in reality, Eritrea’s political instability—of which fractious
modern nationalism was the latest manifestation—can be dated back rather
further than the 1940s and 1950s.The modern liberation movement, broadly
defined, had grown out of the practice of shiftanet in Eritrea and the region
at large; the fronts themselves were essentially shifta in a modern setting,
enhanced by manifestos as well as military hardware, and their modern divi-
sions reflected the cultural and political and geographical fissures which
traversed the northern landscape in antiquity. Eritrea’s was a curious nation-
alism, therefore, rooted in the violent instability of the borderland, and rid-
dled with discord from the outset. In terms of splintered and brutally
internecine nature of the struggle, there is nothing in Eritrea’s past which
bodes particularly well for the future, which is precisely why the EPLF
would later become deeply wary of the past itself. As the ‘old’ encountered
the ‘new’—shifta morphing into guerrillas and revolutionaries—violent
164 colon i al i sms, ol d and new

chauvinisms would be born, and indeed certain forms of these were com-
paratively novel, notably the nationalist militancy of the EPLF itself.
Violent insurgency was fermenting elsewhere, too. Heightened Oromo
activism needs to be understood in the context of the ongoing Somali
problem for the imperial regime, for the two intersected at crucial fault lines
in the south and south-east. Notably, the problem of the Somali frontier in
the Ogaden was intensified with the independence of Somalia in 1960.93
The pan-Somali lobby of the 1940s and 1950s had sought the unification of
all Somali peoples scattered across northern Kenya, eastern Ethiopia, and
Italian, British, and French Somalilands. The project failed, although the
Republic of Somalia was an amalgamation—itself an unwieldy one—of the
British and Italian territories. The pan-Somali failure, and continued
Ethiopian occupation of the Ogaden—not to mention the ‘stranded’ Somalis
in northern Kenya—led to a reigniting of one of the region’s most ancient
zones of conflict. The Somali government provided support to insurgencies
in both eastern Ethiopia and northern Kenya. In the Ogaden, a low-level
guerrilla insurgency began after the Ethiopian state attempted to impose a
new tax on Somali pastoralists, and by the middle of 1963—mirroring events
in Eritrea to the north—Somali fighters were attacking police posts and
making gains against a poorly organized Ethiopian army. Confronted with
guerrillas armed by Mogadishu, Haile Selassie appealed to his American
benefactors for assistance, and at the beginning of 1964 Ethiopia moved a
US-equipped division into the Ogaden and launched a series of more vig-
orous counter-attacks. The insurgency swiftly disintegrated, and within
weeks Ethiopian troops had advanced to the Somali border, threatening
Mogadishu to cease its support for the rebels. Somalia negotiated a ceasefire
in March 1964, and although a standoff ensued, the Ethiopians had at least
won what they believed to be a measure of security on their eastern
flank.94
It would prove short-lived. Indeed just across the border in northern
Kenya, a Somali insurgency had also begun in the course of 1963. The so-
called ‘Shifta War’ was fought from 1963 to 1968 between Somali fighters
and the rather better equipped and trained Kenyan security forces.95 Recent
research suggests that many Somali shifta in northern Kenya were not par-
ticularly concerned about irredentist, Somali nationalist agendas, and that
much of the violence had a specifically local flavour and motivation;96
either way, again Mogadishu was compelled to retreat from its position of
support for the insurgency by 1967–8, and indeed the Somali government
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 165

appears to have lost some interest in the issue. But its retreat was doubtless
compelled in large part by the united front displayed by Jomo Kenyatta and
Haile Selassie, who had much to gain from cooperation on the issue of
frontier insurgencies. Ethiopia and Kenya had signed a defence pact in July
1963, while Haile Selassie later showed a great deal of interest in the situa-
tion in northern Kenya itself. On a state visit to Kenya, the Emperor visited
victims of the Shifta War and donated funds to their recovery. Notably, he
declared—in a somewhat backhanded expression of solidarity with
Kenyatta—that if anyone had a territorial claim to Kenya’s Northern
Frontier District, it was not Somalia but Ethiopia: after all, it was the natu-
ral home of the Boran of southern Ethiopia.97 Certainly Addis Ababa had
long viewed the Kenyan border as a potential weak spot in fortress Ethiopia’s
defences—the empire’s ‘soft underbelly’, as Mburu has it98—although of
course, as noted earlier, it was also a zone of opportunity for bands of vio-
lent entrepreneurs from the highlands. But above all, both the Somali issue
and indeed the Eritrean problem compelled a tough line from Addis Ababa
on the whole question of irredentism and secessionism—as it did from
Nairobi, and many others. While the Somali insurgencies were erupting in
the Ogaden and the NFD, and the Eritrean war was escalating, Haile
Selassie was one of the key architects of the OAU charter, adopted at the
inaugural conference (in Addis Ababa) in 1963, which stated that colonial
boundaries were inalienable and sacrosanct, and that resultant African sov-
ereign territories were inalterable. The somewhat awkward questions of
the Somali and the Eritreans (among other issues then emerging) were
swept aside, one of the hallmarks of the OAU’s conduct of business in the
decades to come; but while most of those present applauded the principle,
the Emperor, sitting astride a patchwork of violent fault lines, had more to
lose than most.
Meanwhile, in the province of Bale neighbouring the Ogaden, an Oromo
and Somali peasantry had become increasingly politicized as the result of an
array of impositions since the early 1950s.99 Land shortages owing to changes
in the tenure system, heavy taxation, the settlement of highlanders, and a
venal administration combined to create an explosive situation by the early
1960s, not dissimilar to that witnessed in Tigray a few years earlier, with
some important differences. In Bale, exploitation of the peasantry was much
more clearly the central cause of the revolt which erupted in 1963, while
Islam provided an ideological framework within which violence against the
state and its representatives could be carried out. The Somali, moreover,
16 6 colon i al i sms, ol d and new

could tap into the irredentist agenda then in currency, and indeed Somali
nationalists perceived Bale as part of the putative greater Somalia. As in
Tigray and Eritrea, the rebellion was at least initially driven by shifta who
had taken to the bush as economic outlaws as much as political bandits, and
it spread rapidly across the district in the mid-1960s. Initially, indeed, the
Ethiopian government attempted placation, but in 1966–7 launched a mas-
sive offensive—just as they were counterattacking the Eritrean rebels,
indeed—and the uprising was gradually, and bloodily, suppressed over the
ensuing three years. The tactics were similar to those employed in Eritrea,
too, with Ethiopian forces attacking rebel encampments and civilian targets
indiscriminately: as in the nineteenth century, and most markedly under
Tewodros, the key to the ‘winning’ of such wars was held to be in the visita-
tion of violence upon the entire community, not simply pitched combat
against men-at-arms.The uprising was finally crushed in early 1970, but not
before another revolt—this time in Gojjam—had begun, in 1968.100 Gojjam
was a rather different proposition, as a largely autonomous province through
the 1940s and 1950s, violently resistant to impositions from the direction of
central Shoan government. Indeed it was a newly aggressive Shoan admin-
istration (headed by the brother, in fact, of the governor of Bale) attempting
to bring various shifta groups to heel and impose a new tax regime that
sparked the uprising in 1968. It was crushed by the end of the year with a
brutality that was by now becoming characteristic of the imperial response
to such challenges—again, there was fairly indiscriminate killing of rebels
and their suspected but often unarmed supporters, as well as a dose of
scorched earth tactics—although in fact the government did subsequently
offer tax concessions and restructured the local administration to render it
more palatable.
Bale and Gojjam were the most prominent of the peasant revolts of the
period, but there were others—in Gedeo, in the south-west, and in Yajju in
Wollo. It is clear that such revolts were often class-based and economic at
root, but they also contained powerful ethnic and/or regionalist elements,
and the latter would become ever more prominent in the years to follow. It
is also the case that, just as in Eritrea and the Ogaden, they exploded along
fault lines of varying degrees of antiquity, zones of contest between ‘centre’
and ‘periphery’ which produced ever more militant and militarized identi-
ties.Yet whatever was lost or gained in the course of these uprisings, worse
was to follow for the peasantries themselves: the 1973 famine ravaged the
northern parts of Ethiopia in particular, and was an apocalyptic backdrop
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 167

to—and indeed a catalyst for—seismic shifts in the higher reaches of the

polity in the early and mid-1970s. Once again, major environmental
change—in this case compounded by chronic state neglect—drove an
upsurge in the direction and quality of violence; prolonged drought and
resultant hunger across the region from the early 1970s and into the 1980s
changed the very nature of the wars themselves. While hundreds of thou-
sands starved to death in the worst famine since the beginning of Menelik’s
reign, the body politic itself swayed feverishly, weakened by chronic mal-
nourishment of a more abstract kind; only massive ingestions of political
radicalism could save it, argued a younger generation of soldiers, students,
and workers. But while Marx might provide food for thought, it remained
to be seen whether revolutionary ideology would be sufficient to revive the
weakened corpus.
The abortive coup of 1960 represented the opening shots in the reasser-
tion of overt military power at the centre of Ethiopian politics. While the
novelty of an increasingly radicalized military in Ethiopia by the late 1960s
and early 1970s lay in its adoption of an array of modernist revolutionary
agendas, its readiness to intervene in politics in fact belonged to a rather
older tradition. Ceremony and symbol notwithstanding, Haile Selassie had
never been a soldier, and thus the bulk of his reign may be seen as somewhat
anomalous vis-à-vis the previous two centuries of habesha political develop-
ment. In the nineteenth century the business of government and military
activity were never distinct, and political leadership usually came about
through military command; rulers had proven themselves in battle, and
power was achieved and wielded through command of men and (increas-
ingly) guns. The state was no secluded concept to be protected by some
monolithic, ‘professional’ military establishment; it was, rather, a living
resource over which influence might be exerted and which might ulti-
mately be captured for particular ends. In other words, simply put, there was
a deep tradition of military involvement in politics in the Ethiopian region.
The 1960 coup attempt had been the first indication of the resurrection of
this tradition. The Ethiopian military from the late nineteenth century
onward, moreover, had been concerned with both internal control and
external adventurism, and indeed often the two were indelibly interlinked.
Soldiers were thus predisposed to intervention in political affairs, notwith-
standing the fact that a core component of Haile Selassie’s modernization
programme had been the creation of what we might loosely term a ‘Western’
military model of professional, apolitical detachment. The tensions between
16 8 colon i al i sms, ol d and new

this model and the Patriots of the 1930s, or the shifta in harness in Eritrea in
the 1940s, are clear enough; but in any case the professionalized Ethiopian
military establishment had an inbuilt propensity for action against elements
within state structures seen as detrimental to particular interests—whether
soldiers’ own, or those of a range of regional or ethnic groups, or indeed
those of the empire as a whole. And to be sure, several of the multiple crises
unfolding in the early 1970s, it was believed, required more decisive military
action. The move against Haile Selassie must be seen at least partly in this
context.
Meanwhile, in the Eritrean context, what is clear is the rejection by a
generation coming of age in the late 1950s and early 1960s of the unionist
sentiments of the generation preceding it. In what would become, in the
fullness of time, something of a generational conflict, the ‘new nationalism’
of the pivotal c.1958–c.1962 period was expounded by many whose parents
had been unionists, for one reason or another; the struggle for Eritrea now
took new forms. Even if the new nationalist struggle would carry along
with it many older Eritreans, men and women of the 1940s and 1950s, it was
the young who led them, and frequently the radicalized, idealistic young, of
the kind likewise discernible in the Ethiopia of the 1960s and early 1970s.101
In Eritrea, it was a process of radicalization which was made possible, at least
in part, by the brutal mishandling of the territory by an apparently ill-
informed and hubristic Amhara elite who had been offered an opportunity
to ‘secure’ the north but who failed to seize it. An institutional myopia pre-
vented the Emperor and his court from perceiving the political and cultural
gorges opening up which would come to swallow armies and state struc-
tures alike; driven by a desire to possess those lands, the state itself was
already being shaped by its frontiers. It would prove to be one of the most
significant political failures in the modern history of northeast Africa, and it
has given rise to a great many crises in its turn.
Ultimately, then, a more strident, more sharply defined Eritrean national-
ism came into being in the late 1950s; it was the product of several converg-
ing dynamics, some of which can be traced to the nineteenth century, and
others which were rather more recent creations, i.e. dating to between the
1900s and the 1950s. Quite how much ‘political awakening’ there was in the
1940s and 1950s is questionable, however, for the roads which led to the
ELM and the ELF have their starting points long before the BMA permit-
ted newspapers and lifted the ban on political parties. Moreover, the Eritrean
nationalist interpretation of the last fifty years has made much of external
 the e m p i re of ha i le se las s i e, C . 19 0 0 – 74 16 9

intervention and betrayal; Eritreans have laid great emphasis on the victim-
hood underpinning their experience.To be sure, it is a thesis which contains
some compelling arguments. But a great deal of caution needs to be exer-
cised here.The federation was not a catastrophe simply because the interna-
tional community sacrificed little Eritrea on the altar of ColdWar expediency;
it was a catastrophe because Ethiopia abused the federal constitution reck-
lessly, and because Eritrean unionists appear to have wholly misunderstood
the intentions of the Haile Selassie government.The ambiguous nationalism
of Tigrinya unionism, in particular, was to prove a flimsy defence against
Shoan ambition. Unionism, broadly defined, has had a long and robust his-
tory in the Eritrean highlands, down to the mid-1990s; but it has always
been conditional in one way or another, and successive Ethiopian political
establishments have consistently misread this, wilfully or otherwise. In any
case, the 1960s witnessed the maturation and the co-option of some long-
standing patterns of violence across the region; and around that violence
there now formed militant identities represented by groups whose modern-
ist programmes made stentorian claims for revolutionary change. The sym-
biotic relationship between ‘centre’ and ‘periphery’ was never clearer than in
the latter decades of the twentieth century, and it is to that which we turn,
now, in the final part of the book.
This page intentionally left blank
 PART
IV
Revolutions,
Liberations, and the
Ghosts of the Mesafint
This page intentionally left blank
 7
Revolution, ‘Liberation’, and
Militant Identity, 1974–91

The violent state renewed: the Derg

The regime that would become known in popular parlance as the Derg—
the deceptively innocuous Amharic word for ‘committee’—had its roots
deeply embedded in the violence and political settlements of the late nine-
teenth and early twentieth centuries, for all the revolutionary and modernist
language and symbol which surrounded it. Mengistu Haile Mariam, who
would quickly emerge as the head of state, bore some resemblance to
Tewodros a century earlier, at least superficially. In some respects, certainly,
the period between the mid-1970s and early 1990s represented the latest
stage in the deployment of overt brute force in the name of a larger, central-
ized Ethiopian unity, a process which had begun under Tewodros in the
1850s and 1860s. And yet it looked, if only for a brief moment, as though it
might have been very different. The revolution which began the process of
demolishing l’ancien regime came on the back of a new radicalized sense of
outrage across several walks of life in Ethiopia, and held out the promise, for
some, of the application of some measure of political idealism to Ethiopian
government and society in a rapidly changing world. That promise came to
nothing—at least in part because of the failure of Ethiopia’s own intellectu-
als to develop creative (and ‘authentic’) solutions to the country’s crises.1
Within a few years, the regime had become one of the bloodiest and most
authoritarian anywhere in the world.2
By 1973–4, imperial Ethiopia was confronted with the worst concatena-
tion of crises since the period following Menelik’s accession. Against a
background of escalating provincial rebellion and a devastating famine, a
series of strikes, demonstrations, and mutinies paralysed Ethiopia in the
174 revolut i on s, l i b e rat i on s, and th e g h o st s

course of 1974. Students, civil servants, and members of the security forces
formed unlikely—indeed often unwitting—collaborators in the revolution-
ary surge; but it was representatives of the military and police forces, who
formed the leadership of the committees, who were in the vanguard of
political change. In September, the emperor was deposed, and died a few
months later; Mengistu’s faction won through, and the Derg embarked on a
programme of Soviet-style socialism—or at least that was the stated inten-
tion. Opposition, whether within political circles in Addis Ababa, in the
rural areas, or in the form of the rebel movements in the north, was not
tolerated.
The Derg was an example—not uncommon across the continent in the
1970s and 1980s—of the so-called ‘radical military regime’ created in the
wake of army takeovers.3 The violence of the Mengistu state was overt, and
that heightened level of violence was a response to the crises confronting
the Ethiopian state—yet both the crises and the response had deep and
intertwined genealogies, and each fed off the other, militarizing political
culture and discourse to a dramatic degree, redolent of the nineteenth cen-
tury. Certainly, again, Mengistu had much in common with Tewodros, in
terms of his apparent belief in the primacy of armed force, and in his con-
viction that all enemies of Ethiopia could, and should, be violently crushed.
His roots—apparently in a Konso slave family in the south—were some-
what more obscure than those of Tewodros, who at least could claim some
kind of noble blood; but they shared a continual anxiety concerning ques-
tions over their legitimacy. Both came from the edges of highland Semitic
civilization—physically, in the case of Tewodros, and also culturally and eth-
nically in Mengistu’s case. The latter’s contemporaries sneered quietly that
he was not even ‘Ethiopian’—his Konso roots meant he was not truly
habesha—and that he was baria. In some respects, both might be held up as
exemplars of how the armed frontier produced those who aspired to the
capture and transformation of the centre. The hostile frontier was their
undoing in turn.
It is important to place the Derg in global context, too, for in this period
the region became the borderland in another, somewhat farther flung set of
contests.4 During the Haile Selassie era, and indeed for the first few months
of the new regime, the US had invested heavily in Ethiopia. From the early
1950s until the mid-1970s, the Americans had supported the Ethiopian army,
which became the largest and best-equipped in sub-Saharan Africa as a result,
and regarded Haile Selassie as a key regional ally in the Cold War.The USSR,
 m i l i tant i de nt i ty, 1974 – 91 175

meanwhile, saw Siad Barre’s Somalia as representing an African foothold, and

offered some measure of support and ideological encouragement to a regime
which presented itself as one of the most revolutionary in the continent, and
which was grateful for the material assistance.Yet within a few months in the
middle of the decade, the situation changed dramatically, a process of trans-
formation which also involved another major war between Ethiopia and
Somalia, in 1977–8. With the ascension of the Derg and the ideological rea-
lignment which resulted, the US was no longer welcome in Ethiopia, and its
facilities in Asmara were abandoned as the Carter administration took office;
the Soviet Union, meanwhile, irritated by Siad Barre’s conduct toward
Ethiopia in the run-up to the war and rather more impressed by the Mengistu
regime, withdrew its favour summarily from Mogadishu and redirected it
toward Addis Ababa.5 American policy toward the Horn was now character-
ized by ‘wait-and-see’; the Carter administration was inclined to become
involved in neither the affairs of the major states in the region nor in those
of the rebel movements. It would come to regret this.
The interjection of the Soviet Union had important consequences, as did
the rather less dramatic support provided by Castro’s Cuba. It greatly
strengthened the Ethiopian military and served to prolong the struggle in
the north, in particular; the EPLF, as we see below, was compelled to with-
draw from its dominant position across the territory in the face of con-
certed new attacks, and it arguably prolonged the war in Eritrea by a decade.
Yet it needs to be kept in mind that those wars were already being fought;
the exigencies of the Cold War sometimes exacerbated, and complicated, but
did not manufacture, patterns of violence across the region. Eritrean national-
ists would condemn—understandably—Soviet imperialism, and make much
political capital out of the failures of the international community; but it is
perhaps worth remembering that the Eritrean civil war did as much as the
intrusions of Moscow into the region to prolong the war in the north. In
any case, from 1974–5, it was clear that the new government was absolutely
intolerant of any suggestion that armed enemies might be negotiated with
rather than conquered. When Somalia invaded in 1977, Mengistu appealed
to ‘ancient Ethiopian patriotism’ in exhorting the populace to resist and
work for the victory of the motherland—with some success, it must be said;
and when the Somalis were indeed thwarted, it was promised that ‘victory
in the east shall be repeated in the north’.6 The murder of General Aman
Andom in late 1974, the one senior figure arguably capable of bringing
about a negotiated settlement in Eritrea, meant that the war—with or
17 6 revolut i on s, l i b e rat i on s, and th e g h o st s

without external assistance—would not simply continue, but increase in

intensity.
The militarization and sovietization of Ethiopia continued apace through
the late 1970s and beyond.7 Against a backdrop of escalating war in the
north, and inter-state war in the Ogaden with Somalia, the Derg arrested,
tortured, and killed thousands of suspected internal enemies in 1977 and
1978, in what became known as the ‘Red Terror’. Extreme violence was
deployed against ‘enemies of the revolution’;8 hundreds were killed or
injured when security forces fired at anti-government rallies in Addis Ababa
in April and May 1977, for example, and torture in detention camps and
extralegal killings were commonplace.9 As the Derg violently consolidated
its power, meanwhile, it developed a philosophy of self-consciously ‘indig-
enous’ or ‘organic’ Marxism aimed at mass mobilization; the philosophy, it
seems, was at least in part designed as a rebuttal to the ‘foreign’ Marxism
espoused by the student movement as well as a range of political parties,
while the mobilization it entailed was to be used to address a number of
crises, from rural development to foreign invasion. Underpinning it all was
the principle of Ethiopia Tikdem, or ‘Ethiopia First’.10 The Mengistu state
sought control through ‘revolution’, not the other way around. Moscow,
moreover, may have had some influence in the introduction of further con-
stitutional reforms through the 1980s, culminating in the proclamation of
Mengistu as President of the renamed People’s Democratic Republic of
Ethiopia in 1987.Yet it was the striking conjunction of political hubris and
environmental catastrophe that rendered comparisons of the Marxist-
Leninist regime with the imperial one that had preceded it most compel-
ling. Drought and famine were the defining phenomena of the 1970s and
1980s across the Sahel belt from Mauritania to the Ethiopian Highlands,
natural in provenance but indubitably exacerbated—and indeed utilized—
by human agency. Just as in the late 1880s and early 1890s, environmental
catastrophe was politicized: it galvanized liberation violence and indeed
compelled various ‘peoples’ fronts’ to search for ever more innovative organ-
izational principles and embrace ever less compromising military strategies.
By the same token, it drove centralizing, imperial violence and enabled the
regime to extend new levels of control over troublesome areas and popula-
tion groups, at least in the short term; entire communities were relocated
from the north to the south of Ethiopia in order to ‘protect’ them from the
ravages of famine, but clearly such strategies were aimed at stripping rebel
movements of their support base. From the mid-1980s, moreover, food
 m i l i tant i de nt i ty, 1974 – 91 177

aid—generated by well-meaning governments and other agencies in the

West—was routinely diverted from insurgent areas and was used by the
Ethiopian army itself.11 However in the longer term, arguably, such human
tragedy was the undoing of the Derg itself.
Reports of actual food shortages or of impending famine had begun to
filter in months earlier, but they were largely ignored; and when the full
force of the famine hit through 1984 and early 1985, shocking in its propor-
tions, the government was virtually powerless, owing to a lack of resources
and initiative within the system.12 Almost eight million people were affected;
around one million died. In eventual response to this crippling economic
and social failure, the government embarked on relocation and villagization
programmes which were bitterly resented by the millions of people involved.
Perhaps half a million people were forcibly moved from the drought-affected
and drought-prone areas in the north and centre to the comparatively more
fertile regions of the south and south-west. As a policy it contributed as
much as anything else to the massive human suffering experienced by
swathes of Eritrean and Ethiopian populations in what was a dreadful dec-
ade for the region; violent radicalism increased markedly as a result, moreo-
ver. It was, at the very least, controversial, and highly unpopular—as was the
decision to create new villages by uprooting and grouping communities
together, relocating them in such a way (so went the official rhetoric) as to
make the provision of aid and assistance easier to rural areas. This policy
began in the east and south-east, for example in Bale and Hararge, and was
thereafter spread across the country in the course of the 1980s.The state had
effectively used catastrophe for political ends, and it rendered millions of
ordinary Ethiopians bitterly opposed to central government, and susceptible
to the suggestions of advancing guerrilla organizations. Potential energy on
so many borderlands would soon become kinetic, and in many respects the
events of the 1980s heightened resistance to and suspicion of central gov-
ernment with which the current Ethiopian government is still confronted,
for all its supposed devolution of power.

Inter-state front lines

Before turning to the ‘internal’ frontiers which would ultimately consume
the Marxist state, it is important to note the inter-state borderlands most
17 8 revolut i on s, l i b e rat i on s, and th e g h o st s

important to the Ethiopian state in the 1970s and 1980s—those of Sudan

and Somalia. First, tensions were rife throughout the period between the
governments in Khartoum and Addis Ababa, thus involving the move-
ments which flitted across their frontiers. Open war was avoided, but proxy
war was ongoing.13 The escalation of the Sudanese civil war in the latter
half of the 1960s had major implications for the wider region, not least
because both Khartoum and Addis Ababa used the other’s internal wars to
enhance their own position as well as their regional standing—lessons in
brutal realpolitik which would not be lost on the liberation movements
who were alternately the victims and the beneficiaries of shifts in regional
politics.
From the mid- and late 1960s there was a marked degree of Sudanese
support and sympathy for the Eritrean struggle, and indeed—at least
initially—a measure of Sudanese influence in the early ELF. When relations
between Sudan and Ethiopia were difficult, Khartoum offered support for
the ELF—which could also draw on left-wing Sudanese political opinion, as
well as various Beja groups along the frontier with whom many members of
the movement shared familial and ethnic links. In this way, too, were frontiers
fertile, producing new forces or variations of old ones to challenge the state.
Meanwhile, Haile Selassie provided succour to Sudanese dissidents, in the
form of the Anyanya from the mid-1960s until 1972. In early 1972, the sign-
ing of the Addis Ababa Agreement between the Sudanese Government and
the Southern Sudan Liberation Movement brought the first civil war to an
end; but the regimes in Addis Ababa and Khartoum remained alert to oppor-
tunities to undermine one another.With the coming to power in Khartoum
of Jaafar Nimeiri in 1969, and then Mengistu in Addis Ababa in 1974, the
situation escalated.14 While in the late 1960s and early 1970s the level of sup-
port which the respective Eritrean and Sudanese rebel movements enjoyed
was very much contingent on the degree of frost in inter-state relations
between Khartoum and Addis Ababa, the hardening of political loyalties and
positions in the 1970s firmed up that support. With Nimeiri increasingly
anxious about Ethiopia’s (and Libya’s) pro-Moscow leanings, he increased
the flow of arms and support to the Eritrean movements—though particu-
larly the ELF, which was more evidently Muslim, and drawn, at least in ori-
gin, from among the groups which straddled the Eritrean–Sudanese
borderlands. He was a little less friendly toward the EPLF—highland, and
Christian, and Marxist after a fashion—but nonetheless both groups enjoyed
access to Port Sudan, and to offices in Khartoum itself.
 m i l i tant i de nt i ty, 1974 – 91 17 9

In retaliation, meanwhile, Mengistu increased Ethiopian support for the

SPLM in southern Sudan from 1983, especially as Nimeiri moved ever closer
into a US sphere of influence.The collapse of the Addis Ababa Agreement and
the beginning of the new phase of north–south Sudanese violence created
new opportunities for Sudanese–Ethiopian antagonism. Again, escalating civil
war in both countries in the early to mid-1980s rendered states themselves
incapable of extending firm control over their frontiers—and in those fron-
tiers, therefore, both the brutalities and weaknesses of the those states were
exposed, while the political cultures of emergent guerrilla movements were
hardened in battles against internal and external enemies alike. The internal
wars of Sudan and Ethiopia became indelibly intertwined. Mengistu permit-
ted the SPLM a base in Gambella region, as well as weapons, training, and
other equipment.The SPLM’s ability to move in and out of south Sudan was
greatly aided by the (falsified) documentation supplied by the Ethiopian gov-
ernment. It is clear, however, that Mengistu was in no position to encourage
secessionist movements in Sudan while attempting to crush those within
Ethiopia, and so Addis Ababa was allegedly instrumental in the formation
within the SPLM of a commitment to a New Sudan—i.e. a united one—
with John Garang emerging as the leader willing to pursue this.15
Other Ethiopian resistance movements—including the TPLF, the
Ethiopian People’s Revolutionary Party (EPRP), and the Ethiopian
Democratic Union (EDU)—also operated out of Sudan to a large extent,
often mobilizing local groups along ethnic lines on the Ethiopian–Sudanese
border, and thus creating, for Khartoum, a buffer zone between the two
regimes. The EDU, for example, preyed on the rootless groups in the bor-
derlands west of Lake Tana, making use of the shifta activities characteristic
of the region for much of the 1980s; in that sense, the EDU was the classic
regional political movement with distinctive historical characteristics, oper-
ating in the frontier zone and harnessing local patterns of resistance and
defiance. While the EPRP was largely vanquished by the TPLF in the early
1980s, a faction evolved into the Ethiopian People’s Democratic Movement
(EPDM) which likewise operated in the Ethiopian–Sudanese borderlands
in alliance with the TPLF, particularly the area of modern-day Benishangul-
Gumuz.16 On and behind the borderlands themselves, there was the swell-
ing population of Eritrean, as well as Tigrayan, refugees who in the course
of the 1970s and 1980s fled the fighting inside the territory and became
long-term residents of eastern Sudan, notably around Kassala and south
toward Gedaref.17
18 0 revolut i on s, l i b e rat i on s, and th e g h o st s

These ‘people in between’ found themselves squeezed between the

competing power blocs in north-east Africa whose own borderlands
became all the more violently volatile as a result. These regions had long
been zones of contest between political systems either side of the Blue
Nile; in the 1970s and 1980s, the level of politicized violence increased
sharply, during which time the military complexity on the ground likewise
heightened considerably.18 Khartoum lent some support—albeit limited,
given the enormity of its struggle with the SPLM—to a range of groups in
the southern Blue Nile and Upper Nile borderlands who were broadly in
arms against the Mengistu state; the EPRP, in particular, was at least tacitly
encouraged in its attempts to mobilize the Berta and Anuak (or Anywaa)
peoples against the Derg. In an ominous portent, meanwhile, the EPLF and
the TPLF also involved themselves in the affairs of the western frontier in
the course of the 1980s, attempting to coordinate patterns of local resist-
ance and mould these into cogent movements for armed liberation on the
model of the Eritrean and Tigrayan organizations. The outcome of these
combined efforts—not to mention the extant wellspring of centrifugal
hostility—was the Berta and Gambella liberation movements, notably the
Gambella People’s Liberation Movement, largely Anuak in composition.
For its part, the Derg, in conjunction with the SPLM, likewise armed
groups of local militia, including the Anuak in Gambella and the Berta in
Benishangul; and thus were new armed front lines opened up along the
Ethiopian–Sudanese borderland, front lines which were arguably critical in
the region’s Cold War struggle.19 The arming of local communities
undoubtedly fuelled extant patterns of violence and retaliation, hardened
causes for which people were willing to fight, and contributed to the
destabilization of an already-complex and volatile frontier zone. In a pat-
tern dating to the nineteenth century, and intensifying in the course of the
twentieth, lines between the ‘civilian’ and the ‘military’ were blurred and
recognized by neither guerrilla nor government.
In particular, these regions were transformed from being important
zones of commercial and cultural interaction, notwithstanding the peri-
odic instability associated with the eastern Sudanese marches, to being
armed frontiers into which the Ethiopian state now aggressively involved
itself. Communities were pulled this way and that, caught in the crossfire
of a multifaceted struggle between Addis Ababa and Khartoum, and
between those regimes and a range of ‘internal’ enemies.20 An additional
protagonist in the area—though not an especially effective one—was the
 m i l i tant i de nt i ty, 1974 – 91 181

Oromo Liberation Front (OLF), which briefly captured the Benishangul

town of Asosa in 1989, but which was more interested in the eastern thea-
tre of operations, as we see below. Between the mid-1980s and early 1990s,
shifts in the Addis-Ababa–Khartoum axis were dramatic, and had equally
dramatic consequences for the wider region. These are more closely
examined in the last chapter; suffice to say here that the collapse of
Nimeiri’s government in 1985, and the overthrow of Sadiq al-Mahdi’s
coalition in 1989 by Omar al-Beshir and the National Islamic Front, spelt
new directions in Sudanese foreign policy, while in 1991 the demolition
of the Derg signalled a new stage in Ethio–Sudanese relations. Sudan
lurched through a series of foreign policy disasters in the early 1990s, find-
ing itself at odds with Egypt, Eritrea, and Uganda; only with Ethiopia,
reconfiguring itself under the EPRDF, did relations improve markedly,21
thrusting independent Eritrea into the role of regional spoiler. The
change would have a major impact on the armed borderlands between
these states.
Along the frontier zones between Ethiopia and Somalia, meanwhile,
violence was episodically explosive, and in between, there was watchfulness
and tension. Inter-state war in our region was confined to these two coun-
tries until the Eritrean–Ethiopian war in 1998; the immediate prelude to
renewed conflict was the emergence of the pan-Somali movement in the
1940s and 1950s, and the clash in the Ogaden between Somali and Ethiopian
forces in 1963–4, which had ended with Ethiopian advances to the border
forcing Mogadishu into a ceasefire. By the mid-1970s, tensions were once
again mounting, this time between Mengistu and Siad Barre.22 The latter,
receiving military aid and other assistance from the Soviet Union, had been
steadily building up Somali forces and continued to agitate for the unifica-
tion of all Somali peoples; he was greatly emboldened by the fact that in the
mid-1970s, Ethiopia appeared in disarray, with the turbulence of the revolu-
tion and the escalating war in Eritrea sapping Addis Ababa’s traditional mili-
tary hegemony in the east and south-east. The Ogaden seemed ripe for the
taking, and even the transfer of the Soviet Union’s favour to Ethiopia was
not sufficient to dent Siad Barre’s confidence. He launched his invasion in
May and June 1977, the Somali army advancing in conjunction with the
Western Somali Liberation Front (WSLF) which was active across the
Ogaden.23 For the first few weeks, Siad Barre appeared vindicated in his
policy of aggression, as the Somali army swiftly overran much of the Ogaden
and the Ethiopians fell back in disarray.
18 2 revolut i on s, l i b e rat i on s, and th e g h o st s

The triumph was short-lived, however, and ultimately Somalia was

undone by both resurgent Ethiopian military might and diplomatic fail-
ure.24 The USSR massively stepped up its military assistance to the Mengistu
state, while the latter itself embarked on a dramatic mobilization and recruit-
ment campaign at home which sought to inspire Ethiopian patriotism
against naked foreign aggression. Once again, the Ethiopian state’s ability to
historicize the violence on its frontiers was a clear political advantage.
Further assistance and support came from Cuba and South Yemen—an
injection of military hardware and counsel that was to have massive implica-
tions for the war in the north, too. Somalia itself, meanwhile, was increas-
ingly isolated. Siad Barre had sought to win over the US to its cause, but the
Americans were not interested; Mogadishu also desperately sought interna-
tional recognition of its rights to the Ogaden, and had somewhat naive
hopes of persuading the OAU in particular. But the OAU—headquartered
in Addis Ababa, a clear disadvantage for the Somalis, as the Eritreans were
also learning—was wholly unwilling to recognize Somali claims, and clung,
with a doggedness born of terror, to the notion of the inviolability of colo-
nial boundaries. It was a notion, of course, which was written into the OAU
Charter in the year that the Somalis had last attempted to raise the issue of
the Ogaden before the August assembly, in 1963.
The Somali position, politically and militarily, quickly collapsed. The
Ethiopian re-conquest was launched in early 1978, with the army making
such rapid headway that by March the Ogaden had been largely recaptured.
The situation remained tense and highly volatile along the scarcely demar-
cated Somali–Ethiopian boundary.25 Siad Barre had attempted to resolve,
once and for all, the explosive question of identity and territory in the east-
ern Horn—a question which dated to at least the late nineteenth century,
and which had escalated slowly but surely from the middle decades of the
twentieth. Not only did he fail to resolve it, however—and it seems unlikely,
in any case, that it could be resolved by force alone—but the 1977–8 war
raised tensions even among the Somali themselves. Refugees fleeing into
northern Somalia from the Ogaden brought with them clan loyalties which
were hardly welcome among the Isaq of the north. Ogaden–Isaq rivalries
had long been a feature of trans-Somali relations; and in the rush to fulfil
the promise indicated by the five-pointed star of the Somali flag, extant ten-
sions within the wider Somali community had been heightened substan-
tially.The full implications of these tensions would become clear a little over
a decade later. In the meantime, the Ethiopians continued to monitor the
 m i l i tant i de nt i ty, 1974 – 91 183

situation across the contested frontier, and were able to turn their attentions
fully to the north, where they anticipated similar results. Addis Ababa and
Mogadishu reverted to the more familiar pattern of supporting one anoth-
er’s dissidents and armed rebels on either side of the Ogaden frontier, a pat-
tern only (in theory) ended by the peace accord signed between the two
governments in early 1988. But the collapse of Somalia by the beginning of
the 1990s only introduced new stresses into the contested zone.

Emerging markets of violence: ethnic

and nationalist borderlands
Wars of nationalist independence and ethnic regionalism escalated dramati-
cally in the course of the 1970s and 1980s.These wars were mainly in Eritrea
and Tigray in the north, and among the Oromo in the south and centre;
there was also a growing, if comparatively low-level, insurgency along the
western frontier with Sudan and in the southern marches of Ethiopia, as
well as political violence in the Ogaden which was part of a package of ten-
sions with Somalia across the border.The liberationist violence of the 1980s
was the historical Ethiopian state writ large: unable to control regionalism
or to channel emerging nationalisms, or to prevent the political radicaliza-
tion of the liberation movements themselves and swathes of the populations
these purported to represent, the Ethiopia built by Menelik and Haile
Selassie was being ripped apart.26 While the patterns of violence stretched
back to the eighteenth century, it was the centralizing brutality of the state
of the late nineteenth and early twentieth centuries that had sown the seeds
of the violence of the 1970s and 1980s. The Derg’s response was increasingly
brutal, its uncompromising stance manifest in the targeting of entire civilian
populations among which guerrillas moved.The Ethiopian security appara-
tus routinely arrested and ‘disappeared’ thousands of individuals suspected of
activities sympathetic toward a range of nationalist guerrilla movements, or
engaging in activities otherwise judged hostile to the state and its goals.
Urban populations and rural communities alike were targeted. Mary Dines,
one of the most assiduous chroniclers of war crimes in Eritrea, identified
numerous attacks by the military on civilian targets: MiG fighters bombing
the village of Dekidashim in August 1986, and then villages near the Barka
province in November; villages set alight by Ethiopian soldiers near Keren
18 4 revolut i on s, l i b e rat i on s, and th e g h o st s

in May and June 1986; the use of napalm on civilian targets.27 In Eritrea,
indeed, Mengistu and his lieutenants declared that they wanted ‘the land,
not the people’—the latter were expendable, indeed were actively loathed
for the troubled and ambiguous landscape which they inhabited. In a vari-
ation on a theme, Mengistu was wont to utter that he would ‘poison the sea
to kill the fish’—that he would visit total war on the Eritrean population in
order to rid Ethiopia of the despised shifta. It was a formula repeated across
the Ethiopian empire in the 1980s, as jails and torture instruments were as
important in the war against recalcitrant populations as AK-47s, tanks, rocket
launchers, and Russian fighter-bombers. Ethiopian centralism, meanwhile,
had ultimately rendered these conflicts unavoidable. The peoples’ wars that
represented the response on the part of the liberation movements became
ever more sophisticated and articulate, and ever better organized.The north-
ern frontier, in particular, would ultimately advance on Addis Ababa. This
violence was attended by a political and a scholarly redefinition of the
Ethiopian empire-state, both within and beyond the region itself; and while
the ghosts of the past returned to haunt the body politic, the new ‘marginal’
struggles of persecuted minorities became ever more visible, and the fron-
tiers of the nineteenth century closed in on the centre.
Across east and north-east Africa in this period, the state was under attack
from increasingly well-armed guerrilla forces, representing what we might
consider a ‘new wave’ of military intervention in African politics.28 The lat-
ter aimed not simply at the capture of the extant system—as previous coups
d’etat had done, for example—but often the complete destruction and radi-
cal remaking of that system, or at least this is what the rhetoric suggested. In
this sphere, if in no other, the influences and irruptions of the Cold War
were relevant, owing to the massive influx into the region of automatic
weapons and a range of other equipment; the continent more broadly was
swiftly awash with the ubiquitous AK-47. Such weaponry fuelled emerging
markets of violence, and enabled a range of populist movements to chal-
lenge the state to much greater effect than previously. Marginalized and
disaffected groups which in the 1950s and 1960s might resist the state using
Second World War rifles had access in the 1970s and 1980s to machine guns
and rocket launchers. The state-level monopoly on military might and the
exercise of ‘legitimate’ violence—in place in Africa since the 1890s and jeal-
ously guarded by colonial and post-colonial governments alike—was gone.
While there may be no direct correlation between the increase in access to
modern firepower and the coalescence of individuals and communities into
 m i l i tant i de nt i ty, 1974 – 91 185

peoples’ liberation fronts, there is little doubt that the latter—born of identi-
ties and grievances crystallizing in various ways since the nineteenth cen-
tury—now sought to dominate increasingly lucrative (politically speaking)
markets of violence for larger political ends.Therefore, guerrilla movements
sought, first, the firepower necessary to challenge the extant order; second,
the moral high ground, in terms of ideas about social revolution inspired by
ideologies which were as important a part of these movements’ armoury as
guns themselves; and, third, political power itself.
These movements were organized in different ways, and motivated by a
range of local factors and dynamics. Guerrilla themselves were ever more
professional fighters, even if they did not always look it, and their com-
manders were increasingly skilled in the military arts, both defence and
offence. Most organizations combined mobile guerrilla warfare, in which
by the 1980s fighters were well-versed—the writings of Mao Zedong in
particular were closely studied—with conventional pitched battle, although
some, notably the EPLF, were rather keener on the latter than others. Most
organizations made extensive use of auxiliary units and cells, notably ‘civil-
ian’ militias and cells of organizers and intelligence agents in the towns and
cities. Some were structured more explicitly than others in terms of ethno-
national identity, such as those in Tigray and among the Oromo, or at least
one ethnic group was dominant; others were overtly anti-colonial and
nationalist, as in Eritrea. Others still were, or became, tactical alliances of an
array of groups with a common goal. In terms of organization and ideologi-
cal orientation, many evoked and developed Leftist doctrine and extolled
the virtues of ‘people’s war’, which it was claimed was necessary in order to
accomplish true ‘liberation’, i.e. through popular social revolution. Guerrilla
leaders, with varying degrees of sincerity and pragmatism, utilized Marxist
rhetoric (and indeed Stalinist practice, in terms of internal control) in
exhorting populations to revolt; certainly, they needed to mobilize and
‘educate’ the peasantry and win their support, for no guerrilla movement
could survive, as Mao had asserted, without the sustenance of the people.
One dramatic manifestation of ethno-nationalism was among the Oromo
population across southern and central Ethiopia.Various Oromo communi-
ties have historically had markedly diverse experiences in the course of their
settlement in the area of present-day Ethiopia from the sixteenth century
onward, and their interaction with a range of different hosts has likewise
been complex. It is this very diversity and complexity which has led to
questions being asked about the validity of the ‘Oromo struggle’, and it was
18 6 revolut i on s, l i b e rat i on s, and th e g h o st s

certainly argued by the Derg that the whole concept of some kind of ethnic
‘homeland’ named Oromia was a nonsense.29 Clearly, Oromo history is not
straightforward—although few such large-scale migratory histories are. Part
of the problem was the sheer geographical spread of the Oromo themselves,
who therefore lacked the compact space within which struggle proceeded in
Eritrea or Tigray. However, it is clear that, in the most overarching manner,
the story in the twentieth century was of conquest, subordination, and mar-
ginalization. This in itself fuelled Oromo consciousness, as did the long his-
tory of racist derogation of the Oromo past and culture at the hands of the
Amhara (and indeed Tigrinya) political establishment.30 While an Oromo
identity was certainly emerging in the 1950s, as was a low-level resentment
of habesha domination, it was during the following decade that political con-
sciousness became rather more radicalized, owing to an increase (ironically,
no doubt) in educational opportunities for Oromo students, and the attend-
ant increased entry of Oromo into the workforce.31 Heightened levels of
education exposed Oromo youth to both casual Amhara disdain toward
them, and the means of redress—namely revolutionary nationalism—in
much the same way that a host of students of other ‘nationalities’ were being
radicalized in the same period. Increasingly aggressive ‘Amharization’ also
had its effect, with the Oromo language sidelined, cultural and political insti-
tutions abolished or undermined, and history denigrated or simply ignored.32
The fruits of the policy—in the form of a disillusioned and politically angry
younger generation of Oromo—were clear by the 1960s. The ‘new’ Oromo
nationalism first gained cogent expression through the Macha-Tulama Self-
Help Association, founded in 1963–4 with a view to improving welfare for
the Oromo population and in so doing mobilizing it.33
Following some initial optimism that the Derg would begin to address
Oromo grievances, Oromo activists turned their fire on the new regime
with the realization that, if anything, the Mengistu state would surpass the
old regime in brutality. The Oromo armed struggle was galvanized by a
newly aggressive approach toward the Oromo population. There was the
massacre at Chercher in eastern Ethiopia in August 1974, which was car-
ried out under the pretext of hunting down Oromo armed dissidents. A
thousand soldiers backed up by a tank company entered the area and killed
a local guerrilla leader and a number of women and children too. Then
there was the murder of two prominent Oromo, Colonel Hailu Regassa
and General Taddesse Birru, in early 1975, both men symbols of emergent
Oromo nationalism. A further massacre of Oromo civilians by Ethiopian
 m i l i tant i de nt i ty, 1974 – 91 18 7

security forces took place in April 1976 in several provinces, including

Kaffa, Illubabor, Arsi, Bale, and Hararge. The killing of several thousand
peasant farmers was a signal of intent on the part of the new regime
that—as in Eritrea—the state was prepared to take the war to the whole
community, not simply a handful of armed dissidents. It was a style of war
which was rooted in the nineteenth century. During the early part of the
1977–8 war with Somalia, Oromo farmers in Harerge were made the
scapegoats for initial Ethiopian setbacks, while hundreds of thousands of
Oromo were caught up in the villagization programme of the Mengistu
government. The internal displacement of Oromo was on a massive scale,
while swathes of Oromo land were in fact cleared to make way for north-
ern settlers in the course of the 1980s. Urban Oromo suspected of the
crime of ‘narrow nationalism’, meanwhile, were targeted by the security
forces.34 Altogether, as in Eritrea and Tigray, Derg strategies of control and
containment merely served to further radicalize significant proportions of
the Oromo population, and spurred recruitment into the OLF. While such
abstract ideas as Oromo unity and a singular Oromo nationhood might
invite debate, as they continue to, there could be little doubt about the
commitment and passion of a newly persecuted generation of outraged
activists, nor the violent brutality which had pushed it forward. In a sense,
the Oromo struggle has been all the greater considering the widespread
ignorance of their history and the hostility to their cause by those who
argued for the cohesiveness and unity of Ethiopia—both within and out-
side Ethiopia itself 35—and considering the relative lack, again, of a neatly
bordered arena of action.36 The notion of the borderland is less straightfor-
ward for the Oromo than for many other groups, owing to the ambiguity
of the Oromo position in the Ethiopian socio-political order. While many
Oromo did indeed inhabit the southern frontier zones, notably the Boran
groups along the Ethiopian–Kenyan border,37 a great many others, as we
have seen earlier in this book, had become integrated into highland society
and culture. This has proven singularly problematic for those who would
espouse a unified Oromo ‘nationalist’ cause, no less than the difficulties
facing those who had sought a singular, united Somali identity.
As for the OLF itself, success in the actual field of combat was limited—
although, according to Asafa Jalata, its mere survival was a matter for cele-
bration, given the challenges confronting it.38 Initially confined to the
eastern provinces, by the early 1980s the OLF had been divided into four
commands, namely eastern, south-eastern, central, and western ‘Oromia’.
188 revolut i on s, l i b e rat i on s, and th e g h o st s

Guerrillas operated in small units within those vast zones, and had some
success—Eritrean- and Tigrayan-style—in setting up local, elected adminis-
trative councils in the ‘controlled areas’. Also corresponding to the Eritrean
and Tigrayan models, the guerrilla forces themselves were offered succour
by the Oromo Relief Association, headquartered largely in Sudan, which
was in the main concerned with the massive refugee problem.39 But despite
some limited successes, the OLF list of military accomplishments is not a
lengthy one alongside those of the TPLF and EPLF—from whom the
movement periodically received assistance, moral and material—and in part,
no doubt, this reflects the fact that even for much of the 1980s Oromo
nationalism was not a mass movement.40 Arguably, the struggle to raise
awareness and mobilize was necessarily given precedence over the armed
struggle, which was somewhat more restricted as a result.

Tigray: revolution and renaissance

The lingering bitterness experienced in Tigray following the crushing of
the 1943 revolt hardened during the 1950s and 1960s as the province suf-
fered from political and economic marginalization. It gave rise to a process
of radicalization which would culminate in the emergence of the TPLF in
the mid-1970s. Tigrayan nationalists occupied a curious, traumatizing space
in the Ethiopian body politic from the early twentieth century: Tigray was
an ancient historical centre, and yet was now a periphery, neglected, abused,
and even despised by both Amhara and the ‘other Tigrinya’ on the Eritrean
side of the Mereb River. The heat of the issue is demonstrated by the writ-
ings of some early twentieth-century Ethiopian intellectuals, one of whom,
Afework Gebreyesus, in his biography of Menelik, represented a certain ele-
ment of Amhara opinion in heaping contempt on Tigray for their lawless-
ness and treachery at the time of Adwa. The Tigrayan writer Gebrehiwot
Baykadan had responded by condemning the impoverishment of the prov-
ince and the tragic migration of its people.41 The debates were no mere
academic trifle: they represented the very real tensions between Tigray and
the Amhara-centric empire, tensions which would become ever more
explosive as Haile Selassie’s reign progressed. These shifts in marginality had
a profound impact on local populations in real terms, as well as on aspirant
new political elites—such as that evident in the Tigrayan student movement
 m i l i tant i de nt i ty, 1974 – 91 18 9

of the late 1960s and early 1970s, and subsequently in the leadership of the
TPLF itself. The usurpation of Tigray’s leadership of Ethiopia in 1889–90
by the Shoan Menelik was within living memory in the 1960s and 1970s,
and clearly Yohannes IV—the last Tigrayan emperor—had iconic status even
among the radicalized younger generation of Tigrayan nationalists.42 Tigray
had been of crucial importance throughout the nineteenth century; in the
twentieth, it represented the classic fertile frontier, producing in time a level
of restorative militant violence which was quite distinct from that further
north in Eritrea, or among the Somali or Oromo further south.
At the same time, Tigrayan nationalism emerged against the backdrop
of—and indeed was in many ways closely linked to—another socio-
economic frontier, namely the migration of Tigrayans out of the province
in search of both work and education. The increasing poverty and neglect
of Tigray—both urban and rural—had compelled the brightest students to
travel to either Asmara University for their higher education, or southward
to Addis Ababa, to Haile Selassie I University.The political radicalism of the
student body in the latter campus is well-known; but Asmara University,
too, was something of a seedbed for Tigrayan student activism in the late
1960s and early 1970s, and would be instrumental in the formation of the
early nationalist movement.43 Meanwhile, poor rural Tigrayans travelled
increasingly to Eritrea to find work, and a Tigrayan labouring class became
a key feature of the mid-twentieth century Eritrean economy, just as an
Eritrean merchant class became a highly visible aspect of the urban socio-
economy of Addis Ababa in the same period. In Eritrea, these Tigrayans
acquired a reputation for thrift—money earned was sent back to villages in
Tigray—and in time for uncouthness and ignorance, as immigrant labour-
ers usually do. Eritreans came to regard them with contempt, and to haugh-
tily describe them as Agame—a reference to the impoverished eastern
district of Tigray from whence many (though by no means all) of the
immigrants hailed. Many also travelled to Addis Ababa.44 What may have
begun as a joking relationship was an increasing affront to Tigrayan sensi-
bilities. Tigrayans believed—with some justification, it must be said—that
Eritreans looked down on them, even despised them, and regarded them-
selves as vastly superior in every respect. It would come to the surface as the
rivalry between the EPLF and the TPLF developed in intensity—and would
return to haunt both communities with renewed war in 1998.
Almost everyone in north-east Africa in this period was always fighting
more than one war.The intensity of the rivalry between various movements
19 0 revolut i on s, l i b e rat i on s, and th e g h o st s

and organizations, the continual ideological positioning and repositioning,

the need for vigilance against the shifting agendas of neighbouring guerrilla
movements and distant states alike, meant that each political community
was compelled to fight on several fronts simultaneously.Tigray was certainly
caught between two fierce militancies—namely, the most immediate enemy
to the south in the form of the Derg, and the less overt but no less natural
antagonist in the form of the Eritrean nationalist movement to the north.
The challenge facing the Tigrayan nationalist front, broadly defined, was to
manage these surrounding threats and simultaneously to bring about the
social revolution headed, crucially, by the revolutionary party, that would
facilitate Tigray’s political resurgence.This sense of betrayal, marginalization,
and righteous, militant anger was no less keen in Tigray than it was in
Eritrea, although there were indeed significant differences between the two
in provenance and articulation; and those seeking the roots of the present-
day political impasse in Ethiopia, engendered by a recalcitrant TPLF ruling
class, might do no better than to appreciate the position of Tigray in the
early 1970s and the effort required to right perceived historic wrongs.
The TPLF was founded in February 1975 in western Tigray, in a barren
lowland borderland with, appropriately, a tradition of shiftanet.45 It quickly
developed a political programme, and sent recruits for training to the EPLF,
from whom it drew inspiration and support in its early years—although
only after it accepted the Eritrean case for secession. Even so relations with
the Eritrean fronts were strained from the outset, as we see in greater detail
below, not least because of the support given to the TPLF’s early, and initially
rather more powerful, rivals: the Marxist EPRP, which was the key move-
ment for revolutionary change in Ethiopia and particularly prominent in
eastern Tigray, received support from the EPLF in the late 1970s, while the
EDU, dominant in western Tigray, was favoured by the ELF. The political
environment was a dangerous one, as it was in Eritrea; the early TPLF was
ill-equipped, badly trained, and vulnerable to the predations of its better-
supplied and more prominent rivals. Its fighters were repeatedly worsted in
pitched battle with the Ethiopian army and the EPRP and EDU alike.Yet
it learnt quickly from its mistakes, and assiduously cultivated peasant sup-
port. It also benefited, in time, from the weaknesses and failings of the
EPRP and EDU. The TPLF programme of land reform contrasted favour-
ably with that of the EDU, which was led by the Tigrayan nobility; while its
espousal of Tigrayan nationalism was highly popular, the EDU suffered
from organization problems and by the beginning of the 1980s it was largely
 m i l i tant i de nt i ty, 1974 – 91 191

sidelined, inhabiting the borderlands outside Gondar. The EPRP, too, suf-
fered from divisions within the leadership, while the TPLF was again able to
outflank the EPRP in mobilizing the peasantry, and benefited from Derg
alienation of the priesthood, with many parish clergy becoming TPLF sym-
pathizers and activists. Increasingly confident militarily, the TPLF inflicted
serious defeats on the EPRP which was soon dispatched out of Tigray and,
effectively, out of the contest to dominant resistance to the Derg. Some ele-
ments later returned to participate in the EPRDF coalition.
Thus the TPLF grew up in a tough neighbourhood, and grew in stature
as a result.46 The movement tapped into both Tigrayan nationalist sensi-
bilities, and the need for social reform, especially in the realm of land
tenure. As in Eritrea, Derg policy fuelled sympathy for and recruitment
into the TPLF: increasing levels of taxation, the purchase of food at less
than market prices, and the forced resettlement of thousands of peasants
to the south (many were seized during the 1984–5 famine when they
came to feeding centres for help) all served to alienate Tigray from the
regime in the course of the 1980s and strengthen the TPLF’s hand. The
Relief Society of Tigray (REST) was the movement’s wing for humani-
tarian assistance and development; and in the early 1980s, recruitment into
the TPLF increased sharply from among the peasantry. Indeed, during the
1984 famine, the TPLF was able to project itself onto a global stage, and it
won significant attention for the first time. In terms of political organiza-
tion, the movement was arguably even more successful than the EPLF in
preventing the emergence of a personality cult by revolving the chair-
manship between two early leaders, Aregowie Berhe and Sebhat Nega; a
third key leader from the outset was Meles Zenawi. Collective decision
making was encouraged, although like the EPLF an elite party at the core
of the movement—the Marxist-Leninist League of Tigray—enhanced the
authoritarian centre and sought to ensure the ideological purity of the
movement. Nonetheless, it was pragmatic, and would later discard its
Marxist programme in favour of a markedly more liberal-capitalist mani-
festo; in a similar vein, early on it had moved away from expounding
Tigrayan secession to regarding itself as in the vanguard in the wider
Ethiopian struggle for democratic unity. But the TPLF’s policy positions
frequently brought it into sharp conflict with the EPLF.
By the late 1980s, the TPLF was a well-organized, battle-hardened
organization with mass support and a clear sense of political direction.47
In the course of 1988–9, it launched a series of new offensives across
19 2 revolut i on s, l i b e rat i on s, and th e g h o st s

Tigray—at Adwa, Axum, Endaselassie, Shire—against a backdrop of some

horrendous civilian casualties as the Derg brought its airpower to bear on
the province.48 But the momentum was now with the liberation forces,
and a unity agreement with the EPLF greatly enhanced their combined
military capacity in the north. In early 1989, the Derg forces in Tigray were
all but crushed, precipitating an abortive coup against Mengistu by several
of his senior commanders, but also raising a new problem in terms of the
prosecution of the war. Several thousand TPLF soldiers now returned to
their homes, believing that with the liberation of Tigray the job was done,
and unconvinced as to the need to carry the war further south. The EPLF
strongly disapproved, and so too did the TPLF leadership, which now
argued that only the total destruction of the Derg and the liberation of all
of Ethiopia (and Eritrea) would bring lasting peace and security to Tigray
itself.
In early 1991, the newly-formed EPRDF forces—including the TPLF,
the OLF, and some smaller groupings, and assisted by mechanized units
from the EPLF—launched the somewhat self-consciously and ironically
named ‘Operation Tewodros’ which cleared Gojjam and Gondar of Derg
forces. Although there had been some confusion in the respective Eritrean
and Tigrayan ranks as to the kind of war this had now become—namely, the
drive for the absolute destruction of the Derg and/or its unconditional sur-
render—the final grand operation was an indication of how far the process
of military maturation had come as far as the EPLF, the TPLF and their vari-
ous allies were concerned.This was now full-fledged conventional war, with
head-on pitched battles against Ethiopian forces and mopping-up exercises
in the rear. As the Derg retreated, civilian casualties continued to mount, for
as ever non-combatants were caught up in a war which was now spread
across a broad front; but as the 1980s drew to a close and the 1990s dawned,
in some areas at least there was something approaching a respite for the
broader population, for as the Derg collapsed, the war swept on into the
Ethiopian heartlands, leaving exhausted communities in the north to enjoy
a brief peace and ponder the uncertainties of the future. Meanwhile, the
advance on Addis Ababa continued, and as it did, the TPLF finally aban-
doned its Marxist-Leninist rhetoric for the language of liberal democracy
altogether more comforting to the US, now watching events with close
interest. Meles Zenawi, who was now the future, was courted by the US as
Mengistu prepared to flee, which he did on 21 May; a week later, the
EPRDF and the EPLF entered Addis Ababa.
 m i l i tant i de nt i ty, 1974 – 91 193

Eritrean epicentre
The Eritrean frontier was the epicentre for much of the violent conflict
across the region, fundamentally destabilizing the Ethiopian edifice in the
late twentieth century.The Eritrean region had been, in one way or another,
the crucial borderland zone since the early nineteenth century; now, the
armed liberation movement to which the frontier had given birth both led
the way in terms of ideology, nature of the struggle and organization—
many movements would model themselves on, or take inspiration from,
the EPLF, ultimately—and reached out beyond the Eritrean mountains
into others’ struggles, into Tigray and across southern and eastern Ethiopia,
and into the Ethiopian–Sudanese borderlands. The EPLF saw its war as
anti-colonial, defining Eritrea in terms of its colonial boundaries (a matter
not quite as straightforward as it was evidently believed to be, as future
events would show), and claiming for the territory a distinctive history.49
While the movement looked and talked like an Eritrean nationalist organi-
zation, the EPLF was also a manifestation of specifically Tigrinya militancy,
and the remarkable expression of a radicalized Tigrinya identity—notwith-
standing the fact that several other groups, including the Tigre and Saho,
had a presence within the organization. Nonetheless the Tigrinya claimed
to speak for the whole territory and laid out a very clear nationalist posi-
tion. Such separatism could never be quite as unambiguous in Tigray, but
the TPLF also toyed with Tigrayan separateness from Ethiopia, represent-
ing a new militant but awkwardly positioned nationalism. Even so, the
TPLF could appeal to a much deeper sense of Tigrinya/Tigrayan identity
than was possible in Eritrea, and it is certainly the case that those differing
visions—essentially an intra-Tigrinya contest—were the source of major
tensions between the two movements from the outset. The EPLF and the
TPLF had a troubled relationship which would spill out beyond the imme-
diate region, and beyond the 1980s; ultimately, the EPLF went its national-
ist way and the TPLF positioned the destiny of Tigray in an Ethiopian
context. It would need to do so in conjunction with a number of other
movements. But in fact it was the EPLF which would take on the role of
regional revolutionary vanguard, attempting to shape others’ agendas and
control the region-wide struggle.
It was at the Eritrean epicentre, moreover, that the violent contradictions
of the modern Ethiopian state were most clearly exposed: debates about
19 4 revolut i on s, l i b e rat i on s, and th e g h o st s

nationalism and nationhood, ethnicity and belonging, the past and the future,
were cacophonous and virulent in the Eritrean context, and have continued
to resonate. The contradictions would be inherited by the independent
Eritrean state. It was, and is, the frontier society writ large, with all its implica-
tions for political and military culture, social cohesion, and individual rights
and obligations. In such an environment it was not possible for cooperation
between movements to ever be more than short term and purely functional;
and in this respect two broad themes are important to note. The first is that
whatever ‘liberal’ political culture there had briefly been in the late 1940s and
early 1950s in Eritrea—or more precisely, Asmara—was dead within twenty
years, and has yet to be revived. By c.1970, political culture in Eritrea was
mercantilist and exclusive, and increasingly violently so; antagonisms were
between organizations and individuals, and were both ideological and per-
sonal. Increasingly, it was a political culture characterized by simultaneous and
successive attempts to monopolize ‘truth’, and while the purest of the goals
espoused by Eritrean nationalists of various hues—the creation of a free,
coherent, pluralistic nationhood—might be laudable, the struggle would
accrue such cost that some might wonder whether the fight had, after all,
been worth it. Certainly, the violence necessary to seize control of the fron-
tier and reconfigure the centre was ultimately inimical to the achievement of
those other goals—popular participation in politics, social justice, basic liber-
ties—in pursuit of which so many lent their labour to the struggle in the
1970s and 1980s. The second theme is the lack of capacity for genuine, long-
term collaboration and partnership between the EPLF and the TPLF. From
the early 1960s onward, the respective nationalisms of Eritrea and Tigray
were fundamentally on a collision course with one another; the forms they
would ultimately take rendered a deep-rooted alliance not simply impossible,
but actually inherently contradictory vis-à-vis Eritrean aims. The best that
could be hoped for—and for a time it was achieved—was arm’s-length, tacti-
cal, short-term cooperation in pursuit of the defeat of the common enemy,
without the removal of whom neither Eritrea’s nor Tigray’s aspirations could
be realized.This is true even taking into account early EPLF nurturing of the
TPLF—in any case held by Tigrayan leaders to have been grossly exaggerated
by the former. For a time, too, that cooperation appeared to offer some genu-
ine hope, for it accomplished a cease-fire in the long frontier war, between
1991 and 1998; but war did indeed resume, and the question should not
be—as it often has been—why did the two countries go to war?, but rather why
did they not go to war sooner?
 m i l i tant i de nt i ty, 1974 – 91 195

The 1970s was the decade in which the Eritrean liberation movement
came close to tearing itself apart, and certainly the territory witnessed a civil
war—as well as a pronounced degree of brutality within the movements
themselves—which remains a dark basement in the political structure of
independence.50 It has certainly had serious implications for plurality and
inclusivity in independent Eritrea. Nonetheless, given the frictions of the
1940s and 1950s, it was entirely predictable that the ELF would fragment.
With its roots in the shifta tradition of the Muslim western lowlands, and in
the pan-Arab oriented and largely aristocratic leadership in exile, the ELF
faced challenges from within, and fissures soon opened up between lowland
and highland; but these were rather less to do with Muslim-Christian ten-
sions—even if these were important at times—than with the distinct his-
torical experiences of metahit and kebessa, regions which in some respects
were fighting rather different wars for much of the 1970s and 1980s.51 These
were frontiers within the frontier, and although, for example, Nara and
Tigrinya might at length find common cause against the Ethiopian enemy,
their local visions of the violent frontier were informed by very distinct
historical experience. Fragmentation was also about the very nature of the
‘revolution’—as it often is—and was brought about by new recruits imbued
with the Marxist-Leninist leanings of the age confronting what they saw as
the reactionary or counter-revolutionary leadership and agenda of the ELF.
In a sense, the struggles within the Eritrean nationalist movement, culmi-
nating in the hegemony of the EPLF, thus represented the early Derg in
miniature.
The EPLF was essentially the product of the coming together of two
splinter groups which had broken away from the ELF in 1970.52 The first
was largely Muslim from the Massawa area but also comprised some
Christians, and gathered in the Danakil where the dissident ELF leader
Othman Saleh Sabbe arranged for them to be supplied with weaponry; the
group became known as the Popular Liberation Forces (PLF). The second
group was Christian Tigrinya, and fled from the ELF following the killing
of some Christians by an ELF commander. Gathering at Ala in Akele Guzay
province—it was known as the ‘Ala Group’—this small band was centred
on the young Isaias Afeworki, and in its statement, Our Struggle and its Goals,
it accused the ELF of Islamic sectarianism and failure to develop revolution-
ary leadership. Between 1970 and 1972, Ala Group and the PLF conducted
negotiations and slowly integrated; it was a process of unification which
culminated in a joint plan of action in October 1972, and the following year
19 6 revolut i on s, l i b e rat i on s, and th e g h o st s

in the creation of an integrated fighting force. For a time the ‘Obel Group’
of some Beni Amer fighters was also involved, but they later withdrew. The
fractious and violent political environment within which the EPLF was
born is crucial to understanding its subsequent development, ethos, and
nature. Even regardless of the Ethiopian military, the movement was imme-
diately confronted with the external threat of the ELF, and by internal dis-
sent. Civil war between the dissident movements and the ELF had erupted
by 1972, and continued sporadically throughout the 1970s, notwithstanding
a period of quasi-reconciliation in the mid-1970s.53 Most of the clashes
between the ELF and the EPLF took place in the mountainous Sahel region
in the north of Eritrea, and took various forms, from minor skirmishes
between small bands of fighters, to rather longer, pitched battles over several
days, in which fighters dug trenches and exchanged prolonged fire. Civil
war can be considered to have ceased by 1981, when an increasingly power-
ful EPLF finally succeeded in expelling the ELF from the field of combat,
into Sudan; thereafter, it became a beacon of opposition to EPLF hegemony,
both among the Eritrean refugee community in Sudan and among the por-
tions of the Eritrean diaspora in Europe and North America, although the
ELF itself would continue to splinter and reform in the years that followed.
The eventual triumph of the EPLF in the internecine struggle has come to
be understood by many—certainly many within Eritrea itself—as somehow
inevitable, as Eritrea’s unassailable destiny; it was not, for the ELF continued
to recruit even from among Tigrinya highlanders throughout the 1970s, and
the movement subjected itself to a fair amount of self-criticism and conse-
quently underwent significant internal reform. Clearly, however, the EPLF
was both politically and military more efficient—clinically so, indeed—
more disciplined, and ultimately more successful in terms of recruitment,
dissemination of propaganda, and social programmes. Nonetheless the
Eritrean civil war was a conflict that ripped communities apart and undoubt-
edly set the independence struggle itself back several years; the experience
of Eritrean killing Eritrean, moreover, an apparently bloodier sequel to the
violent divisions of the 1940s, confirmed the EPLF leadership in its belief
that absolute loyalty and commitment to a single political line was necessary
to the successful pursuit of sovereignty.The movement would kill to achieve
it; only through killing, and through the application of the severest disci-
pline, could the volatile frontier be brought under control. The experience
of the civil war further contributed to the idea that open debate was ulti-
mately a fruitless distraction, and that disagreement or divergence in opin-
 m i l i tant i de nt i ty, 1974 – 91 197

ion was tantamount to betrayal, a tendency that has persisted in Eritrean

political culture to the present day.
If conflict with the ELF persuaded the EPLF leadership of the need to
take Eritrea by force, likewise the internal crisis in 1973 convinced it of the
need for brutal discipline. A dissident group comprising a number of new
recruits, nicknamed manqa or ‘bat’ as it convened at night, emerged to criti-
cize the leadership for what it perceived as worryingly authoritarian ten-
dencies, as well as flaws in military strategy and organization more broadly.54
Isaias Afeworki himself was singled out for particular criticism. The leader-
ship moved decisively against its critics, with a number of people summarily
executed and others forced to publicly recant. The manqa crisis, like the
experience of civil war, was a formative episode in the EPLF’s development:
thereafter, debate was strictly controlled, and while it might occasionally be
initiated by the leadership itself, fighters were otherwise expected to obey
unquestioningly the decisions emanating from the central committee. By
the time of the First Congress in 1977, the structure and ethos of the EPLF
was largely in place, although political maturation was to attend the with-
drawal of the movement to its rear base in Sahel for much of the 1980s.
While the overthrow of the imperial regime in September 1974 gave
pause for thought—as it did among Oromo nationalists—it was swiftly clear
that the Derg was no more disposed toward a negotiated solution to the
Eritrean question than was Haile Selassie. Any hopes that Aman Andom
might have achieved a settlement were quickly dashed; his brief attempt at
winning hearts and minds in Asmara was followed by the violent retrench-
ment of the Mengistu regime.55 It is perhaps an open question whether the
Eritrean fronts would indeed have negotiated with Addis Ababa had they
been given the opportunity to do so; in any event, for Mengistu it was
beyond question.Yet in the mid-1970s the EPLF was more concerned with
the ELF than with the Ethiopian threat per se; political disarray at the centre
weakened the Ethiopian army, while the Somali invasion of the Ogaden in
1977 was a further boon to the liberation fronts. By the end of 1974, an
armistice between the ELF and the EPLF was followed by a series of attacks
on Asmara, and between them the two fronts controlled up to ninety per
cent of Eritrean territory, with the Ethiopian army beleaguered in the major
towns. The Ethiopian response was fierce, targeting Asmara, rounding up
civilians, and instigating something of a reign of terror in those parts of
Eritrea which it could still reach;56 Ethiopian security forces routinely car-
ried out extra-legal executions in order to instil fear in urban and rural
19 8 revolut i on s, l i b e rat i on s, and th e g h o st s

communities alike; thousands languished in prison, where torture was com-

mon and where women were exposed to the additional torment of sexual
assault.57 Derg policy drove recruits into the ranks of both fronts, while mili-
tary successes involved the capture of equipment from the Ethiopians—
arguably, captured ammunition, guns, and tanks constituted the EPLF’s
single most important supply of hardware throughout its struggle. The bal-
ance of military power shifted in 1977–8, however, and the period proved
debatably the most decisive in the history of the Eritrean armed struggle—
both in the prolongation of the war itself, and in the political implications.58
Having defeated Somalia, and now receiving massive military assistance
from the Soviet Union, the Ethiopian army turned its full and replenished
attention to the north, launching a series of offensives which threatened to
extinguish the nationalist movement altogether.
Inspired by Maoist teaching and recent Vietnamese experience in the arts
of prolonged liberation struggle, the EPLF put into action its long-planned
strategy of staged retreat to the rear base in Sahel.59 In what became known
as the ‘strategic withdrawal’, the EPLF pulled out of its positions across the
territory and converged on the mountainous Nakfa area in the north.60 It
was a watershed: militarily, the movement survived, and consolidated; politi-
cally, it enabled the EPLF to refine its organization and further develop its
visions and programme for government. While many at the time began to
write the movement off, in fact the struggle had entered a new phase; what
marred an otherwise brilliant military manoeuvre was a new outbreak of
fighting with the ELF at the end of the 1970s and early 1980s which resulted
in the expulsion of the latter into Sudan. In the Nakfa area, the EPLF con-
structed an elaborate and impregnable defensive structure, and withstood
repeated Ethiopian assaults.61
The EPLF was a remarkable organization, certainly among the most
remarkable movements for popular liberation anywhere in the world since
the Second World War.62 It prided itself on its revolutionary progressiveness,
its self-sufficiency, and its highly disciplined membership; it stood for the
unity in diversity of the various Eritrean peoples and cultures, and argued—
relatively simply—that the Eritrean nation, forged through the bitter expe-
rience of colonial rule (Italian, British, and Ethiopian), must wage an
anti-colonial war against Ethiopian occupation which was illegal and his-
torically unjustified. It was an argument, of course, which was summarily
and violently rejected by the Derg, which dismissed the northern rebels—
often lumping the TPLF in with the EPLF, much to the chagrin of the
 m i l i tant i de nt i ty, 1974 – 91 19 9

former—as mere bandits without a cause, or as bankrupt secessionists.63 But

even as the Mengistu state poured contempt on the very notion of Eritrean
secession, it was proving singularly difficult—and by the mid-1980s impos-
sible—to defeat the movement. Again, its major source of weaponry was
materiel seized from the Ethiopians themselves, although in their early period
the movement was supported by the People’s Democratic Republic of
Yemen; subsequently, however, the movement made much of its independ-
ence, and certainly from its entrenched position in Sahel it was free of reli-
ance on a host state. The EPLF’s strength lay in several key areas. Vigorous
recruitment programmes from the early and mid-1970s made the EPLF a
serious fighting force, notwithstanding the vast numerical supremacy con-
fronting it: estimates of frontline fighters through the late 1970s and early
1980s ranged between 30,000 and 50,000, but certainly by the eve of libera-
tion in 1991 there were some 100,000 men and women under arms. Ruthless
authoritarian leadership was scarcely challenged after the 1973 crisis, and an
unquestionably highly gifted group of people remained at the helm of the
movement politically and military until the achievement of independence.
The EPLF instigated ambitious and generally successful health and edu-
cation programmes in the liberated rural areas, and in the rear base around
Nakfa.64 The movement campaigned for political awareness—itself of lim-
ited scope, as events would prove—but more importantly for literacy and
rural development, for women’s rights and land reform. The rear base itself
was a striking achievement, comprising an underground world—hidden
from the daily predations of Ethiopian fighter bombers—of hospitals,
schools, and factories. A new generation was born and raised in the Sahel
rear base—the ‘Zero School’ children—who were either orphans or the
offspring of fighters. It was a veritable revolution in the mountains, and one
which subsequent mythologizing should not diminish. Cults of personality
were avoided—although everyone knew who was who—and much was
made of egalitarianism within the movement, and of relationships between
fighters based on respect and total loyalty. Meanwhile, the Eritrean Relief
Association (ERA) channelled aid to refugees and the rural poor, and
became a vibrant and impressive agency for humanitarian assistance; it was
especially important when Eritrea was struck by the famine of 1984–5.65
The EPLF was also adept at mobilizing the considerable Eritrean popula-
tion abroad, raising funds and awareness, and contributing to the diplomatic
lobby in Europe and North America. Although, increasingly, the ranks of
Eritrean fighters in the actual field of combat would come to regard them-
200 revolut i on s, l i b e rat i on s, and th e g h o st s

selves as something of a distinct, exclusive class—fiercely proud of a bond

forged in mortal danger—it is doubtful that the struggle would have suc-
ceeded in quite the way it did without the efforts of the Eritrean
diaspora.66
Above all, the EPLF was a state in waiting, and a remarkably tightly run,
disciplinarian state at that.67 The strategic withdrawal, necessitated by a wave
of new Soviet-backed Ethiopian offensives, arguably prolonged the armed
struggle by a decade; but it also facilitated the creation of the EPLF’s model
society around Nakfa, and enabled the movement to prepare, in effect, for
government—things would have been rather more chaotic, bloody, and
perhaps liberal had victory been achieved in the late 1970s. As it was, the
1980s witnessed the EPLF honing its political structure, eliminating its ene-
mies, and preparing to administer the Eritrean people: the political machine
built in Nakfa would then be transposed onto Asmara and Eritrean society
at large.The First Congress of the EPLF in 1977 had witnessed the comple-
tion of the unification process—Ramadan Muhammad Nur and Isaias
Afeworki were the key figures—as well as the publication of the Front’s first
‘national democratic programme’;68 in the years that followed, the move-
ment’s structure was further developed. It comprised, in summary, a clan-
destine party—the Eritrean People’s Revolutionary Party (EPRP), which
sat, concealed, at the centre of the movement, and whose existence was
unknown to the vast majority of fighters—the political cadres, and the
security apparatus. The Revolutionary Party developed strategy, dissemi-
nated policy from the leadership downward, and monitored the movement
as a whole; its activities were obscured from view until relatively recently.69
The cadres were highly trained activists, deployed throughout the move-
ment as commissars, organizers, and political education officers; the cadres
were responsible for the obligatory programmes of political ‘awareness’
through which all Eritreans who became involved with the Front were
passed. Internal security was concerned with internal and external threats:
in terms of the latter, it was concerned with the Ethiopians and the ELF, but
in terms of the former it maintained strict discipline within the Front and
essentially functioned as a military police force. In addition to these core
elements, there were popular organizations of women, workers, and stu-
dents, rural militias, secret urban cells in the major towns and cities, and
elected committees in the liberated areas.70
In the meantime, however, the Front’s other major conflict—and one
which would have long-term consequences—was with the TPLF across
 m i l i tant i de nt i ty, 1974 – 91 2 01

the border. I have argued elsewhere that the disputes between the two
movements in the late 1970s and sporadically through the 1980s sowed the
seeds for future violence;71 here, suffice to say that the multiple disagree-
ments were both old and new, and represented a series of political tremors
which presaged a catastrophic tectonic event which continues to destabi-
lize the wider region. Arguably the single most remarkable aspect of this
story is the degree to which it was overlooked in the literature for much
of the 1980s and 1990s.72 Even seasoned observers of either movement
apparently failed to pick it up, or if they did, regarded it as fairly unimpor-
tant.73 After the outbreak of war in 1998, naturally enough, there was a
rush to press of various analyses highlighting one or other of a series of
‘causes’ and dynamics. Some began to ‘notice’ the strained relations
between the EPLF and the TPLF during the armed struggle itself, and
more generally the history of tensions—as well as intimacy—which char-
acterized Tigray–Tigrinya cross-border relations.74 Others began to
‘remember’ those tensions rather more clearly, both publicly and privately,
and to recall incidents which now took on the characteristics of por-
tents.75 This was not merely a matter of journalistic and scholarly over-
sight; it is in fact part of the story, for the ability of the two fronts to
suppress their deep-rooted conflicts from the late 1980s onward was
remarkable, as was the apparent pact of forgetting which was as important
to the stability of Eritrea (and of course Ethiopia) in the early 1990s as was
silence over the civil war with the ELF. In a very real sense, the drama of
the two fronts’ respective triumphs in 1991 served—for a time—to con-
ceal both the true nature of the bloody civil wars which defined their
formative years, and the deep-seated conflicts between them once they
were hegemonic in their respective fields of combat.
The violent instability of the northern frontier zone, dating back to the
nineteenth century, would serve ultimately to produce militarized cultures
and identities, the ‘modern’ manifestation of which was the EPLF and the
TPLF. Violence, in sum, had become customary within political organiza-
tions seeking to inherit power across the region, in terms of both ‘internal’
affairs—the treatment of dissidence, notably, and the ordering of society—
and ‘external’ matters, in dealing with outside threats, whether real or imag-
ined. A robust, muscular approach to issues of import had become
characteristic of the region’s political culture, which itself had long been
shaped by force of arms. Broader populations, too, became, if not impervious
to, then certainly intimately familiar with, violence as an indelible aspect of
202 revolut i on s, l i b e rat i on s, and th e g h o st s

public life and therefore in some senses tolerated and accepted it to a marked
degree. Nor was this merely a matter of violence or the threat of it as a top–
down imposition: wider communities had long been willing to be galva-
nized into supporting and participating in violence against enemies old and
new, and, again, both real and imagined.Violent conflict, in other words, had
long become part and parcel of public political discourse, the most natural
part of basic human relations.This, therefore, is the broader psychosocial and
cultural context within which intense inter-organizational and inter-
community (and later international) violence in the Horn of Africa in the
later twentieth century must be understood.
But there was also a range of rather more novel dynamics which settled
on top of earlier layers of issues, converting the latter into the kind of human
fossil fuels which would continue to drive forward conflict as well as politi-
cal and cultural creativity. Broadly, we are concerned with the themes of
identity and territory. The EPLF did provide the early TPLF with military
assistance, but this has perhaps been exaggerated by the former and
downplayed by the latter.76 More important was the issue of the TPLF’s
objectives—and its definition of ‘Tigray’. Initially, the movement appeared
to lean toward a secessionist stance, which would involve the creation of a
‘people’s republic of Tigray’—later a source of some embarrassment—and
in 1975 it defined a Tigrayan as anyone who spoke Tigrinya, as well as such
border communities as the Irob and Afar.77 At the same time, however, it
claimed to support the Eritrean case, and prided itself on having changed
the ‘negative mindset’ of the Tigrayan people toward the Eritrean ques-
tion.78 This glaring ambiguity notwithstanding, it is clear that the EPLF
leadership regarded with suspicion the TPLF’s definition of ‘Tigrayan’,
incorporating as it did the highlands of Eritrea; it expressed concern over
this aggressive new form of Tigrayan nationalism with its reliance on such
alien (to Eritreans) historical reference points as Yohannes IV and Ras Alula.
The EPLF position had been that the TPLF must abandon talk of an inde-
pendent Tigray, and fight for a democratic, multi-national Ethiopia. The
EPLF argued, in essence, that Tigray had always been part of the ‘Ethiopian
empire’, dismissing the idea of an independent Tigrayan nationhood as eco-
nomically, socially, and historically ungrounded.79 In an apparent rebuttal,
the TPLF had declared that theirs (Tigray’s) was a truly ‘national’ struggle, as
all peoples within the territory were Tigrayan, but that that of the EPLF was
multi-national—for there were many nationalities in Eritrea, and the EPLF
should cooperate with other liberation movements within Eritrea.80
 m i l i tant i de nt i ty, 1974 – 91 2 03

Ultimately, the various nationalities of Eritrea—which was, after all, a wholly

artificial colonial creation—should have the right of secession.
The EPLF found this hypocritical and provocative. The EPLF position
was that while the right of secession might be granted to oppressed groups,
democratic unity would effectively remove the need for secession; the
EPLF was determined to put paid to any notion of Eritrean disunity, and
to place clear ideological distance between their struggle and that of
Tigray, which again should proceed in the pan-Ethiopian context. The
Tigrayan leadership saw this as a violation of the fundamental right of
peoples to self-determination.81 In sum, the TPLF was, at the very least,
raising questions about the very legitimacy of the Eritrean struggle, as
defined and led by the EPLF; the EPLF was dismissive of Tigrayan
nationalism, and was only prepared to support the TPLF struggle should
they become part of the larger war for Ethiopian democratic unity. Each
side was telling the other to do much the same thing, and each resented
it. Of course, the TPLF did soon abandon its secessionist position and
widen its vision to the liberation of Ethiopia; but tensions over the nation-
ality and identity questions remained, while the perceived arrogance and
aloofness of the EPLF merely further offended Tigrayan sensibilities and
fuelled resentment toward their would-be partners north of the Mereb.
The close linguistic and cultural—and, often, even familial—links between
the two Tigrinya communities notwithstanding, there were sharp differ-
ences on either side of the Mereb.
Some of these more recent differences were manifest in border disputes,
which can be dated to at least the mid-1970s. Early instances include the
claim by the ELF—then dominant in the west and south-west of Eritrea—
that the region up to Sheraro in lowland north-west Tigray was part of
Eritrea, with armed clashes with the TPLF breaking out in the Adiabo area
in 1976.82 Nevertheless the TPLF—then weak—‘permitted’ the ELF to
establish militias in the area and administer it for the time being, recognizing
that there were many ‘Eritreans’ in the region, itself a significant problem
when it came to aligning boundary with nationality.83 With the destruction
of the ELF by 1981, the TPLF brought much of the disputed region back
under its control, but in the early 1980s the question of borders was once
more high on the agenda—at least on the agenda of the TPLF, for the
Eritreans were apparently rather less interested in the issue, again much to
the annoyance of the TPLF. At the core of the issue, it seems, was the TPLF’s
desire for Tigray’s borders to be fully demarcated, which the EPLF interpreted
2 04 revolut i on s, l i b e rat i on s, and th e g h o st s

as Tigrayan territorial expansionism. In the mid-1980s, the EPLF issued a

statement:
In the period between 1979 and 1983, at different times the TPLF had caused
many serious problems and tensions on the question of boundaries.The EPLF
stated then that it was not the right time to raise such questions and that the
colonial boundary was clear. However, the TPLF [continued to raise] the
boundary issue by claiming the territory of the district of Badme, in the cen-
tre Tsorona, and in the south Bada, while the TPLF inhibited the EPLF’s
movement in and administration of those areas. Moreover, the TPLF [claimed
that] Tigray had an outlet to the outside world through Dankalia.84

The TPLF felt that clarification and demarcation was indeed needed, not
least as they felt that Tigrayan territory had been eroded during the reigns
of both Menelik and Haile Selassie. The Eritreans were dismissive: ‘Eritrea’s
legal boundary’, they declared, ‘is its colonial boundary’, and there could be
no talk of ‘any other fabricated boundary’.85
Frustrated, the TPLF shelved the issue, and indicated that ‘such problems
can be solved in the spirit of negotiation’, but warned that
[t]his does not mean that there is no need for negotiation concerning the
Eritrean–Ethiopian boundary ...The TPLF’s viewpoint is that Eritrea as a
nation was created during Italian colonial rule. Therefore, by official agree-
ment between the Ethiopian king Menelik II and the Italian government, the
boundary between Eritrea and Ethiopia was clearly demarcated ...However,
we cannot say that there will be no problems in implementing this Agreement
between Menelik II and the Italian government, because (1) for instance,
some places are clearly demarcated on the map but not clearly demarcated on
the ground, and (2) moreover, in the agreement ...some places are recognized
as belonging to either Eritrea or Ethiopia, but again the area may not be
found in the respective country as it is stated in the agreement.86

The EPLF’s silence on questions which the Tigrayans considered of vital

importance was taken as further evidence of Eritrean arrogance, and ulti-
mately of the EPLF refusal to treat the TPLF as an equal.87 What is certainly
clear from the exchanges between the late 1970s and early 1980s is that both
sides were very well aware of the serious problems over border demarcation.
That nothing was done is an indictment of both sets of leaders.
There were other problems, too, of course; again, we have already noted
the somewhat less tangible issue of Tigrayan sensitivity toward a perceived
Eritrean sense of superiority and hubris—which certainly could not have
been more clearly manifest than it was in the leadership of the EPLF.
 m i l i tant i de nt i ty, 1974 – 91 2 05

Tigrayans, maligned and marginalized, had much to be righteously indig-

nant about; the nationalism of the TPLF tapped into that wellspring of
resentment, and when they looked north they perceived a movement that
had disengaged from their immediate environment with a haughty brutality
which quietly angered many in the TPLF. The EPLF, of course, had its own
worries, and distrusted the Tigrayan front, its programme and (initially, at
least) its military weakness. Famously, moreover, the EPLF cut the TPLF’s
supply link through Eritrea to Kassala in Sudan at the height of their disa-
greements; the TPLF responded by organizing 100,000 peasants in the con-
struction of a new route linking western Tigray to Gedaref. The episode
would leave a lasting bitterness.The movements also disagreed over military
strategy.The TPLF prided itself on fighting a genuine ‘people’s war’, operat-
ing within the peasant communities whose liberation they sought, while
the EPLF—especially following the withdrawal to Sahel at the end of the
1970s—was accused of fighting a conventional, ‘bourgeois’ military cam-
paign, distant from ‘the people’. (Nonetheless, the TPLF did commit several
brigades to the EPLF’s defence of Sahel in the early 1980s, contributing to
the thwarting of the Red Star campaign.) Between 1985 and 1988 there was
a more or less total break in relations between the two fronts;88 but in 1988
tactical cooperation resumed, with both sides displaying their pragmatism in
the face of a faltering regime. Successes for the EPLF at Afabet in March
1988, and at Shire for the TPLF in February 1989, cleared the way, effec-
tively, for a final joint push toward Addis Ababa. But the problems which
had dogged relations between the two movements over the preceding dec-
ade were not forgotten, merely set aside, and close to hand.
Politically, the EPLF demonstrated its essential pragmatism at the 1987
Congress, at which Marxism-Leninism was largely abandoned in favour of
democratic institutions, pluralism, and market principles.89 Militarily, the
movement at first staunchly defended its position in Sahel, and then increas-
ingly took the war to the Derg, through a combination of guerrilla activity
behind enemy lines and conventional attacks from the north. Mengistu’s
1982 ‘Red Star’ offensive—the sixth major effort by the Ethiopian army—
came close to overwhelming the Eritrean positions, but they held firm;
from this point, gradually, the military momentum began to seep back to
the EPLF.90 In 1988, EPLF forces won a major battle at Afabet, south of
Nakfa, and finally broke out of their defensive ring, destroying a sizeable
Ethiopian force in the process.91 Famously, it was described by the scholar
and commentator Basil Davidson as the most significant defeat of a national
206 revolut i on s, l i b e rat i on s, and th e g h o st s

army by a guerrilla force since the French loss at Dien Bien Phu in 1954.92
Afabet was indeed crucial, as it effectively split the forces of the Derg in
Eritrea in two—those concentrated in the kebessa, and those now cut adrift
in the western lowlands. In the course of the following year, the EPLF
attacked through the eastern side of the territory, and captured Massawa in
early 1990, though at considerable cost.93 The armed frontier was now on
the march toward the centre.
In the late 1980s, recruitment into the EPLF increased dramatically—
including contingents of former ELF personnel who now detected the
unmistakable whiff of inevitable victory in the air, and who could not but
be impressed by the EPLF’s extraordinary achievement. Indeed, the political
capital of the movement at this time was inestimable, and that of Isaias
Afeworki particularly so; to his supporters and neutrals alike, there was
something of the quiet visionary about him, and certainly genius in the core
leadership. Even if some, with rather longer memories and first-hand expe-
rience of the Front’s attitude toward dissent, had reservations about the
imminent victory of the EPLF, they were, for a time, silenced by the sheer
magnitude of the accomplishment and the general exuberance which would
follow liberation from Ethiopia. The end came swiftly: with pockets of
demoralized Ethiopian forces encircled across Eritrea, and surrendering in
their droves, the EPLF finally made its entry into Asmara on 24 May 1991,
while Eritrean units were seconded to the EPRDF forces simultaneously
advancing on Addis Ababa, seized several days later. The EPLF had seized
control of the frontier; the politics of the borderland now loomed large in
the north.
Just as the TPLF was the product of a century of violent struggle, political
upheaval, and ultimately marginalization in the north, so the EPLF had
similarly deep roots, and was the outcome of a troubled, turbulent, and ulti-
mately extraordinarily fertile frontier. Yet what was remarkable about the
EPLF itself was the manner in which it sought to impose order on the fron-
tier, a military solution to the problems of provincial and cultural complex-
ity, and political ambiguity—although, as we shall see, the full implications
of that order, the nature of the solution, would only become clear in the
years after independence. Besides the importance of the decade or so in
which the EPLF was able to prepare for government, two other long-term
issues need to be reiterated here.The first is that Eritrea became, in the 1970s
and 1980s, an extraordinarily brutal and violent political environment, char-
acterized by violence and counter-violence between a militarized state and
 m i l i tant i de nt i ty, 1974 – 91 2 07

its militarized frontier. Lessons in control and force would be learnt by the
leadership of the EPLF.The political culture of the rear base, of the siege and
the fortress, would continue to develop into the 1990s and beyond. There
were high hopes in the early 1990s that a new kind of politics would prevail;
but the war of liberation had taken rather more of a toll than was immedi-
ately apparent. Second, the prolonged and deep-rooted disputes with the
TPLF, and the subsequent propulsion of the latter into power in Addis
Ababa, meant that it was not a matter of if, but when, Ethiopia and Eritrea
would return to arms against one another; the bitterness and suspicion
between the two movements, while often unspoken and low-level, were
nonetheless poisonous.
In many respects, the deposition of Mengistu’s regime signalled the end
of an era in which professional soldiers—the descendants of colonial armies
elsewhere in Africa, though not in Ethiopia itself—had both a monopoly on
physical force, as well as on available weaponry, and also a monopoly on
moral rectitude.The army barracks of the 1960s and 1970s had been the sole
repositories of the moral and certainly the physical order—however wrongly,
it had been believed that the barracks were the wellsprings of political and
social change and development. But a new wave of armed force had broken
both monopolies, and the movements involved had demonstrated them-
selves more than equal to the task of challenging the state-level purveyors
of violence. These represented aggressive new markets of violence, fuelled
by heady levels of material and ideological investment, spurred by emergent
nationalisms and ethnocentrisms. And those marketplaces were crowded
indeed: both the EPLF and the TPLF were as much the products of pro-
longed and bitter civil wars as they were the victors over some ‘external’
force. The internecine dimension of their violent formative experiences
cannot be underestimated, although it often is, owing to the drama of their
victories over the Derg.Yet in other ways, the events of 1991 were merely the
latest—albeit arguably the most dramatic—manifestation of a pattern of
long standing in regional history, namely the advance of the armed border-
land on the centre. In 1991, the armed frontier seized the centre, both in
Eritrea and in Ethiopia; it remained to be seen, however, whether the poli-
tics of the frontier might be left behind as easily as the frontier itself.
 8
New States, Old Wars
Violence, Frontier, and Destiny in the
Modern Era

The dead governing: liberation and federation

By the middle of 1991, the armed frontier had recaptured the centre, both
in Eritrea and Ethiopia. The fact of prolonged armed struggle had impor-
tant implications for the nature of state and society across the region. To
varying degrees, these movements exhibited a belief in the absolute primacy
of armed force in the resolution of ‘problems’, whether internal—in the
suppression of opposition—or external, in the confronting of adjacent ene-
mies. Violence, simply put, produced violence exponentially—as the very
existence of the movements in question demonstrated.The success of armed
force in the achievement of political power persuaded a new generation of
leaders that it might be deployed in other arenas; and there would be the
arming of new frontiers to confront those new centres in their turn.
Moreover, despite the various claims made for democratic, social revolution
by movements of liberation, the latter were organized necessarily around
unquestioning obedience; guerrillas, in sum, would prove unenthusiastic
democrats, as wars fought by vanguard movements on behalf of broader
populaces invariably produced cultures of authoritarianism. Closely con-
nected to this, blood sacrifice was often translated into a moral right to rule,
which in turn was reconfigured into the fulfilment of some kind of national
destiny. No ‘civilian’—that is to say, non-combatant—could possibly be
justified in challenging this, and if they did, it might be interpreted as a new
form of treason. The exclusivity of participation in the armed struggle bred
new political elites. Eritrea provides the best example. Here, the memory of
 new state s, ol d war s 209

the martyrs—those who fell in action during the armed struggle—would

become sacred, and increasingly, though almost imperceptibly, political. The
struggle itself—‘the field’—would become the stuff of legend, and ex-fight-
ers themselves constituted a new elite, save the permanently disabled, who
were provided with accommodation and basic sustenance but who lived on
the periphery of society. Ethiopian martyrs likewise joined the great pan-
theon of the fallen in defence of the Ethiopian motherland, and became
part of the great military tradition—even if veterans might be seen begging
in the streets of Addis Ababa. In both Eritrea and Ethiopia, these new elites
became the righteous, battle-scarred guardians of national destiny, and pos-
sessors of arcane knowledge.
Initially, however, there was a fair amount of political creativity in both
countries.1 First, they had to separate: that had been the arrangement reluc-
tantly accepted by the TPLF several years earlier, even if they had periodi-
cally questioned the entire basis on which the EPLF was pursuing and
organizing the struggle itself. While the new Ethiopian government might
regard itself as magnanimously setting the Eritreans free, and paternally sup-
porting the latter in their bid for international recognition, for its part the
Eritrean government saw itself as having long since won the argument
through its own achievements in ‘the field’, and as the guarantor of order
and stability in Ethiopia (never mind Eritrea) during this difficult period of
transition. The EPLF, some suggest, agreed to wait two years before holding
its referendum on independence, while the EPRDF government stabilized
itself in Addis Ababa. Meles Zenawi, meanwhile, made it absolutely clear
that he would support the Eritrean bid for independence from the outset.
It must be borne in mind, of course, that what made these outwardly effi-
cient, cooperative arrangements especially impressive was the stark contrast
they offered with events further south, in Somalia. There, with the over-
throw of Siad Barre, the state had ceased to exist by the early 1990s; along-
side such chronic disintegration, Ethiopia and Eritrea appeared the resolute
harbingers of a new order in north-east Africa. The situation in Somalia, of
course, would eventually come to consume both Eritreans and Ethiopians—
while relations with Sudan likewise augured ill for the future.
Nonetheless, for now, there was indeed collaboration between Addis
Ababa and Asmara in the lengthy operation to separate the two countries.
In the end, it was something of a hatchet job, but it appeared to work: both
countries survived the procedure, although no-one had noticed the massive
internal damage inflicted. The Eritreans probably did contribute to the
210 revolut i on s, l i b e rat i on s, and g h o st s

stabilization of the EPRDF regime by holding off from an immediate ref-

erendum; it is also the case that Ethiopia, bluntly, had little choice in Eritrea’s
‘secession’, because militarily the argument was indeed won. The facts on
the ground were such that no Ethiopian force could have prevented the
creation of an independent Eritrea. Nonetheless, Asmara would probably
not have been able to make good its bid for international recognition of its
sovereign status had it not been for the strong support offered by senior
EPRDF personnel, and Meles in particular; it is also the case that a stunned
Amhara and Tigrayan population, in particular, within Ethiopia needed to
be persuaded of the righteousness of the Eritrean cause by TPLF cadres.The
lingering doubts over whether the Eritreans would really, in the final analy-
sis, become independent were removed in the course of 1992 and early
1993, even if many Ethiopians—including a great number of people within
the TPLF and the EPRDF more broadly—never actually accepted it. But
they, like the new government in Addis Ababa itself, had little choice.
Eritrea’s departure dramatically and apparently permanently altered the
balance of power in north-east Africa.
The referendum itself, however, took some organizing, and by the time
the moment arrived in April 1993—attended by hundreds of UN and
other observers—Eritreans had long been primed as to the issue at stake,
or at least the issue on the ballot paper.2 They were being asked to approve
or disapprove of Eritrean independence; the EPLF case was clear, and
was deeply appreciated by many Eritreans, based on a history of forcible
incorporation, violent persecution, and the fundamental right of self-
determination. The vote was 99.8 per cent in favour of independence, and
had been, on the whole, well-organized and was declared free and fair by
the UN reporters, notwithstanding some chaotic anomalies around the
country. Eritrea was de jure independent, after two years of de facto. The
rejoicing and justified self-congratulations were marred by the events of
20 May 1993, when a group of fighters, annoyed by the unilateral
announcement by the government that EPLF personnel would effectively
continue unpaid for the foreseeable future, seized control of government
offices and the airport in protest.3 They met with Isaias, and later dis-
banded peacefully; but if they had fears about a newly-detected highhand-
edness in the government’s attitude, it was only the beginning. A few days
later, the protest forgiven if not forgotten, Eritrea celebrated its formal
independence. In a ceremony in Asmara, the key figure at the feast Meles
Zenawi, now Ethiopian president—neither wicked stepmother nor
 new state s, ol d war s 211

Banquo, just yet—declared his hope that ‘the wounds of the past will be
healed’.4 This was the EPLF’s moment.
The first few years of Eritrean independence were dominated by three
broad and intimately interrelated themes, namely the struggle for economic
development, the difficulties encountered in the quest for a stable political
system, and a problematic and complex relationship with neighbouring
states, particularly Ethiopia. While in the immediate term these need to be
understood, naturally, in the context of new-found independence, in fact
each of these challenges was rooted in the deeper past. The rhetoric of the
‘new beginning’, and the bullish and indeed naive optimism of the early
1990s, would soon evaporate. Nonetheless, expectations in 1993 were indeed
high.5 On the face of it, at least, the EPLF possessed an enormous amount
of political capital; few questioned (publicly, at any rate) the EPLF’s assump-
tion that overwhelming popular support for independence in fact meant a
popular mandate for the movement itself, even though that was not the issue
at stake during the referendum. The sheer scale of the EPLF’s military suc-
cess, and the euphoria which attended the achievement of independence,
was sufficient to enable the movement to contemplate the enormous task
of reconstruction with some confidence.The EPLF initially went about this
task with the same vigour and commitment that had characterized its armed
struggle, and Eritrea won plaudits from the donor community for its deter-
mination to avoid excessive national indebtedness, and to ensure that all
much-needed international aid would go directly to the people who needed
it. The government’s much-publicized stubbornness when dealing with
NGOs and donor countries—notably, it demanded absolute control of aid
money6—won it admiration from foreign observers, some of whom became
enthusiastic supporters for a time, although aid workers and would-be inves-
tors increasingly found Eritrean institutions and bureaucracy impossible to
deal with. The relationship had already begun to sour in 1996 and 1997,
when several major NGOs were compelled to leave the country as the gov-
ernment decided it could do certain things better by itself. This was rash,
and would come back to haunt it.
In political terms, too, initial prospects had seemed good. Shortly after
EPLF forces entered Asmara in May 1991, a senior figure in the movement,
Sebhat Efrem, later Minister of Defence, remarked in an interview with the
journalist Dan Connell that a time would come for the EPLF to ‘disappear’,
its mandate fulfilled.7 Others, including Isaias Afeworki himself, would
intimate something similar in the months before the independence
212 revolut i on s, l i b e rat i on s, and g h o st s

referendum. Nominally, of course, the EPLF did indeed ‘disappear’ soon

afterwards: at the movement’s Third Congress, held at Nakfa in February
1994, the EPLF abolished itself and was replaced by the People’s Front for
Democracy and Justice (PFDJ). A National Assembly was created as the core
forum for political debate. In reality, the PFDJ was—and remains—the
EPLF in everything but name, and would continue to dominate Eritrean
politics and society, while liberation-era figures continued to occupy the
vast majority of key government posts from the presidency downwards.
Meanwhile, the constitution-making process received a great deal of atten-
tion down to 1997, when a comprehensive draft was produced; but it was
never ratified, still less applied. The constitution itself, under the overall
direction of Bereket Habte Selassie, was a remarkable achievement, and
involved much consultation and input from Eritreans from many walks of
life, as well as some outside specialist counsel;8 but it was doomed. The war
that broke out a year later was used as the excuse for its abeyance, and it has
collected dust ever since.
The natural authoritarianism of the EPLF’s leadership, as well as attitudes
toward rivals and dissidents alike during the struggle, meant that there had
been, in all likelihood, no genuine commitment to the constitution.9 From
the early 1990s, the government was similarly intolerant of perceived non-
conformers and dissidents—notably certain Protestant denominations and
Jehovah’s Witnesses—and it was clear early on that ‘human rights’ were not
a particular priority for the Eritrean state.10 All this said, in the period imme-
diately following independence, many suspended their natural scepticism;
human rights abuses were relatively limited and did not, on the whole, tar-
nish the reputation of what was still seen as an essentially progressive, prin-
cipled and well-intentioned political movement. Meanwhile, a measure of
amnesty had been granted to members of the ELF, for example, some of
whom returned to Eritrea to participate in the nation-building exercise.
Many of these worked with refugees, in the rebuilding of social and eco-
nomic infrastructure, and in the education sector. Troubled relations with
Sudan and Yemen notwithstanding, there was enough positivity to engen-
der a degree of optimism that the great northern war was indeed over. It
was to prove an illusion.
Nonetheless, while the process of temporary disarmament—or at least
standing down—of the northern frontier was taking place, plans were being
developed within Ethiopia for novel political arrangements, for there were
other frontiers to be mollified and indeed incorporated. In essence, the
 new state s, ol d war s 213

ethnic federalism which the EPRDF government initiated had its roots in
the zemene mesafint, and in many respects was the logical response to a cen-
tury and a half of centrifugal violence, even if this was not the explicit
justification for it—the deep past rarely makes its way into mainstream
political discourse. More directly, it represented an attempt to address the
chronic problem of the multi-ethnic nature of the Ethiopian empire-state—
and more expressly, the emergence of militarized ethnicity—which had
become clear since the 1960s and 1970s.11 A federal system based on an
equitable distribution of power among ethnic groupings was the only means,
it was argued, by which Ethiopia could survive at all—especially in the light
of Eritrea’s departure—and it was no historical accident that the political
experiment now unfolding was spearheaded by a Tigrayan elite represent-
ing a group which accounted only for some ten per cent of the total popu-
lation of Ethiopia.
Some foreign observers of long standing predictably embraced the new
regime with gusto, and proclaimed the ‘resurgence’ of Ethiopia;12 for many
Ethiopians, however, the question of the government’s intentions was not
long in the answering.There was an initial welcome given to the Transitional
Government of Ethiopia’s announcement in mid-1991 that it would recog-
nize the country’s ethnic diversity by reorganizing Ethiopia as a federal state
divided into ethnically-defined regions. It quickly became clear, however,
that regional administrations were to be politically and ideologically affili-
ated with the EPRDF. By the end of 1992, the OLF had already withdrawn
from the coalition, followed early the next year by members of the Southern
Ethiopia People’s Democratic Coalition. The events of 1992–3, indeed,
were salutary.13 Confused and deeply flawed elections were held in June
1992, in which the OLF and the All-Amhara People’s Organisation (AAPO)
refused to take part, and during which opposition was routinely intimi-
dated. Disorganization at the local level meant that extraordinary authority
was enjoyed by EPRDF officials during the polls themselves. Nonetheless,
although it was difficult for anyone to say whether the result really was an
endorsement of the EPRDF—it seems safe to suggest that it was not—
Meles declared that he had won a popular mandate as president. Within a
few months, various groups had withdrawn from the coalition; and then, at
the beginning of 1993, the police violently suppressed a demonstration by
students in Addis Ababa, marching in protest against the imminent inde-
pendence of Eritrea. Meles brooked no dissent on this issue—and swiftly
moved to purge Addis Ababa University of dozens of academics who were
214 revolut i on s, l i b e rat i on s, and g h o st s

accused of inciting student violence.14 These were mostly Amhara—some

were active members of the AAPO—and it was not difficult for many to
reach the conclusion that the government had now embarked on a form of
ethnic cleansing made possible by an increasingly authoritarian regime.
Notably, of course, the perception grew that Meles Zenawi himself—with
Eritrean heritage—was suspiciously ‘close’ to the Asmara government, and
appeared to Amhara critics to be Isaias’ stooge in Addis Ababa; with the
TPLF in power, Ethiopia was losing its coastline, and a ‘natural’ part of
‘ancient Ethiopia’. It was one part of the bitter ethnic melange of Ethiopian
political culture in the 1990s and into the new millennium, characterized
by shifting and multi-dimensional enemies, positioned along different
borderlines at the same time, straddling issues and overlapping loyalties.
The government pushed ahead with its constitutional programme. In
1994, the new federal constitution was unveiled, of which Article 39.1 stipu-
lated the right of regions to self-determination up to and including secession.
The events of the early 1990s in Ethiopia contained more than a touch of
irony: a strictly applied constitution along the lines of the 1994 version,
forty or even thirty years earlier, might have kept Eritrea inside Ethiopia; it
took several decades of violence, including that in Tigray itself, to arrive at
this constitutional reality. Ultimately it was an attempt on the part of the
TPLF at the constitutional disarmament of the frontier, of which the TPLF
itself was a product. Even then, of course, it has not been at all clear that the
EPRDF government truly accepts the spirit of the new constitution; the
TPLF leadership is more aware than any regime previously of unfeasibility
of centralism, covert or otherwise, and yet it still needs to remain in power—
and so it has tended to embrace the political strategies of its predecessors,
with some novel elements thrown in. But as Prendergast and Duffield put
it, the ‘brilliance’ of the federal system lay in the fact that ‘it regionalised
conflict that might have been expressed at the national level’.15 Ethnic
regionalism, according to this view, represents the attempt by the EPRDF
to quarantine violence which might spill over onto the national arena; if so,
it represents a chronic misunderstanding of Ethiopian history, beyond a rec-
ognition that regional conflict has been extremely destabilizing. Ethnicity,
in this context, has been used at the level of the state to devolve violence to
the regions to which it properly belongs. As Donald Donham suggests, the
ethno-federalism of the EPRDF ‘was a form of “indirect rule” based on
official definitions of ethnicity’.16 Above all, of course, this was a system
brought into being by a militant Tigrayan leadership acutely aware of its
 new state s, ol d war s 215

vulnerability at the centre of the Ethiopian polity; if Tigray was to maintain

its rediscovered dominance, or even simply survive, the TPLF would need a
deft combination of armed force and political guile—and some degree of
cooperation with the Amhara and the Oromo.
In recent years, there has been increasing interest in the relationship
between both ethnicity and power in contemporary Ethiopia, and between
violence against particular groups by the state and ethnic identity, whether
claimed or imposed.17 Furthermore, in the case of both Ethiopia and Eritrea,
there has been an examination of the manipulation of ethnic imagery in the
development of group cohesion—particularly in the context of state-
formation from the early 1990s onward—and the flexibility and indeed
ambiguity of ethnic belonging.18 There is little need to recover the ground
traversed by much of this excellent scholarship, constituting as it does an
unprecedented level of scrutiny of the role of ethnic identity in Ethiopian
political life, notably from the perspective of democracy and human rights.19
This relatively new focus is to be warmly welcomed, because this is schol-
arly research which involves holding political systems both to account and
up to close scrutiny. I argue, however, for the need to avoid foreshortening
the analysis of ethnicity, identity, and conflict across our region, and to
lengthen as well as deepen our understanding of some key phenomena and
processes in north-east African political culture. Again, the region is crosscut
with a series of interlocking fault lines and borders, which sometimes
follow what we can loosely call ‘ethnic’ fractures, and at other times run
across ethnicity and involve the boundaries of the state and of state-forma-
tion. There can be little doubt that the modern Ethiopian government has
used ethnicity to manipulate public opinion and discourse, and that ethnic-
ity itself is frequently characterized by ambiguity: the most obvious example
in our region is the Tigrinya bloc straddling Tigray and highland Eritrea, the
‘trans-Mereb’ zone. But this does not mean that therefore ethnicity itself is
a deliberate creation, a product of the instrumentalism at the heart of mod-
ern human relations and power structures. Rather, such episodic appeals (or
reversions) to ethnicity co-exist with deeper-rooted, primordial senses of
what it is to belong to a particular group, state, or ‘community’, while the
latter have themselves often been forged through violent struggle over the
longer term. It is also clear that ethnicity is a political weapon, as we have
noted earlier, and that particular ethnic groups are ostracized or persecuted
because of their perceived propensity for opposition—although clearly
‘political groups’ might also be ethnically ostracized, i.e. labelled as belonging
216 revolut i on s, l i b e rat i on s, and g h o st s

to a particular ethnicity which has always ‘caused trouble’, or which is per-

ennially ‘lawless’. Either way, the end result is invariably the formation of
kinds of armed borderlands described through this book.We return to these
issues below.
The new constitution was approved by the Ethiopian parliament in mid-
1995, with Meles becoming prime minister and Negasso Gidada, an Oromo
activist, becoming a largely ceremonial president, a post he held until falling
foul of the leadership in 2001. Indeed efforts were made to create the
appearance of inter-ethnic cooperation in government, with Amhara,
Oromo, Somali, and Gurage, among others, being represented in the coun-
cil of ministers; but few close observers could doubt that the key ministers
and assistant ministers were Tigrayans.The army was reorganized, too, under
the direction of Tsadkan Gebretensae and Yemane (‘Jamaica’) Kidane, both
central figures in the military and political wings of the TPLF. Again, senior
officers from other ethnic groups were appointed, including Oromo and
Amhara; clearly, however, these were carefully vetted, ideologically recon-
ciled individuals, and favour was still given to Tigrayan officers.20
In the immediate aftermath of the war with Eritrea, there was the first
major split within the TPLF over its prosecution. In 2001, there was a purge
of the politburo and cabinet in which Meles effectively removed those who
accused him of failing to finish the job, of pulling his punches in dealing
with the Eritreans, who had been, it was supposed, on the ropes in the mid-
dle of 2000 when Meles agreed to a ceasefire.21 It was an indication—or a
reminder, to those rather better acquainted with the movement’s history—
of the Meles faction’s brutality and intolerance of dissent at the centre of the
organization. Indeed, in the early twenty-first century, new opposition
movements appeared within Tigray itself, bitterly opposed to the hegemony
of the TPLF and its cronyism, and serving as a stark reminder of how the
TPLF itself had risen to prominence through the destruction of rivals within
the province. But the TPLF found it increasingly difficult to engineer the
failure of those which would challenge it, whether within Tigray or else-
where in Ethiopia. Certainly, in the early years of the twenty-first century,
the TPLF appeared increasingly besieged, with a mentality to match, at the
heart of the Ethiopian polity; for that reason, too, many Tigrayans rejected
the movement, for they saw it as ultimately jeopardising Tigray itself, and
exposing it to some violent future backlash, as had befallen the Amhara in
the 1980s. It was increasingly common to hear mutterings that Tigray was
in effect being ‘held hostage’ by the TPLF; born aloft on a new wave of
 new state s, ol d war s 217

Tigrayan nationalism, the TPLF was now placing Tigray in grave danger. In
the years following the Eritrean war, the EPRDF government became ever
more authoritarian—or at least, ever less willing to conceal it, much to the
annoyance of those foreign governments which had backed Meles Zenawi
with such enthusiasm. It was increasingly clear that the TPLF-led state had
no intention of surrendering power for the foreseeable future.
The Ethiopian government also took advantage of the US interest in the
region as part of the war on Islamic extremism. Under this kind of cover the
Ethiopian security forces could certainly pursue the growing number of
armed insurgencies across Ethiopia with relative impunity. Below, we look
briefly at the Ethiopian intervention in Somalia; here, it is important to note
the allegations made concerning the existence of detention centres in east-
ern Ethiopia for Muslim ‘terror’ suspects.22 There was novelty here in the
context, and in the international climate, and arguably in the degree to
which state apparatus could be deployed to political ends; but there was
precedent in the nineteenth century and earlier in the treatment by the
highland polity of Muslim threats.Tewodros and Yohannes alike would have
recognized the ‘mission’ embarked upon by the EPRDF state to quash the
radical Islamic threat looming to the south and east, and indeed within
habesha society itself. Elements of Somali Islam were indeed more markedly
radical—there were foreign models to provide inspiration in motive and
organization, notably in Iraq and Afghanistan23—but this was only the latest
stage in a long and ambiguous religious war in the Horn. As elsewhere,
it would prove self-perpetuating: Ethiopia would become an even more
vulnerable target for the region’s Islamists as a result of its close association
(however misconstrued) with the US in the latter’s ‘war on terror’, and its
own bloody adventures in Somalia. None of this prevented the Ethiopian
government from blaming the Eritreans for ultimately being behind occa-
sional bomb blasts in Addis Ababa and elsewhere.

The great unexpected war

There is an expanding body of literature on the causes and nature of the
1998–2000 war between Eritrea and Ethiopia. At various times and in vari-
ous contexts, writers have been struck by the apparent futility of the con-
flict, and have broadly lent support to the ‘bald men fighting over the comb’
218 revolut i on s, l i b e rat i on s, and g h o st s

notion which has undermined comprehension of events from the begin-

ning. It is a characterization, indeed, which has warped understanding of
African warfare more generally for a very long time, as I have argued else-
where.24 Others have dwelt on the ‘family feud’ concept, resting on the idea
that the EPLF and the TPLF, or Eritrean Tigrinya and Tigrayans, or even
Eritreans and Ethiopians generally, are somehow ‘brothers’ who then spec-
tacularly fell out.25 Some of the more nuanced literature has highlighted the
shocking humanitarian as well as wider political and historical implica-
tions.26 Of course, in the post-Cold-War world, this curiously inexplicable
conflict—reminiscent, so many journalists could not resist informing us, of
the Western Front in the First World War—was supposedly a reminder of
the simmering barbarism and atavism of politics in the Horn. In that sense
it was easier to draw out the peculiarly African nature of the war. Nonetheless,
while the initial rush of scholarly and journalistic analysis was characterized
by a fair amount of bias for one side or the other, and certainly routine
accusations and counter-accusations of subjectivity, the academic treatment
of the war—while somewhat in abeyance at present—is growing steadily
more sophisticated and less overtly subjective.
This was, then, the great unexpected war; no-one, or very few, saw it
coming. Perhaps this is the most singularly remarkable aspect of the entire
conflict. It became less ‘unexpected’ as time went on, naturally enough; cer-
tainly, within a short space of time, there could be scarcely any doubt about
the degree of mutual hostility, nor about the existence of in-built cultural
and political mechanisms to facilitate such animosity. It is also clear that our
understanding of the war has been hampered by geographical reductionism
and temporal foreshortening, misrepresentation in terms of both time and
space. It was not simply about who owned Badme, but concerned the com-
petition for hegemony across the wider region, as well as access to the Red
Sea coast, in itself an issue of some antiquity. Nor was it simply about the
stupidly dangerous disagreements between two men, or even two organiza-
tions, but rather the resumption of the war in mid-1998 after a brief armistice
was the latest—and arguably the most violent, to date—manifestation of a
pattern of violence both within and around the contested Eritrean zone
dating to at least the middle of the nineteenth century.The borders between
Eritrea and Ethiopia mattered; but what was more important, now, was the
fact of Eritrea as frontier nation, built on a fault line, and of the militaristic
and chauvinistic EPLF as the product of that volatile environment. The
Ethiopia of the TPLF-EPRDF, too, was the product of a concatenation of
 new state s, ol d war s 219

frontier wars; but once again it was Eritrea which became the epicentre of
a regional network of violence.
The 1991–8 period was a mere armistice in a very long war. Initially,
however, as in other spheres, the signs were positive. The Ethiopian govern-
ment accepted and blessed the outcome of the 1993 referendum; this was
quickly followed in September 1993 by a series of agreements between the
two states encompassing trade, banking, defence, and nationality.27 Indeed
through the mid-1990s, a form of neo-unionism had emerged, espoused by
senior Eritrean figures who spoke warmly of the intimacy between the two
countries. Economic ties appeared particularly close, not least because
Ethiopia represented Eritrea’s commercial hinterland; the agreement of
September 1993, followed by another as late as January 1997, demonstrated
an official expectation, at least, that this economic interdependency was to
continue, indeed grow stronger.28 Whether wishful thinking on both sides,
or the clever deployment of rhetoric in advance of betrayal, or simply politi-
cal naivety, this public language masked deep problems, and there was a fair
amount of protesting too much doubtless willed on by the need for a period
of recuperation and consolidation. This element, indeed, is clearly critical,
and often overlooked; neither the TPLF nor the EPLF was in a position to
adopt an aggressive stance toward one another in the early 1990s, and even
if—as is likely—leaders on both sides anticipated some future trial of arms,
a little breathing space was necessary given the massive military effort of
recent years. Legitimacy, moreover, both internal and external, would have
been less forthcoming had a clash come sooner.Whether either side expected
a war on quite the scale of 1998–2000 is another matter; it seems, on balance,
they did not.
Eritrea’s introduction of a national currency, the nakfa, in 1997—previ-
ously it had used the Ethiopian birr—led to a trade war which escalated
rapidly in the months prior to the outbreak of war in mid-1998.29 This was
surprising, on the face of it, as a separate Eritrean currency had been antici-
pated in the 1993 agreement. Ethiopia boycotted the port of Assab, in effect
an Ethiopian port in any case, for it hardly served the Eritrean economy in
any meaningful manner, and accused the Eritreans of imposing hefty duties;
the Eritreans denied this.Trade between the two countries swiftly withered.
It was the beginning of an economic war which Eritrea could not win.
Meanwhile there were problems at the border, too, along the ancient fault
line which ran across the northern region. Border problems, as we have
seen, existed during the liberation struggle, and were manifest in clashes
220 revolut i on s, l i b e rat i on s, and g h o st s

over administrative zones between the Ethiopian and Eritrean movements.

They continued through the 1990s, approaching crisis point by 1997–8.30 In
July and August 1997, Ethiopian incursions in two areas—Badme in the
west, and Adi Murug in the centre—prompted an exchange of letters
between Isaias and Meles concerning the need for restraint and eventual
demarcation,31 something in which the TPLF leadership had been inter-
ested since the early 1980s. Indeed a joint border commission had met
sporadically through the 1990s, but in time each side would accuse the
other of failing to take this seriously. These were no mere teething troubles:
the real issue was the very nature and indeed existence of Eritrean state-
hood, however important were other, more specific, issues at the local level.
The question remains how much the Ethiopian government, increasingly
self-confident and secure, genuinely accepted Eritrean independence,
involving as it did the loss of Ethiopia’s entire coastline and a substantial
population held by most outside Ethiopia itself to be ‘Ethiopian’. Similarly,
the question remains how much the Eritrean government actually had faith
in the viability of its new nation, and certainly in its regional security.
Nonetheless it would adhere doggedly to the argument that the border
question was ‘straightforward’: it was a matter of colonial boundaries, no
more and no less. Ethiopia argued that these boundaries had never been
demarcated on the ground, and that therefore the question was one of actual
occupation and administration. As the war unfolded, depths of antagonism
between the two states were rapidly revealed which were clearly related to
a much wider range of issues. This is not to suggest that borders did not
matter; on a certain level, they did, of course. But border demarcation was
only part of a much more complex historical and geopolitical equation.
What happened at the frontier between these two states was symptomatic
of tensions at their respective centres.
The events of the war itself are well known and require only a summary
here.32 In May 1998, an exchange of fire took place at the hamlet of Badme
in which several Eritreans were killed, and the Eritrean army moved in
force to retake the disputed area. The Eritreans appear to have been rather
better prepared for this eventuality; the Ethiopian army, for all the evidence
of armed militia in the area, remained under strength for several weeks.33
Fighting erupted along the central portion of the border, too, and by early
June the conflict had begun in earnest, complete with Ethiopian air-raids
on Eritrea and retaliation by Eritrea on the Tigrayan town of Mekele. Early
attempts at mediation, through both a US-Rwanda peace plan and the
 new state s, ol d war s 2 21

OAU, failed, and although the war entered a quiescent phase for several
months the conflict itself was increasingly disastrous for Eritrea, not least
because of the mass mobilization which took place, placing enormous
strains on a fragile society and economy. Moreover, a new ‘front’ opened up
in the form of deportations: beginning in June 1998, Ethiopia deported tens
of thousands of ‘ethnic Eritreans’ and Ethiopians of Eritrean descent, and
the vitriolic language which characterized the Ethiopian government’s pro-
nouncements spoke of a deep well of resentment and bitterness regarding
the new state to the north, and its people, broadly defined. The Eritrean
government did not immediately respond in kind and generally showed
restraint in its press releases, as well as something close to over-compensa-
tion in terms of its protective treatment of Ethiopians living in Eritrea.34
Only later did the Eritrean government initiate an extensive and at times
brutal programme of ‘repatriation’ of Ethiopians, especially Tigrayans, from
the country.
In February 1999, a major Ethiopian offensive recaptured Badme, and the
Eritrean government indicated a willingness to accept the OAU peace plan
in what was widely interpreted as an embarrassing climb down. It should
have signalled the end of the war; but the war continued, with some of the
worst fighting taking place along the central front between April and June
1999. Once again, the war entered a lull, and diplomatic activity on the part
of the OAU, the UN, and various individual interested parties was frenzied.
However, in May 2000, Ethiopia launched its most overwhelming offensive
to date, breaking the Eritrean lines in the west and advancing deep into
Eritrean territory. The Eritrean army was compelled to abandon the west of
the country, and withdrew into the mountain plateau from which they
managed to halt the Ethiopians, at great cost to the latter. Even so, for a
country which had prided itself on its military capacity and ability to with-
stand any external threat, the offensive of May and June 2000 was shocking
to the Eritrean populace.35 In addition to the material and economic disas-
ter which it brought, these setbacks also had considerable political fallout
which would only become clear after the war, when deep rifts inside party
and government over the prosecution of the war would come to light. In
reality, the two sides had fought each other to a standstill; it was a war
neither could hope to ‘win’, at least in the unconditional sense. Eritrea
could not have withstood further assaults, certainly not in economic terms,
while strains were already telling on broader society as well as within politi-
cal structures and relationships. The Eritrean government could claim to
222 revolut i on s, l i b e rat i on s, and g h o st s

have withstood Ethiopian aggression, and thwarted Ethiopia’s expansionist

plans; and certainly the implicit Ethiopian aim of removing Isaias Afeworki
had not been realized, however much the latter’s position had been shaken
by the events of the war. But Eritrea was exhausted, and large swathes of the
population were weary and disillusioned—although not as much so as they
would become. When the ceasefire of June was followed by the formal
agreement of December 2000, it was greeted in Eritrea with a marked
blend of relief and tired sadness. Many would wonder, after all, what the
conflict had been about; at least Eritrea itself was intact after an experience
as bruising in its own way as the decades of liberation struggle, but as
the political cost of the war became clear, many would begin to question
even this.
Until 2008, a demilitarized zone patrolled by the UN divided the two
countries, and the border remained undemarcated, while two of the largest
armies in Africa faced one another down, in places only metres apart. In
early 2002, a border commission produced the document to which it was
hoped both sides would adhere, outlining how the border would be delim-
ited and demarcated.36 While it was accepted by Eritrea, the Ethiopian
government rejected key parts of it—notably the awarding of Badme to
Eritrea—and in the years that followed requested further negotiation on the
matter. Eritrea, adamant that the findings were to be final and binding,
refused, and has become increasingly outraged that more international pres-
sure has not been placed on Ethiopia to implement the agreement. Pressure
on the two sides to reach compromise has produced no results. As the UN
finally pulled out in 2007–8, the border was ‘virtually’ demarcated—a politi-
cal nonsense which was nonetheless endorsed by Isaias. Relations with
Ethiopia are not yet normalized, nor do they look like being normalized in
the near future.37 Cross-border incidents occur regularly, and at the time of
writing the government-owned media in Eritrea regularly warns of further
Ethiopian expansionist plans, even as thousands of Eritrean conscripts leak
into both Ethiopia and Sudan in search of asylum. The spectre of war,
whether real or imagined, continues to stalk Eritrea, and in truth the gov-
ernment has utilized the external threat in stalling demobilization, keeping
society at a high level of military preparedness and justifying tight political
discipline.
A great deal of attention has been paid, again, to the creation of hated
enemies, the manipulation of ethnicity and nationalism, and the perception
of the past in both Eritrea and Ethiopia, but particularly the latter, as the
 new state s, ol d war s 2 23

result of the war.38 On the Ethiopian side, it is in fact a book that pre-dates
the war, rather than anything that comes after it, which best demonstrates
the sense of simmering antipathy toward the EPLF in particular, and per-
haps the kebessa Tigrinya in general. Tekeste Negash’s study of the federal
period of the 1950s is in fact an ill-disguised condemnation of the EPLF,
which,Tekeste suggests, embodies the fickleness and cynical Machiavellianism
of the Christian highlanders.39 Before the war, too, Alemseged Abbay tried
a gentler tack by suggesting that Tigrayans on either side of the Mereb were
the same people but the EPLF chose a ‘divergent path’, rejecting primordial
links with their southern brothers and developing an ‘instrumentalist’ iden-
tity.40 Certainly, regardless of any inter- or indeed intra-ethnic nuance, the
pain of the loss of coastline once again coursed through the Ethiopian body
politic, the amputation all the more traumatic as now the inherently aggres-
sive, confrontational sha’abiya were once again on the march, the ungrateful
recipients of Ethiopian political and economic largesse. Isaias’ military hubris,
apparently well known during the liberation struggle years, was rediscov-
ered through hindsight and was now exposed for all to see. Eritrea was
indeed, as one contemporary polemic had it, Ethiopia’s ‘problem child’.41
Ethiopia was the victim of a northern militancy which—not content with
the attainment of its sovereignty—had now gone so far as to encroach on
Ethiopia’s own sovereignty, much to the ‘shock’ and ‘disappointment’ of
ordinary Ethiopians.42 When the Eritrean government accepted the origi-
nal OAU peace plan following its defeat on the Badme front in February
1999, it was rejected by Addis Ababa as a cynical and disingenuous ploy to
buy time in the face of unexpected military setback. Eritrea and Eritreans
could not be trusted—witness the ongoing expulsion of tens of thousands
of ‘Eritreans’ from Ethiopia. This would be softened somewhat later on
among many ordinary Ethiopians, who—like increasing numbers of exiled
Eritreans—came to focus on the megalomania of Isaias himself, pitying
their Eritrean ‘brothers and sisters’ who were now languishing under his
dictatorial rule.43
For now, however, ‘Eritrea’ was a condition to be addressed as aggres-
sively as it had erupted. Ethiopia’s war was righteous, and just;44 and in that
sense the events of 1998-2000 fed easily into a pattern of celebratory mili-
tarism which had been a core element in the Ethiopian nation-building
project since the mid-nineteenth century, and which had been a part of
habesha political culture for several centuries. Days after Badme fell in late
February 1999, there came the annual celebration of Adwa across Ethiopia.
224 revolut i on s, l i b e rat i on s, and g h o st s

In Addis Ababa, predictably, the event took on new meaning, and the pub-
lic celebrations were frenetic. Once again, a northern invader had been put
to flight; once again, Ethiopian troops had stood firm on the northern
frontier.45 Ultimately, victory over Eritrea was a ‘second Adwa’, and indeed
placed clear psychological as well as physical distance between the Eritrean
aggressor and the Ethiopian victor—a kind of ‘firming up’ of the northern
boundary, although the creation of this would hardly bring stability, as we
see below, but rather would merely institutionalize and formalize instabili-
ty.46 For some, there was resonance too with the Italian war of the mid-
1930s, especially among the few surviving ‘Patriots’.47 There were, of course,
overlapping pools of identity involved in this: Tigrayans could frame the
war in very local terms, not least because of their proximity to the historic
frontier zone and cultural affinity to the Tigrinya of the Eritrea plateau, for
example; others (not least the EPRDF government itself ) needed to appeal
to a pan-Ethiopian nationalism, remarkable in its durability, and overriding
ethnic or regional tensions.Tigray, for example, had fought many of its past
wars with the Amhara, not exclusively with enemies lurking in the so-
called Mereb Melash, the land beyond the Mereb. For the Amhara, the
sha’abiya represented the loss of Eritrea and the culmination of the great
northern war which had smashed their hold on power, for the time being;
the TPLF, indeed, had inherited their losses. In any case it scarcely mattered:
a generic tradition of heroic violence could be mobilized according to
current needs, and tailored to current circumstances. It meant that, come
the Eritrean war and the new ethno-political dispensation of the 1990s,
there were various vintages of violence which could be carried up from
the great cellars of public memory, and which all could sup, at least in
principle.
On the Tigrayan side of the Mereb war zone, the response to the war was
an ambiguous melange of shock, sadness, and prepared resignation.48 Eritreans
were interlinked with Tigray through a complex mesh of personal relation-
ships, cultural and linguistic ties, and economic activities; but they were also
aggressive, untrustworthy, haughty: characteristics writ large in the EPLF—
and indeed ultimately in the person of Isaias himself, whom Tigrayans would
have had to invent had he not already existed, as the perfect Eritrean cari-
cature. Yet these ambiguities—characteristics, surely, of most frontier zones
throughout human history—were mirrored on the kebessa side of the bor-
der, as I have demonstrated elsewhere.49 At the larger level, Ethiopia—
conceived as a monolith when the need arose—remained an expansionist
 new state s, ol d war s 2 25

empire-state, as it had been since the late nineteenth century, and remained
unreconciled to the loss of Eritrea. In this context it did not matter who was
at the helm in Addis Ababa.Yet on another level it clearly mattered a great
deal. Tigrayans had been part of Eritrean communities along the border;
there had been intermarriage and shared culture, and a rich common herit-
age; peoples on both sides of the Mereb, notably, had struggled against
Amhara domination, and here they had much in common with western
lowlanders, too. And yet Tigrayans were never to be fully trusted; they were
venal, mean-spirited, and scheming, characteristics incarnate in the person
of Meles Zenawi, just as Isaias was the personification of kebessa hubris.
Tigrayans were, in short, despised. Indeed the closeness of Tigray appeared
to make this war even more bitter than any past conflict with the Amhara
far to the south—partly, indeed, because Tigrayans were commonly held in
such contempt, while the Amhara were viewed as more ‘natural’, and more
respected, opponents. The proximity, in the end, only made the differences
starker and, it seemed, more deadly; the ambiguities of the Tigray–Tigrinya
relationship were clearly more dangerous than any sharp edges.
From the middle of 1998, apparently deeply held feelings of contempt as
well as the bitterest of memories were aroused in Eritrea, and were expressed
both privately and publicly. Some writers reflected on Tigrayan expansion-
ism from a historical point of view, and were drawn inexorably to compari-
sons of the TPLF with Yohannes IV.50 Predictably, patterns of ‘Eritrean’
resistance were discerned in the distant past, patriotic spotlights cast into the
dark corners of history much as they had been in the enemy’s camp. Others
recalled how the TPLF had been driven by deep-seated resentment, or were
gripped by an unnerving inferiority complex, or had laid claim to Badme
during the liberation struggle, long before (it was suggested) it became a
serious issue.51 Along the border, many expressed the view that the Woyane
had been planning this for a long time, and that they were simply following
a much deeper tradition of Tigrayan military leaders making violent incur-
sions into the kebessa.52 And of course, what clearer evidence was there of
the TPLF’s hatred for Eritreans than the deportations?53
The pact of forgetting was broken, and almost every view expressed on
one side had its echo on the other: untrustworthiness; dangerous ambition;
prior planning for this war; military aggression born of insecurity. So much
of this was the inevitable product of highly charged emotion at a time of
national crisis, clearly; and scholars need to exercise caution in the utiliza-
tion of such testimony. But the alacrity with which communities and states
226 revolut i on s, l i b e rat i on s, and g h o st s

alike presented their animus is telling. Real historical experience shaped

views on both sides; the trans-Mereb relationship was one built on the
region’s major fault line, the epicentre of so much violent—and creative—
instability, and therefore was both ambiguous and highly volatile. This was
no disingenuous passion, but the reflection of a troubled association between
communities, peoples, political organizations, and, ultimately, state struc-
tures. To a very real degree, the events of the late twentieth century were
rooted in the violent century between 1850 and 1950. In their different but
overlapping ways, each of the Eritrean, Tigrayan, and Oromo struggles
developed out of past experience over the long term, and were the products
of layered identities which were organic, even as they were increasingly
exploited by modern state apparatuses.Too often the functional element has
been exaggerated, leading scholars to think too much on the present at the
expense of historical experience. Certainly, a great deal of the imagery was
the product of modern states’ manipulations; but wholly instrumentalist
interpretations are misleading.The past mattered; lived experience was criti-
cal. This was now clearly manifest in the reality of Eritrea as the frontier
society, the institutionalization of violence which was the outcome of both
residual Ethiopian territorial and political ambition and, in particular, the
EPLF’s conquest of the contested Eritrean frontier zone. It is this to which
we now turn.

Borderline nation: Eritrea as frontier society

The hope invested in the new state of Eritrea in the early 1990s—in truth,
as much the product of relief at the ending of the long war as of any real
understanding of the political culture of the EPLF—evaporated in the space
of a decade, although in fact the real shift came quickly, in the space of a few
months in 2001. The extraordinary clampdown on dissent in that year was
in fact only a reprise of the 1973 ‘crisis’ within the movement, when ques-
tions had been asked about the authoritarian tendencies of the leadership;
but now, in the context of independent statehood, and in full view of a wary
citizenry still recovering from the war, it took on rather more critical pro-
portions. The immediate roots of the 2001 crisis are to be found in the war
itself, for there had been growing dismay at Isaias’ direct handling of opera-
tions which culminated, many senior figures believed, in the Ethiopian
 new state s, ol d war s 227

breakthrough in May 2000.Whether or not the stories of an imminent coup

against Isaias at that point were accurate, his standing in the army protected
him as the war reached its climax in the hills south of Asmara; but the dis-
sidents were not to be silenced. The private newspapers, which had briefly
flourished just prior to and during the war, became outlets for criticism of
the leadership. In October 2000, a number of prominent Eritreans in the
diaspora produced the so-called ‘Berlin Manifesto’ which criticized the
President’s increasingly personalized and dictatorial system of rule. Early in
2001, an ‘open letter’ to Isaias, written by several high-ranking political and
military figures in the government and ‘leaked’ to the broader membership
of the PFDJ, accused the President of autocratic and high-handed leader-
ship, and of failing to consult either the Party or the National Assembly in
terms of the direction the country was taking.54 For a time, there was no
public response from the President’s Office; but it is now clear that plans
were indeed being hatched, and doubtless many of the dissidents—all too
familiar as they were with the autocratic and indeed at times paranoid polit-
ical culture of the EPLF, for they had been part of it—anticipated detention.
And so it came, in September 2001—a week after the attacks in the United
States, notably—when 11 of the 15 key critics, who included some of the
chief architects of the EPLF’s victory in 1991, were arrested and imprisoned.
(Three others were abroad at the time, and one recanted.) These dissidents
remain in jail at the time of writing.55 The private press was closed down,
and journalists and editors arrested. Public debate was swiftly stifled. It sig-
nalled a new strategy of oppression and intolerance of opposition which
shows no sign of abating.
In truth, of course, many in the EPLF may have seen something like this
coming, for such behaviour was hardly unknown within the Front since its
inception in the early 1970s. There was a recognition that certain senior
figures, in their readiness to follow his leadership in the pursuit of the ‘revo-
lution’, had ultimately facilitated the emergence of Isaias’ own dominance.
He might have been checked in the early 1990s, arguably, as the EPLF was
dissolved and the PFDJ founded, or in the mid-1980s with shifts in the
central leadership; but he was not, and the Isaias faction gradually strength-
ened its hand within the movement in the course of the 1990s.The entrench-
ment of the President’s position would continue in the decade after the
2001 clampdown.56 While through the 1990s, and during the struggle, there
had been no real evidence of the personality cult, it was increasingly clear
that such a creeping cult did indeed exist, and that Isaias had in fact carefully
228 revolut i on s, l i b e rat i on s, and g h o st s

cultivated the image of the simple, selfless, and incorruptible leader, the only
individual capable of guiding Eritrea into the future, the guardian of the
achievements of the ‘revolution’. He became an increasingly isolated figure,
surrounding himself with those who were in complete concurrence with
his views, and creating a series of rivalries and tensions among those below
him. The President’s Office, characterized by a heady conjunction of para-
noia and self-righteousness, swiftly became the sole source of power in
Eritrea. This sense of government under siege was reflected in society at
large as the ‘no-war, no-peace’ situation looked ever more permanent.
Prisons swelled with dissidents, critics, Pentecostal Christians, and Jehovah’s
Witnesses, with draft dodgers and deserters, with ‘spies’ and journalists, and
suspiciously successful businessmen. The judiciary was rendered impotent,
the media remained under the tightest of controls, the Orthodox Church
was watched.57 Movement in and out of the country—both for individuals
and commodities, even of the most basic kind—was soon severely restricted.
Within a few years, Eritrea was widely regarded as one of the most oppres-
sive countries in the world, as well as one of the most militarized.58 While
there was little space for serious opposition to the government inside the
country—with the partial exception of the shadowy armed movement of
Islamic jihadists operating in the Sudanese border areas—opposition move-
ments proliferated abroad, and critics in exile increased in number and
became more vocal. Disparate elements in Europe and North America con-
tinually combined, split, and recombined to form a confusing constellation
of opposition to the Isaias regime, but they were—and by and large remain—
disorganized and disoriented.59 They had little credibility inside Eritrea
itself, as many long-suffering Eritreans regarded them sceptically as cut from
much the same cloth as Isaias himself, while the decision of some to accept
Ethiopian hospitality from time to time further undermined their integrity
and authority in most Eritreans’ eyes.
Much of this was against a background of, and was in many respects
directly linked to, rapid economic deterioration.The war had devastated the
fledgling Eritrean economy—fragile to begin with, following the neglect
and destruction suffered between the 1950s and the 1980s, and hardly assisted
by the government’s centralist and regulatory interference during the 1990s.
After 2000, trade was reduced to a trickle, inward investment was virtually
non-existent, and the effects of massive unemployment were only partially
and in any case temporarily masked by mass military and national service.
Dependency on foreign aid—despite the increasingly shrill rhetoric related
 new state s, ol d war s 229

to self-reliance which had now become dangerous dogma—worsened in

the wake of a series of bad harvests which compounded the material devas-
tation wrought by the war itself. This was true despite the Eritrean govern-
ment’s deteriorating relationships with the European Union and the US,
neither of which possessed the patience or the wherewithal to understand
the journey which had led the country to its grim condition. As 2010
dawned, indeed, reports emerged that the Eritrean government was actually
confiscating incoming food aid, declaring that such outside assistance made
people ‘lazy’.60 The utilization of economic desperation for political ends
appeared to be taking new and ever more brutal forms; but in fact food has
long been politicized—both as a military target and as a weapon, by being
withheld—and evidence of the intersection of food and violence abounds
throughout the nineteenth and twentieth centuries. The destruction or the
confiscation of food has long been used to destabilize or punish enemies
across an environmentally fragile region; the EPLF, it seems, has learnt the
lesson well, the enemy in this case being its own citizenry.
Ever greater distance opened up between government and populace, as
the former maintained a high level of militarization—arguing, only partly
disingenuously, for the need for vigilance against future Ethiopian aggres-
sion—and the latter increasingly faced a future bereft of hope or opportu-
nity.61 The system of fixed-term military service, begun in the early 1990s
with the establishment of Sawa training camp in the western lowlands, was
initially conceived as a nation-building project; in the early twenty-first
century such service was actively undermining the nation. It had become
indefinite, and most young people (and not a few older ones) thought only
of how to escape. The government mistrusted its own population, espe-
cially youth, to a marked degree.62 The University of Asmara, for example,
had long been seen as a source of potential trouble, owing to the agitation
of some students during the war, and the increasing activity of the bravely
(but briefly) independent Students’ Union immediately after it; its eventual
closure stood in stark contrast to the massive expansion of Sawa, which in
many ways was the real heart of the educational system.The University was
replaced by a series of ill-developed ‘colleges’ around the country—a
reflection of Isaias’ own apparent distaste for Asmara, which was seen as
corrupted and troublesome—where students lived in military conditions
and received very little actual ‘education’. Indeed the EPLF’s anti-intellec-
tualism was marked, and was reflected in the contempt expressed not sim-
ply for the University but for the ‘so-called’ scholars in exile who were
230 revolut i on s, l i b e rat i on s, and g h o st s

increasingly critical of the movement. While Isaias declared that an ex-

fighter could do the same job as someone with a university degree, others
sought weakly to defend an education policy that claimed a developmental
agenda but which appeared rather more motivated by control freakery and
growing nervousness that young people were losing the ‘values’ of the
struggle—forgetting, of course, that tens of thousands had been killed or
maimed in the trenches between 1998 and 2000. That nervousness was
reflected in the opening of a cadre school in Nakfa, where selected youths
would be ‘reminded’ of those values; the government’s fear of popular dis-
engagement from politics, indeed, also led to the holding of public meet-
ings in the course of 2005 to invite some degree of ‘dialogue’. It was
short-lived, and a resounding failure, but it provided a brief glimpse of
concern in government circles.63
An oppressive, militarized social system led to a haemorrhaging of
youngsters in the army across the border into Ethiopia and Sudan. It was a
steady leakage of sovereignty, and a matter of political legitimacy much
more serious than any squabble over the demarcation of a border. While
asylum centres across the region, as well as in Europe and North America,
swelled with young Eritreans, their supposed abandonment of the ‘cause’
confirmed in the minds of those still in Isaias’ circle that only he possessed
the character and direction to finish the job: ‘the dogs bark’, he was alleg-
edly fond of saying, ‘but the camel continues to march’. Few Eritreans were
at all clear where exactly the camel was headed, or even if he knew himself;
but the dwindling core of loyalists only had their mistrust of ‘the people’
confirmed, reinforcing the extraordinarily solipsistic belief within the
movement that those who had not been in ‘the struggle’ could not truly
understand it, or the sacrifices made. What I have elsewhere termed the
‘Nakfa syndrome’64—the essentialist conviction that ‘Eritrea’ had been
truly forged in the EPLF’s struggle in the 1970s and 1980s, and that some-
thing of an authoritarian utopia had been created in the rear base around
Nakfa—had defined Eritrean society from the moment the EPLF entered
Asmara in 1991. The celebration of the ‘martyrs’—those who had given
their lives in ‘the field’—was an annual pageantry which was both pro-
foundly emotional but also deeply political, for it represented the last word
in all political discourse.Whatever might happen, whatever problems might
be encountered, however occasionally disgruntled people might become,
the EPLF was untouchable: the memory of the martyrs, and the sacrifices
the movement had made on behalf of the people, guaranteed it. This was
 new state s, ol d war s 231

the culture that hardened into something altogether more malignant in the
years after 2001. Independence Day and Martyrs’ Day were once about the
creation of national cohesion; in later years, they only served to accentuate
the distance between the governing and the governed. It seemed that, after
all, the Nakfa utopia had only been for the few who had actually been there;
everyone else could, at best, only be associated with, but never truly part of,
that great achievement. And so there were fighters, and there was everyone
else; and yet after 2001 it seemed that neither could really be trusted by an
ever mightier executive whose authority was rooted in military command,
and a talent for intrigue.
A similar belligerence, the product of the bitter isolation of the EPLF’s
struggle, was evident in Eritrea’s foreign relations.65 Even apart from Ethiopia,
independent Eritrea’s relations with its neighbours were turbulent, leading
to the frequently made charge that Eritrea is an aggressive and destabilizing
force intent on building a hegemonic or at least spoiling role in the region.
While this may be true to some extent, it has to be understood in the con-
text of the Eritrean region’s long-term historical experience, culminating in
the political culture of the EPLF.66 In fact, some of the problems with neigh-
bours have not been of Eritrea’s deliberate making, and in any case are
comprehensible in the context of young nationhood. Nonetheless, at the
very least, Eritrea’s foreign policy has been characterized by a clumsy naivety,
a reflection, in part, of the inexperienced diplomatic corps which was drawn
from the same pool of ex-fighters making up the ranks of government.
Eritrea’s responses to foreign policy problems have invariably involved some
degree of force, at least in the first instance. Regarded by outsiders as prickly
and uncompromising, the government’s position has always been that
Eritrea’s hard-won sovereignty will come first and foremost in policy con-
siderations, and any perceived threat will be met with whatever force neces-
sary; certainly, independent Eritrea’s political establishment had become the
domain of a liberation movement whose military skills were proven but
whose willingness to negotiate and compromise was hardly a notable char-
acteristic. A belief in the absolute primacy of armed force lay at the heart of
that political establishment.The military hubris of the Eritreans was marked
after 1991; their unconditional military triumph, and the attendant semi-
mythology which grew quickly around it—they had defeated the super-
power that was the Soviet Union, as well as Ethiopia with the largest army
in Africa—had generated an extraordinary degree of self-belief. At the same
time, however, it is important to remember that the Eritrean government
232 revolut i on s, l i b e rat i on s, and g h o st s

was keenly aware of its potential vulnerability in a region dominated by two

much larger, economically stronger, and better established powers, namely
Ethiopia and Sudan. The attempt to destabilize its neighbours was in that
sense a rational policy decision for Eritrea. It might also be pointed out that
it has proved remarkably easy for outsiders to blame the Eritreans for a great
many of the region’s problems; a relative newcomer, Eritrea has lacked the
pools of empathy or even knowledge among western observers and there-
fore has suffered somewhat in terms of representation.
Nonetheless, Eritrea did itself very few favours. Thus, early on there was
distaste for the OAU and its failure to serve Eritrea’s interests in particular
and those of the continent more widely; Isaias declared that ‘to mince our
words now and applaud the OAU would neither serve the desired purpose
of learning lessons from our past, nor reflect positively on our honesty and
integrity’.67 This initiated a difficult relationship which would haunt Eritrea
during the war with Ethiopia, the latter having much greater influence
within the organization, as well as being its permanent host. Relations with
Sudan quickly collapsed and were problematic for the first decade or more
of Eritrean independence;68 the rise to power of the National Islamic Front
in 1989, and the unfolding of its Islamist agenda, created severe strains in the
relationship with independent Eritrea in the early 1990s.The EPLF, after all,
had its roots in no small part in opposition to Sudanese Arabism and
Islamism, and the two countries were soon providing succour to one anoth-
er’s rebels and political opponents in order to gain leverage over their inter-
nal and therefore their external affairs. Eritrea sponsored or supported the
ongoing rebellions in southern and eastern Sudan through the 1990s, and
would later do the same in Darfur; Sudan provided bases and ideological
sustenance to jihadist groups operating in the northern mountains, seeking
the overthrow of the EPLF state.
Meanwhile, Eritrea clashed with Yemen over the Hanish Islands in 1996,
although this was resolved by outside arbitration;69 there were sporadic ten-
sions with Djibouti, with whom diplomatic ties were cut during the 1998–
2000 war, and there is a military standoff along a contested stretch of border
at the time of writing.70 There was adventurism in Somalia—after 2000, this
was a means of instigating trouble on Ethiopia’s southern flank—although
the scale of the Eritreans’ involvement was somewhat smaller than that of
Ethiopia. In the years after 2002, when the boundary commission published
its findings, Eritrean condemnations of the US, the EU, and ultimately the
UN—all of whom Asmara saw as complicit in failing to place pressure on
 new state s, ol d war s 233

Ethiopia to abide by the ruling—became ever more shrill. Relations with

the US in particular reached a nadir by 2007,71 when Washington—
convinced, as it remains, that the Eritreans were essentially arming Somali
terrorists—declared that it was considering placing Eritrea on the list of
state sponsors of terrorism. At any rate, after vigorous lobbying by both
IGAD and the AU, the UN Security Council in December 2009 imposed
sanctions on Eritrea for its involvement in Somalia, an unprecedented expe-
rience for an African nation; Eritrea’s isolation, notwithstanding its growing
interest in friendships with such states as Iran, seemed complete. The sanc-
tions, predictably, were greeted with angry outrage in Asmara, and it seemed
that Isaias was at loggerheads with just about the whole of the ‘international
community’—and, moreover, that he was happiest in that position. Of
course, it might be pointed out that over the last twenty years Ethiopia has
hardly been more successful in terms of regional relations: a major war with
Eritrea, military involvement in Somalia, border problems with Kenya. In
that sense—so the Eritrean leadership would have it, at any rate—Ethiopia
is the centrepiece of the region’s instability. However, it is the Eritrean ‘zone’
which is in fact the epicentre of much of the area’s conflict, while Eritrea
itself is the most visible manifestation of the politics of the frontier. The
EPLF inherited not a nation but a historical fault line, a thin arc of con-
tested territory with a history of recurrent conflict and intrinsic instability.
This environment, of which the movement itself was a product, profoundly
influenced Eritrean foreign policy, and thus has had a major impact beyond
Eritrea itself; but it also influenced Eritrea’s domestic politics in ways which
were not, perhaps, foreseeable at the point of liberation—many had their
judgement clouded by euphoria—but which had become clear by the end
of the decade.The EPLF’s mistrust of the very space they had inherited, and
its knowledge of the region’s violent, unstable past, meant that they in turn
governed through a violent militarism which suggested that Eritrea’s own
people were apparently not to be trusted any more than the country’s
rapacious neighbours.
In other words, contemporary Eritrea—in both foreign and domestic
affairs—must be understood as much in terms of the politics of the fault
line as of the failures of the post-colony. It is the case that recent analyses of
Eritrea’s increasing difficulties have tended to focus on the authoritarian
proclivities of the EPLF;72 the country’s predicament, in sum, is the product
of modern misgovernment, and is rooted in the political failures of the
EPLF itself, and even more specifically, in the personality of Isaias Afeworki.
234 revolut i on s, l i b e rat i on s, and g h o st s

The nation may be battered from above, but it remains intact, and the cen-
tral assumption is that a nation does indeed exist, or at least that there is a
nation to be had in the future, given the requisite political guidance. These
analyses are of course valid in their own right; but here I propose an alterna-
tive approach, related to historical experience and resultant political culture.
Eritrea’s politics, again, are those of the borderland, and the outcome of
long-term instability; the pressures of the frontier zone, dating to the nine-
teenth century and perhaps earlier, have produced a politically unstable
environment and a restless militarism. Eritrea cannot simply be understood
to have ‘failed’ in the conventional post-colonial sense. The roots of its fail-
ure, rather, lie in its troubled and unstable history as a violent frontier zone,
and the political tectonics which have shaped responses to particular prob-
lems. It was an extraordinarily fertile frontier, too; and that fertility—the
vitality of violence, if we prefer—is critical to understanding the region’s
modern history.
Ethiopia remains the sine qua non of the Eritrean state. Ethiopia has been
differently imagined by different sets of people at different points in time,
but in the end no satisfactory solution has been reached in Eritrea about
how to deal with either Tigray, or the Amhara, or the putative ‘empire-state’
as a whole. The EPLF has not yet developed a coherent or persuasive strat-
egy—either at home or abroad—which is not at heart anti-Ethiopian. At
the same time, of course, it must be remembered that Ethiopia does indeed
constitute a ‘threat’, however distorted or exaggerated by the government in
Asmara; multiple strands of what we might call the Ethiopian ‘political
establishment’ do not accept Eritrean independence in its current form, and
there is residual hostility from the TPLF and Amhara elites alike. In other
words, although much blame for the troubled relationship might be laid at
the feet of the EPLF, there is a very real and complex set of problems to be
addressed which requires a degree of political creativity, flexibility, and
imagination of which the current government in Asmara seems increasingly
incapable. Yet in many respects Ethiopia is only the gatekeeper of sha’abiya
anger: beyond it lies the global hinterland. Ethiopia reminds Eritrea of how
the latter has been habitually betrayed and misunderstood; the repeated
failure of the international community to do its duty, or even to pay atten-
tion—in the 1950s, in the early 1960s, during the struggle itself, and latterly
during the 1998–2000 war and its immediate aftermath—has persuaded the
leadership that Eritrea is always the victim of international intrigue and
ignorance. National neurosis, perhaps, but unfortunately it is grounded in
 new state s, ol d war s 235

some measure of historical reality—and needs to be taken rather more seri-

ously by foreign representatives charged with dealing with the Eritrean gov-
ernment. Thus the logic of the militarized isolation of Sahel, and beyond:
the fortified and undemarcated border represents Eritrea’s besieged and
beleaguered position in a hostile world. Much of this also explains the
export of the frontier war; the ‘frontier on the move’ is manifest in Eritrean
involvement in Somalia, and Darfur, and southern Ethiopia, policies aimed
at undermining both Sudan and Ethiopia. In this sense the EPLF is a form
of super-shifta, its foreign policy representing shifta rampant. The move-
ment’s regional adventurism is a reflection of the restless violence of the
historic Eritrean frontier, and at least in part symptomatic of its frustration
with and indeed contempt for the ‘nation’ it supposedly inherited.
The rise of the Eritrean armed nationalist movement represented, ulti-
mately, an attempt to achieve mastery of the frontier through violence. It was
not, in the end, ‘revolution’, but evolution, in which the Tigrinya, not unnat-
urally, emerged pre-eminent, even if that was not how the liberation war
itself began; the militant identity thus created is the outcome of an attempt
to resolve a historically uncertain and ambiguous state of affairs, born of
mistrust and insecurity. The EPLF has attempted, further, to make the fron-
tier a moral state, to weave a sense of morality around the experience of the
frontier, driven forward by the Front’s own sense of militarized morality and
‘revolutionary’ zeal. In sum, it was a process which involved making a moral
virtue of being Eritrean, rooted in a sense of bitter experience on the con-
tested frontier, with all its attendant sacrifice and suffering, the critical moral
elements of the frontier people. In effect, the EPLF represented a form of
militarized moral mastery of the frontier;73 such mastery was not, as the
EPLF had it, in itself revolutionary—although clearly it comprised novel and
innovative elements—but rather profoundly evolutionary. Arguably the most
novel element, in fact, was the brutal imposition of a monopoly on the use
of violence across the Eritrean frontier zone, which was essential if the EPLF
was to succeed in ‘modernizing’—i.e. creating a functioning nation from—
the contested space which it had captured. The Eritrea that the EPLF cre-
ated, then, involved the modernization and institutionalization of a frontier.
However, it seems clear that even while striving for mastery, the Front did
not truly accept the viability of its estate, and certainly it could not take for
granted its stability; the roots of modern authoritarianism, thus, lie in the
long-term violent struggle to achieve control over an inherently unstable
environment—both internal and external.
236 revolut i on s, l i b e rat i on s, and g h o st s

Oromia
The OLF withdrew from the coalition in 1992, and shortly afterwards
resumed its struggle for ‘Oromia’, the imagined homeland of the Oromo
people. The profound divisions within the Oromo liberation movement
notwithstanding, the exponents of the cause became ever more articulate,
and were increasingly supported by an academic establishment in exile,
determined to pursue its revision of the Ethiopian past—essentially an exer-
cise in the demythologizing of the Menelik empire and its successors. Some
of this work was more polemical than scholarly, but it was all aggressively
revisionist.74 The situation was horrendously complex, however: the very
shape of the new federal state of Oromia indicated the scale of the problem,
for it was flung like a splash of paint across the south and centre of the repub-
lic, wholly lacking in the geographical logic of Ethiopia, or Tigray, or even—
for all its political and clan-related complexity—the Ogaden. Millions of
Oromo, moreover, lived at relative peace with the Ethiopian state, unwilling
to embrace the renewed armed struggle advocated by the OLF and the
myriad other Oromo movements which soon emerged. But the fact remained
that the withdrawal of Oromo fighters back to the proverbial bush was per-
haps the most visible manifestation of the failure of the EPRDF’s ethnic
federalist experiment. At the time of writing, those fighters—of whatever
faction—are some way from overthrowing the state, or anything approach-
ing it; their main areas of activity are Wellega in the west, to some extent in
Gambella region, and in southern Sidamo close to the Kenyan border. But
they do have the potential to represent the single most dangerous internal
challenge to the federal state, not least in that the Oromo nationalist move-
ments purport to represent close to half the country’s total population.Those
who are contemptuously dismissive of the Oromo military challenge are
missing the point, for as long as it exists, the Federal Democratic Republic of
Ethiopia is inherently unstable. As the 1990s developed, Oromo nationalists
situated their struggle in an ever more sophisticated historical context; the
basic thesis, however, was relatively straightforward, viz. that Ethiopia was an
imperial state created by Menelik, and the Oromo were the largest of a
number of ‘colonized peoples’ under the habesha yoke since the late nine-
teenth century.75 While the ‘new’ EPRDF regime made some effort to con-
trol Oromo hearts and minds, largely through the illusion of ethno-federalism
and political partnership—from the mid-1990s, there was a significant
 new state s, ol d war s 237

Oromo presence in the Ethiopian government—the proliferating nationalist

movements advocated its violent overthrow and talked of redress for decades
of persecution and oppression of the Oromo at the hands of the Amhara-
Tigrayan political establishment.
The issue for Oromo activists has been one of unity. Already from the
late 1980s the OLF was in competition with the Oromo People’s Democratic
Organisation (OPDO), formed in 1989 by a group of Oromo who had
been cooperating with the TPLF. The OPDO accepted the pan-Ethiopi-
anism of the EPRDF, while the OLF—which had long had uneasy relations
with other Ethiopian guerrilla organizations—situated its struggle solely in
terms of Oromo rights; certainly, at the very least, the OLF demanded that
it be recognized as the sole representative of the Oromo people. When this
was not forthcoming, and the TPLF favoured the OPDO, the OLF with-
drew from the coalition, declaring that the OPDO was a government pup-
pet organization which did not represent the interests of the Oromo
population. In the years that followed, the OPDO was vigorously promoted
as the Oromo ‘official partner’ in the new political dispensation, and any
Oromo suspected of supporting rival movements, including the OLF, faced
systematic persecution.76 In 1997, talks between the government and the
OLF broke down, the former inviting the latter back into its fold, but the
latter stipulating that power in Oromia should be transferred from the
OPDO to the OLF, something which the government could not counte-
nance. There was a faith-based element, too, for the OLF leadership—pre-
dominantly Muslim—believed that Addis Ababa’s US-assisted military
operations against Islamic extremism (mostly Somali) in the mid-1990s ulti-
mately rendered any serious dialogue between the secessionist OLF and the
EPRDF pointless. The OLF was branded a terrorist organization bent on
overthrowing the legitimate political system by force—as the EPRDF itself
had done, but it seemed it was a matter of interpretation and context—and
the state security apparatus was brought to bear on an Oromo population
toward which the state itself remained suspicious and hostile.77 The OPDO
itself was subject to continual monitoring—there were major purges of its
membership and leadership, for example, in 1997 and again in 2001, when
former Ethiopian president Negasso Gidada was expelled—which was
clearly important for the government, as it was through the OPDO that
‘Oromia’ would be administered, from the new federal capital at Adama
(formerly Nazret). Sporadic public protests by Oromo activists were dealt
with brutally; by the late 1990s and early 2000s, human rights organizations
238 revolut i on s, l i b e rat i on s, and g h o st s

were reporting horrific stories of abuse of Oromo suspects in detention,

with often entire communities being targeted owing to supposed OLF links
and women in particular being subjected to sexual assault.78 At the other
end of the political scale, senior Oromo political leaders—accused, invaria-
bly, of ‘narrow nationalism’—continue to be arrested and detained regularly
at the time of writing.
Yet the response to ‘state terrorism’ has been splintered, and again the
Oromo cause, if it can be thus singularly rendered, has been undermined by
a chronic disunity which makes even the Eritrean opposition groups appear
relatively cohesive. The proliferation of movements through the 1990s
caused such consternation within the broad church of Oromo activism that
various groups came together in 2000 to form the United Liberation Forces
of Oromia (ULFO).79 Strains quickly emerged, and by 2006 the movement
had collapsed, the trigger being the formation of the rival Alliance for
Freedom and Democracy (AFD), in which the OLF—apparently long sus-
picious of the rise of the ULFO—was involved. Confusingly, however, there
were now three distinct movements using the nomenclature ‘OLF’; there
was also FIDO (the Front for Independent Democratic Oromia), UOPLF
(United Oromo People’s Liberation Front), FIO (Front for Independence
of Oromia), and COPLF (Council for Oromia People’s Liberation Front).
By the time the Oromo Forum for Dialogue and Reconciliation emerged
in exile in 2007–8 in an attempt to at least provide a context for open dis-
cussion, it was clearly badly needed, and there was more than a hint of
desperation in the preamble: there needed to be an end to ‘infighting, mis-
trust and innuendos’, and a new unified struggle to halt ‘the escalation of
the suffering of the Oromo people and other colonized peoples at the hands
of the TPLF minority government’.80 To anyone acquainted with Ethiopian
and Eritrean liberation politics in the 1970s and 1980s, it was a familiar tale.
Meanwhile the actual numbers involved in the field of combat on the
Oromo side had dropped from perhaps 15,000 in the early 1990s to some-
thing around 5000–6000.
The Eritreans would increasingly if surreptitiously lend their support—
certainly moral, and allegedly material—to the cause, especially after 1998.
The irony, but also the motivation, was that tens of thousands of the bodies
scattered across the Eritrean battlefields of 1998, 1999, and 2000 belonged
to Oromo soldiers in the Ethiopian army. With the eruption of the war in
mid-1998, there had been a massive recruitment campaign across the south,
especially in impoverished, rural Oromo communities; they were to become
 new state s, ol d war s 239

cannon-fodder, in the true Napoleonic sense, although many joined up

because of the opportunities promised to them by recruitment officers.81
Stories abounded of Oromo soldiers being marched over minefields in
order to clear the way for more elite, glory-hunting habesha units; of Oromo
units being deployed in hopeless or static situations, while Amhara and
Tigrayan units were given rather more attractive postings; of Oromo being
ill-trained and ill-equipped, and being deployed only to draw fire; or of
Oromo dead being abandoned in mass graves by retreating Ethiopian forces
without any but the most basic of markings, while Tigrayan and Amhara
soldiers and officers were treated rather more carefully, often given indi-
vidual graves or being carried back.82 Whatever the precise truth of these
stories, their very existence was suggestive of brutal ethnic differentiation
within a military establishment now under new management, and of trou-
bled attitudes more broadly. Disenfranchised, rural Oromo youth—as much
a socio-economic group as an ethnic one—clearly provided EPRDF offic-
ers with a source of cheap and replaceable manpower, although perhaps
their recruitment might also be seen as a form of the political co-option
necessary to the great national (re)building of the 1990s. Nonetheless, where
the project itself was evidently failing, the EPRDF reverted to the brutal
control and repression of Oromo communities well learnt from earlier
regimes.
In that sense, the great EPRDF-led nation-building project—supposedly
greatly spurred by the 1998–2000 war—was a resounding failure, riddled as
it was with deep-seated tensions.83 While the Eritrean war was meant to
galvanize Ethiopian patriots and enable the government to appeal to an
ancient Ethiopian nationalism, it was clear that millions of Ethiopians—
Oromo prominent among them—could scarcely endorse such a project.
Many Oromo leaders themselves rejected completely the nationalist rheto-
ric deployed during the war, perceiving the conflict, rather, as an exclusively
northern affair between Tigrinya rivals, which Meles was now manipulating
to consolidate the TPLF’s position in Ethiopia more broadly. From this per-
spective, arguably, the Eritrean war actually heightened, rather than reduced,
simmering ethno-regional tensions within Ethiopia—because it led inevi-
tably to the questions, whose war is this and who stands to benefit? Regardless
of the inefficacy of the Oromo armed struggle and its political weaknesses,
the fact remained that ‘Oromia’—both as a concept and as a place—was also
a fertile frontier within which creative new identities were being cultivated,
and along which the true nature of the EPRDF state revealed itself.
240 revolut i on s, l i b e rat i on s, and g h o st s

It remains a frontier which may yet come to redefine the very polity of
Ethiopia itself.

Continuities and transitions: Somalia and Ethiopia

The disintegration of Somalia between 1988 and 1991 stood in stark con-
trast, so it initially seemed, to the progressive state-building projects in
Eritrea and Ethiopia in the early 1990s. It is not my purpose here to explore
the Somali situation in the kind of depth which is in any case much better
left to specialists;84 rather, my aim is to briefly outline the manner in which
Somalia came once again to constitute a major frontier of violence for both
Ethiopia and Eritrea, though especially the former, and involved the con-
tinuation of the 1998–2000 war by proxy. While the Ethiopian involvement
in Somalia is very much on the public record, as it were—even if they reject
some of the accusations made regarding motive and behaviour—it should
be pointed out that the Eritreans continue to reject allegations that they
are involved, beyond offering ‘moral support’ for the ‘Somali struggle’.
Throughout the 1990s, the EPRDF government had naturally taken a close
interest in the violence playing out along Ethiopia’s historic southern fron-
tier, with security operations being periodically conducted along the Somali
border while the ‘warlordism’ of the Hawiye clans in southern central
Somalia descended into violent competition over (as much as anything else)
commercial opportunities. Between the fourteenth and the sixteenth cen-
turies, Christian highlanders had forged statehood in struggles with Somali
Islam; it was an ongoing contest in the nineteenth and early twentieth cen-
turies. Now, Addis Ababa had the full support of the US in its mounting
battle against Islamic extremism, and in particular against the fundamentalist
al-Ittihad movement. As a ‘failed state’—in the conventional sense of that
increasingly utilized catchphrase—Somalia was seen as a fertile breeding
ground, and a hiding place, for Muslim extremism, especially after September
2001, when US intelligence detected the presence of al-Qaeda cells there.85
The US and Ethiopia had shared objectives in this respect throughout both
the Clinton and the Bush administrations, even if they differed over
strategy.
Ethiopian policy through much of the 1990s was to support certain key
‘secular’ warlords, notably Abdulahi Yusuf, in opposition to the Islamic
 new state s, ol d war s 2 41

groups; in the meantime, repeated reconciliation conferences were held in

the attempt to bring together various Somali factions, usually in vain. The
last of these gave rise to the Transitional National Government (TNG) in
2000.86 This ‘administration’ lacked support from certain key Mogadishu
warlords, while its Islamist leanings soon provoked Ethiopian hostility, and
shortly another peace progress—led by the regional Intergovernmental
Authority on Development (IGAD)—was in motion, ostensibly to recon-
cile the TNG with its Ethiopian-backed opponents. By the end of the proc-
ess in 2004, however, the TNG had largely evaporated. Meles Zenawi was
instrumental in the international negotiations leading to its replacement,
the Transitional Federal Government (TFG), whose parliament (meeting in
Kenya, not on Somali soil) elected Abdulahi Yusuf as president. The
Ethiopians have long been forced to defend themselves against allegations
that they bribed the parliament to do so. Increasingly, the TFG was some-
thing of a fiction, desperately casting around for wider Somali support;
meanwhile, the Union of Islamic Courts (UIC), an umbrella organization
for various loosely affiliated interests, including both moderate and more
radical wings of political Islam, was becoming ever more influential and
popular across southern Somalia. By early 2006, supporters of the UIC were
increasingly hostile to the warlords, and especially some warlord factions’
dominance of the profitable Mogadishu port business.
Clearly, external intervention further complicated the mire of Somali
politics and allegiances.The Ogaden, of course, had long been a natural field
of intrigue, dating to the era of the liberation struggle.The Ogaden National
Liberation Front (ONLF), dating to 1984 when it took up the struggle of
the earlier Western Somali Liberation Front, had long been a key ally of the
EPLF in the region; Siad Barre had long hosted movements against the
Derg, at least until the agreement of 1988, and both TPLF and EPLF leaders
had travelled on Somali passports. The ONLF’s main support was drawn
from the southern Ogadeni clans, while the major northern clans, as well as
the non-Ogadeni clans living in the Ogaden, including the Hawiye and
Isaq, opposed the movement. Supported by the EPLF, it fought the Derg for
several years before (briefly) enjoying legitimate status as a political party in
the early 1990s; it followed the OLF out of the coalition, however, and
returned to the bush, in the process absorbing a number of Muslim Oromo
recruits.After independence, Eritrea sought further opportunities in Somalia:
in 1998, it struck up a short-lived relationship with the key warlord Hussein
Aideed before the latter transferred his loyalty to Addis Ababa. Eritrea also
242 revolut i on s, l i b e rat i on s, and g h o st s

continued to support the ONLF, with rumours emerging by 2005 that

ONLF fighters were being trained in Eritrea itself. Occasional bomb out-
rages in Addis Ababa and attacks on oil installations notwithstanding, the
‘success’ of the ONLF was limited—not least because of its apparent inabil-
ity to develop broad support within the newly named ‘Region 5’ of the
federal Ethiopian state. A complex interplay between Ethiopian and Somali
authorities in the Ogaden had led to something approaching a stable politi-
cal settlement there, or at least one in which ‘resistance’ to Ethiopian ‘occu-
pation’ had been largely neutered.87 Nonetheless, the Ogaden remained a
key area of radical political activity, enabling Ethiopian forces to justify
brutal security operations there,88 and the region was a natural point of
leverage against Ethiopia for Eritrea; when the UIC took control of
Mogadishu in mid-2006, it greatly opened up Eritrea’s access to the Ogaden
and the ONLF.
Eritrea’s problem—and, of course, Ethiopia’s—was the growing Islamism
within the region. Asmara had long discouraged the growth of Islamism in
the ONLF, but it nonetheless developed links with radical Islamic leaders
within the UIC, notably Sheikh Yosef Indohadde and Sheikh Hassan Dahir
Aweys, and was reportedly training and arming UIC fighters. In sum, the
profoundly fragmented ‘arrangement’ of a besieged and ineffectual TFG
surrounded by an array of warlords was at least partially improved through
the rise of the UIC, which briefly, in 2006, governed the southern region of
Somalia rather more effectively than any administration since the end of the
1980s. An apparently popular revolt (precipitated by a long-running busi-
ness rivalry) saw the warlords expelled from Mogadishu in mid-2006. The
UIC itself was divided between moderates and hardliners, but it was the
militant Islamist wing which held sway, for now. Their ‘mistake’, however,
was to make threatening noises in the direction of Ethiopia, which feared
both shari’a and jihad on its southern flank, a much more serious threat, in
its way, than that posed by the Eritreans in the north. At the end of 2006 and
the beginning of 2007, Ethiopia invaded. The US had reportedly advised
against such a risky adventure, warning Meles that Somalia would become
‘Ethiopia’s Iraq’, but once it had begun they supported it, materially and
morally. And at first, Ethiopia seemed vindicated: the UIC forces scattered
into the bush, and Ethiopian troops occupied Mogadishu with relative ease.
But it was a remarkably wrongheaded war, apparently ignorant of history;
few Somalis of whatever clan, or political or religious persuasion, would
welcome Ethiopian troops as ‘liberators’ and guarantors of order, and a
 new state s, ol d war s 243

violent backlash was entirely predictable. Ethiopia was the ancient enemy,
and radicalized Islam fuelled the violence of the response.
The insurgency grew in strength—again, the Eritreans were allegedly
supporting it with small arms and basic training, although they refuted the
charge—and the Ethiopian forces were increasingly beleaguered in their
bases. Civilian casualties spiralled across southern Somalia, and a humanitar-
ian crisis of shocking proportions loomed; meanwhile, Ethiopian troops
were reportedly involved in appalling abuses of civilians and suspected fight-
ers alike,89 and did little to win over proverbial ‘hearts and minds’—however
unlikely that would have been in any case.There were reports that Ethiopian
forces regularly and deliberately shelled civilian areas of Mogadishu in
response to insurgent attacks, wiping out entire districts in what human
rights organizations described as ‘war crimes’. While for much of 2007
occupying Ethiopian forces reportedly behaved in a restrained and disci-
plined manner, in 2008 eye-witness accounts abounded of troops looting
and pillaging, raping women, and carrying out on the spot executions, as
well as firing live rounds into crowds of civilians and attacking mosques.
There were also ‘credible reports’ that the Ethiopian military was diverting
urgently needed food aid from drought-affected areas in order to flush out
rebels.90 There was nothing novel about such tactics, for they had been seen
before, most recently in the tactics adopted by the armies of the Derg. If
anything, however, atrocities in Somalia (as also in the Ogaden) took on the
appearance of culture- and race-war, aimed at the decimation of the very
economic and cultural bases of Somali existence.
Addis Ababa engineered the creation of a new administration, as ‘transi-
tional’ and as ineffectual as that overthrown by the UIC, against the back-
drop of mounting catastrophe. The invasion stoked up nationalist as well as
Islamist hatred of the government in Addis Ababa. Somali Islamist insur-
gents—known popularly as al-Shabaab, literally ‘the youth’—escalated their
attacks through 2007 and 2008. There was a certain irony in Eritrea’s sup-
port, as the government had long had problems of its own with Islamic
extremism, and a wave of Islamism across the region was hardly in Asmara’s
long-term interests—but from Eritrea’s point of view the policy was both
logical and consistent.91 The EPLF saw not allies but short-term interests
and opportunities, a chance to undermine Ethiopia along a new front; it was
decidedly not in Eritrea’s interests to see Ethiopia succeed in Somalia. But
the outcome was increasing international isolation—ironic, again, given
that Eritrea’s Somali policy was driven by the same self-interest and concern
244 revolut i on s, l i b e rat i on s, and g h o st s

for self-defence that had motivated Ethiopia’s aggression. The US, again,
identified Eritrea as behaving in the manner of a state sponsor of terrorism,
and indeed Asmara had become the meeting place for various strands of
Somali opposition—UIC leaders, some warlords, elements of the former
transitional administration. Most importantly, from the Americans’ view-
point, was Eritrea’s hosting of the Alliance for the Re-Liberation of Somalia
(ARS), which included Sheikh Hassan Dahir Aweys, a leader of the al-Itti-
had Islamist group, and named in the UN list of persons linked to terrorism;
its secretary-general was Sheikh Sharif Sheikh Ahmed, a key religious leader
but regarded as something of a moderate. Eritrea supported the ARS in its
refusal to work either with the temporary Somali government or the
Ethiopians, and many of its members stayed away from the Djibouti confer-
ence in 2008 which brought some Somali groups together for preliminary
negotiations. Indeed, the fact that Djibouti was treading on Eritrean ‘politi-
cal territory’, as it were, was almost certainly one of the factors behind the
confrontation which took place on the Eritrean–Djiboutian border that
same year. Eritrea’s isolation seemed complete at the time of writing in
mid-2009 when the AU took the dramatic and unprecedented step of call-
ing for international sanctions against Eritrea in view of the latter’s support
for Somali insurgents. These were imposed by the UN Security Council in
December 2009.
Meanwhile Ethiopia became increasingly frustrated by the inability of
the Somali administration to settle even their internal differences, never
mind cope with the growing insurgency; Meles made threatening noises
about pulling out if Ethiopia did not receive the anticipated international
support, or indeed the gratitude which he believed he was due for ‘dealing
with’ the Somali crisis. But direct UN intervention was ruled out, and only
a token AU force was made available. It was enough for the Ethiopian gov-
ernment, which declared its mission achieved, and withdrew its troops at
the beginning of 2009. Ethiopian forces remained close by, however, and
reports soon emerged of military operations just inside the Somali border,
activity which will surely continue for the foreseeable future. With the
Ethiopian withdrawal, and a new Somali administration under former UIC
and ARS leader Sheikh Sharif Sheikh Ahmed in place, Aweys (dubbed the
‘kingmaker’ by foreign journalists) returned to Mogadishu and there is, at
the time of writing, a belief that Somalis are on the threshold of their best
chance at a lasting political settlement for nearly two decades. But it is only
a threshold: while it is hoped that increasing numbers of al-Shabaab might
 new state s, ol d war s 2 45

be lured into a ceasefire by a power-sharing arrangement with key Islamist

leaders, in the meantime violence between government and Islamist forces
continues.92 There is clearly a long way to travel, while the conflict on land
spills over onto sea lanes in the Gulf of Aden and Indian Ocean in the form
of Somali piracy.93 For perhaps obvious reasons, the upsurge in piracy has
captured international attention, but it is ultimately only a reflection of the
turmoil—and of economic collapse—on land.94 Ethiopia and Eritrea,
moreover, will continue to harness whatever forces within Somalia they see
fit to further their own national interests—lessons well learnt from the era
of liberation struggle.
Mention should be made, finally, of the self-proclaimed Republic of
Somaliland, the territory of former British Somaliland, which has both dis-
tanced itself from the mayhem further south and made significant progress
toward independent statehood.95 Its campaign for international recognition
has been as yet unsuccessful, for a range of reasons, some more difficult to
understand than others; but to all intents and purposes Somaliland is already
a functioning state, and the government in Hargeysa has a powerful claim to
sovereign status which it must be hoped will be accepted sooner rather than
later. Somaliland is by no means wholly free of the disturbances which have
defined the south,96 and it is not quite the haven depicted by some of its
more ardent supporters; but it has established a relative stability much appre-
ciated by Ethiopia which has developed a close relationship with its neigh-
bour. At the time of writing, Ethiopia is reportedly preparing to champion
Somaliland’s cause, which will prove an extremely interesting and important
development for the region. There can be little doubt that it is very much
in Ethiopia’s interests to have a ‘friendly’ Somali regime so close to hand,
while it also further undercuts Eritrea’s influence in the region; on this fron-
tier, if on no other, Meles’ strategy is proving markedly superior to that of
Isaias, who would have done rather better to cultivate his own relationship
with a government in Hargeysa that may yet play a major role in the stabi-
lization of the region.
 Epilogue
Armed Frontiers and Militarized Margins

F rom the early 1990s, there emerged across the region an extraordinary
patchwork of conflict. Axes of violence and strategic machinations
intermeshed and pulled peoples and places together in new ways. In truth
some of these conflicts began as localized struggles and crises, and were not
initially connected to others in the region: the collapse of Somalia at the
beginning of the 1990s was one such example, having its causes and its own
dynamics. But given that one of the key themes of the region’s history in
the second half of the twentieth century was the enlargement of scale in
terms of violence, it was inevitable that such local battles should become
others’ battles, too. Each war represented someone else’s foreign policy
opportunity and formed part of a war being fought for other reasons some-
where else in the region. The Eritrean–Ethiopian war, various insurgencies
in Ethiopia, in particular that in the Ogaden and Oromo territories, the
civil war in Sudan and the conflict in Darfur, the escalating violence in
Somalia: these were the interconnected frontiers of violence, the political
tectonics which had come to define the region.
This was not how it was supposed to be, of course: the end of the Cold
War was supposed to usher in a new era of democratic development, an era
which saw greater efforts toward the resolution of conflict and the applica-
tion of pragmatic solutions to long-standing problems—and all against the
backdrop provided by the triumph of liberal, democratic, capitalist values.
At the very least, there would be no superpower rivalry to spur on wars in
Africa. It was a remarkably naive, not to say curiously patronising, approach
to African warfare in the late twentieth century: it seemed that, in killing
one another as much as in any other sphere, Africans required some kind of
external inspiration.To many outsider observers, the death of the Siad Barre
regime in Somalia and subsequent violent chaos was an unfortunate anomaly,
 e p i log ue 2 47

and indeed reminded many of just how savage Africa might be, and had
been; still, the US-led intervention would hopefully go some way to restor-
ing order. And, more importantly, further north there were apparently
reasons to be optimistic. In Isaias Afeworki’s EPLF and Meles Zenawi’s
EPRDF, the region appeared to have organizations capable of engendering
stability and political and economic development. The term ‘pragmatic’ was
the one most often used to describe Isaias and Meles themselves—and cer-
tainly, the EPLF and EPRDF respectively appeared to offer solutions to
problems which had plagued the Ethiopian region since the eighteenth
century. One was the issue of the frontier zone of Eritrea and access to the
coast, and the other was the question of ‘nationalities’ within Ethiopia itself.
Events in Somalia no doubt fed the West’s inertia when it came to the mass
killing in Rwanda, and together the two crises darkened the international
mood in the first few years after the supposed ‘end of history’. Nonetheless,
the optimism was still there in 1998 when US President Clinton declared
his belief in an ‘African renaissance’, within which were included Isaias
Afeworki and Meles Zenawi.
To suggest that such optimism that the end of the Cold War produced
was misplaced would not merely be an understatement, it would in itself
be a chronic misrepresentation of reality. This was an optimism which was
underpinned by a criminal ignorance of the dynamics of the region’s his-
tory and political processes; whether such ignorance was wilful is some-
thing which will only become clear to the historians of the future, but for
certain it was born of a remarkable arrogance.The West saw what it wanted
to see in the mid-1990s, and what it saw served its interests. Neo-imperialism
now came in the guise of free-marketism and aggressive humanitarianism.
The notion that Africans could now stop fighting because ‘liberal capital-
ism’ had supposedly won the last great battle of history was redolent of the
same conceit which had driven the partition of Africa a century earlier. By
its very nature it was a marginalization of Africans from the engines of
change: they would be the dumb recipients of wisdoms developed in
another place. Isaias Afeworki and Meles Zenawi were indeed pragmatic,
but not quite in the way that Western governments had hoped: for these
men, war was now a matter of unfinished business, an extension of national
policy, and something over which their new-found sovereignty allowed
them complete control. Liberation war would have multiple legacies in the
region, but one of them was clearly the institutionalization of violence in
the new states of Ethiopia and Eritrea. In both countries, governments
which were led by and to a large extent comprised of former guerrillas
248 e p i log ue

could draw on long histories of violence and use these to craft notions of
national destiny.
In the case of the ‘new’ nations, which both Ethiopia and Eritrea eventu-
ally became, these were not merely militarisms which were mutually antago-
nistic, but rather they were in many respects actually defined one against the
other. The peculiar political and cultural systems of the Eritrean state can be
explained by the fact that it was built on a historical fault line—and one
which had come to fundamentally destabilize the entire region.The Ethiopian
state was also profoundly militarized, and defined by its frontiers, viz. the
pointed celebration of Adwa in March 1999, just days after the Ethiopian
army’s success in capturing Badme.This was an exercise in telescopic remem-
brance, in which Isaias Afeworki took the place of Oreste Baratieri. Regardless
of the enemy, the 1999 commemoration of Adwa provided a meaningful
glimpse into the role of militarism at the heart of the Ethiopian polity and
public life. The epicentre of the region’s instability lay along the Eritrean–
Ethiopian border, and thus the roots of much modern conflict lay in the
history of troubled relations between the EPLF and the TPLF; but they were
also to be found in the difficult relationship between newly independent
Eritrea and Sudan. Certainly it might be argued—and some in the region
have been only too happy to make the argument—that much of the region’s
instability stems from the emergence of one of the most successful armed
liberation movements in Africa, and perhaps anywhere in the world, namely
the EPLF. The EPLF has had a profound impact on the political shape of the
region, and thus it may be possible to see Eritrea as the pivot around which,
and through which, so much contemporary violence occurs. The very pres-
ence of Eritrea has affected the political dynamics of the region; the EPLF
itself is a complex, ambitious, and—until very recently—highly effective
organization. It has had considerable regional reach. Yet the emergence of
Eritrea is only part of the story, because of course the contemporary history
of the region has been shaped by the ‘success’ not only of the EPLF, but also
of the TPLF in Ethiopia and the SPLM in Sudan. Nationalism and ethnicity
remain, at the time of writing, powerful shapers of regional destiny—but
religion has also long been significant and, likewise, remains so today, perhaps
more than ever. In particular, the clash between Christianity and Islam is
potentially more violent than previously.
Above all, perhaps, the period between the 1970s and the early twenty-
first century has witnessed a dramatic expansion in the scale of ‘frontierism’,
especially since the early 1990s. It is the interconnectedness of various
 e p i log ue 249

frontiers of violence across the region which is the single most important
manifestation of ‘modernity’ in recent years. In particular, frontiers by proxy
have opened up, or perhaps more aptly, frontiers have been co-opted within
the region by parties wishing to extend their own confrontations in new
ways. Ethiopia’s hosting of a range of Eritrean opposition movements, and
multi-faceted Eritrean involvement in Somalia, eastern Sudan, Darfur, and
inside Ethiopia itself, has created a set of what we might term ‘virtual bor-
derlands’, and the proliferation of these has been much more the product of
the organic, restless militarism of the region—dating at least to the nine-
teenth century—than the result of any kind of transformative modernity or
external intervention, either during the Cold War or after.
North-east Africa, then, in common with virtually all zones of human
habitation, is characterized by peripheral areas with long histories of eco-
nomic distress, political marginalization and oppression, social dislocation,
and ultimately violent conflict.1 These regions are but rarely glimpsed in
contemporary reportage, unless they happen to advance on the centre;
and when they do receive attention, these militarized margins often attract
humanitarian interest—usually relating to refugee flows and attendant
nutritional and medical issues—rather than attracting interest in the con-
flict itself, or even in the dynamics behind the marginalization of those
zones in the first place. The upshot is often that the wars being fought
there represent the grime on the underbelly of human progress, and seem
curiously pointless and endless; descriptions of them are sometimes redo-
lent of nineteenth-century depictions of endless cycles of borderland
violence. Thus, while the Eritrean–Ethiopian war attracted a flurry of
attention at the end of the 1990s, a host of ‘smaller’ conflicts were prolif-
erating across the region—many of them, indeed, related at various
removes to the Eritrean–Ethiopian war itself. Some were located in the
turbulent Ethiopian–Sudanese borderlands, involving a range of local
liberation fronts and rebel groups hoping to exert leverage on either Addis
Ababa or Khartoum or indeed both. Other groups again operated in the
southern extremities of Ethiopia, linked in different ways to the civil war
in Sudan, or to the Oromo and Somali in northern Kenya and eastern
Ethiopia. There was nothing new, for example, in the violent instability of
the western-Ethiopia–eastern-Sudan frontier zone; cross-border violence
had been endemic for much of the twentieth century. In this sense the
area broadly had much in common with the Eritrean frontier zone.
Nonetheless, owing to such specific factors as population density, and
25 0 e p i log ue

political and ethnic ‘visibility’, these militarized margins were as yet unable
to produce the kind of cohesive and effective militarism witnessed in
Eritrea. But their importance was not less for that.
Indeed, many of the conflicts which have been at the centre of this study
started out as ‘marginal’, fizzling into life on the peripheries of major states;
this is true, for example, of the modern Eritrean and Tigrayan insurgencies,
dismissed by successive Ethiopian governments as mere shifta. Struggling on
the edges, they strove to make themselves heard. Both Eritrean and Tigrayan
insurgencies, of course, were ultimately successful, marching on the centre
and capturing the state.Yet our concern here is those struggles which do not
‘succeed’ in quite this manner—or at least, have not done so at the time of
writing—and which have not captured the state, or even much attention,
save from NGOs, anthropologists, and of course internal security forces.
Here, violence became a way of life, and life was violence, however ‘low-
level’. Such violence was the product of the neglected place, often begin-
ning in the proverbial backwaters of nineteenth-century polities and early
twentieth-century colonial states; these were backwaters, the ‘wild places’
and frontier country of the fevered metropolitan imagination, often becom-
ing caught between colonial territories as buffer zones, scarcely governed
and ill-developed. The inhabitants invariably acquired bad reputations, as
warlike and troublesome. Such rough places could be treated accordingly by
state militaries, which could behave there in ways unthinkable elsewhere. As
we have seen, a good example is the Northern Frontier District in Kenya,
prone to raids from the Ethiopian side of the border, a frontier zone where
the far, diluted reaches of British authority in Nairobi met the southern-
most reaches of imperial Ethiopia. With shifting links across the border in
Italian Somaliland, to their episodic ‘brothers’ in the Ogaden and southern
Ethiopia, the inhabitants were regarded by the independent Kenyan gov-
ernment very much as the British had before them—wild, dangerous, unre-
liable in their loyalties, irreconcilable to modernity. Volatile and scarcely
visible, it was the proverbial militarized margin.
Conceptually, no doubt, these places are inevitable in human history, as
necessary to the functioning of society as any of the other brutal pillars on
which communities are raised; and they provide essential reference points
for states and metropoles in the processes of objectification which feed self-
image there. Such frontiers are often kept at arms’ length, whether they are
armed against an external enemy, or whether they serve as an ugly reminder
of what society once was at some indeterminate point in time—yet, as
 e p i log ue 251

I have argued throughout this book, such frontiers are both fertile and
primordial, the well-springs of both deep-seated insecurity and dynamic
creativity. These are areas characterized by a delicate balance of local
resources, notably access to water, healthy pasture, and arable farming land,
and they are areas which are extremely vulnerable to environmental shifts
which can have—and have had—catastrophic consequences. These are
zones where conflict is easily sparked, sucking adjacent states into them,
where economic war quickly becomes political, and vice versa.2 Our region
has long been crisscrossed by a series of these frontier zones, places where
there exist groups at odds with ‘law’ and ‘authority’, places which are no-
man’s-lands, but also places which are zones of transition, where cultures
and peoples meet, and new communities are formed.These have been places
of refuge as well as resistance, and new or hybrid social structures appear;
crucially, the states that persecute or neglect or lay claim to them have lim-
ited, if any, authority here, and as one travels further from the centres, the
oxygen of state power becomes ever thinner.
These have also been places to which states may export wars started
elsewhere—another key feature of our region’s history, but especially visible
in recent years. Eritrea has certainly diversified its responses to the ‘Ethiopian
problem’, by simultaneously courting the southern Sudanese in terms of
commercial and other links, and siding with Khartoum over such issues as a
UN force for Darfur and the ICC arrest warrant for Bashir himself. A ruth-
less pragmatism lies behind Eritrea’s ‘foreign policy’ initiatives in Darfur
itself, and in Somalia and among the Oromo; Eritrean realpolitik has involved
violent leverage in those spaces between states. While it is outside the scope
of the present work, the Darfur region exemplifies the point.3 Sudan under
Nimeiri had been a relatively consistent supporter of the EPLF; under
Bashir and the National Islamic Front from 1989, however, Sudan was an
enemy. Eritrea sought to exert pressure on Khartoum, which, it was believed
as early as 1994, was backing jihadist rebels against Asmara. Following the
severance of diplomatic relations between Eritrea and Sudan at the end of
1994, Isaias dispatched a mission to Chad in the attempt to persuade
President Deby to open a ‘Western Front’ against Khartoum; when Deby
demurred, Asmara had to content itself with supporting the SPLA in south-
ern Sudan, and the Beja guerrillas in the north-east, known as the Beja
Congress, composed mostly of Beni Amer. The latter were in alliance with
a smaller group, the Rashaida Free Lions, under the banner of the ‘Eastern
Front’.These were brought together, under Eritrean supervision, within the
25 2 e p i log ue

anti-Khartoum National Democratic Alliance, and Eritrean support for

Sudanese opposition movements intensified with Khartoum’s increased
assistance to Eritrean Islamist fighters—some of whom were trained and
armed by Osama bin Laden, then active in north-east Sudan. The irony of
Eritrea’s later support for Somali Islamists, thus incurring the wrath of the
US, clearly lies in the fact that the government in Asmara was doing battle
with bin Laden long before most of the world had heard of him. When the
Darfur rebellion erupted in 2003, the Eritreans were quick to take advan-
tage, offering support to both the various factions of the Sudan Liberation
Army (SLA) and the Justice and Equality Movement (JEM), while Asmara
became an important conduit for supplies to the rebels. Isaias, it seemed,
hoped to ‘manage’ the revolt and become the key intermediary between the
rebels and Khartoum, and indeed the wider international community, ren-
dering himself indispensible—hence his opposition to any kind of interna-
tional peacekeeping force, which would clearly have undermined his own
position. In a sense this was diplomacy not between ‘states’ per se, but between
powerful organizations—the EPLF, TPLF, NIF, SPLM—with access to the
resources and manpower and ultimately the leverage associated with states.
They were the products of political tectonics, and their diplomacy was that
of the fault line. For Eritrea, in particular, this was about the export of the
frontier war—whether to Somalia, or Darfur, or any other volatile place
where it might be successfully repackaged.
The Ethiopian–Sudanese borderlands also constituted a vast zone in which
violence had become an increasingly marketable commodity against a back-
drop of economic neglect; long a place of refuge for individuals and their
followers fleeing from either Khartoum or Addis Ababa, in the 1990s and
2000s it continued to be characterized by low-level violent insurgency4—and
a place where the true characters of the states involved were sharply revealed.
Note should be made in particular of the Gambella and Benishangul regions,
where the Gambella People’s Liberation Movement (GPLM) and the
Benishangul People’s Liberation Movement (BPLM) operated in defiance of
metropolitan authority. While Eritrean–Sudanese relations swiftly deterio-
rated in the early 1990s, the relationship between Khartoum and Addis Ababa
was rather more ambiguous: initially, the EPRDF sought more positive deal-
ings with Bashir’s regime—although, like the Eritreans, the Ethiopians were
also concerned with the export of Islamic radicalism from Sudan. The assas-
sination attempt on Egypt’s President Mubarak in Addis Ababa in 1995 seemed
to bear these fears out. Meles kept dialogue open with Khartoum, but simul-
 e p i log ue 253

taneously offered support to the SPLM who were able to operate from bases
inside the Ethiopian border. On the Sudanese side, Islamist foreign policy
directives emerged within the NIF under the leadership of Hassan al-Turabi
with the aim of cultivating clients in the border areas who were hostile to the
EPRDF. Thus the BPLM, with its major support from the Muslim Berti
group, fell under Khartoum’s influence, and by the mid-1990s was engaged in
‘irredentist’ (i.e. pro-Sudan) violence across the Benishangul region. Further
south, in Gambella, violent competition between the two dominant groups,
the Nuer and the Anuak, was the key dynamic: because a Nuer elite had
leaned toward the Derg, the Anuak-dominated GPLM had become a key ally
of the EPRDF forces in the region, a relationship which continued into the
1990s. But it collapsed when, amid allegations of corruption, Addis Ababa
instituted direct rule and effectively sidelined the GPLM; moreover, a 1994
EPRDF-run census concluded that the Nuer were the majority group in the
region, a finding angrily rejected by the Anuak who thus became ever more
alienated from the new Ethiopian regime.5
The outbreak of the Eritrean–Ethiopian war in 1998 signalled a new
phase in cross-border dynamics. Ethiopia was ever more anxious to repair
relations with Sudan, which was happy to oblige considering its own inse-
curities—and economic and strategic agreements involved the ending of, or
at least the reduction in support for one another’s dissidents and rebels.This,
for a time at least, weakened the position of both the BPLM and the OLF.
Nonetheless, the GPLM received Eritrean support from 2000 onwards.
Meanwhile the EPRDF created the joint Anuak–Nuer front, the Gambella
People’s Democratic Front, largely controlled by Addis Ababa; the Anuak
responded by forming the Gambella People’s Democratic Congress, whose
espoused aim was the expulsion of both Nuer and habesha settlers from the
region. Indeed tension in the area is multi-dimensional, with that between
Nuer and Anuak playing out against the larger issue of local groups’ hostility
toward highland (Amhara, Tigrayans, and Oromo) settlement in the region.
Most of these movements, however, were crippled by internal factionalism
and their impact, in the end, has been negligible. Nonetheless, Ethiopian
security forces carried out operations against the Anuak in late 2003, when
several hundred civilians were massacred, demonstrating the willingness of
the state to make its presence felt at the edge in the most violent manner
and to bring to heal recalcitrant groups.6
Coffee-rich Sidamo region, too, has found itself the focus of unwelcome
attention from central authority. Here, a tradition of violent resistance to
25 4 e p i log ue

interference from Addis Ababa—the peasant uprising in Gedeo in 1960 was

in response to the seizure of coffee-producing land by highland settlers, and
was brutally crushed7—evolved into the Sidama People’s Democratic
Organisation (SPDO) which took up arms against the Derg. The SPDO
became part of the Southern Ethiopian People’s Democratic Front (SEPDF),
which was itself part of the EPRDF coalition. From the early 1990s, however,
the attempt by central government to dominate politics in outlying areas such
as Sidamo, and the consequent persecution of Sidama groups which oppose
it, intensified considerably. In the early 2000s, much disturbance was centred
on the move by the government to alter the federal status of Sidamo, bringing
it under the regional government and thus under closer central control. In
May 2002, a demonstration against these proposals took place in Awassa, the
federal capital of Sidamo, and was apparently peaceful until security forces
opened fire, killing at least twenty-five people and wounding many others. In
the days that followed, scores of the demonstrators and others who supported
the demonstration were arrested and allegedly tortured while in detention.
While the southern region government set up an enquiry into the demon-
stration, the enquiry was not independent—in other words it was dominated
by central government—and no report was published.8 Further repercussions
followed in August 2002, when several leading figures in the SPDO were
dismissed from the party and arrested, accused by central government of
fomenting the May violence; however, the president of the SPDO, Girma
Chuluko, openly condemned the security forces for opening fire unprovoked,
and strongly opposed the change in Sidamo’s status.9 The SPDO has found
itself harassed and pressurized by central government since that time. The
alacrity and force with which the Ethiopian state imposes itself on frontier
zones—often in the pursuit of economic goals—has also recently been dem-
onstrated by the Gibe dam project on the Omo river. Here, local needs and
environmental concerns are swept aside in the name of state-led moderniza-
tion—in this case, increased electricity production.
Other small wars proliferate in contested frontier zones across the region—
often brutal at the local level, but fought beneath the horizon of visibility,
among groups long neglected and marginalized. Thus, Afar guerrillas—the
Afar Liberation Front, or the Red Sea Front—continue to operate in the
politically and physically hostile geography of the Danakil region, for exam-
ple, and might be utilized by the Ethiopian military should the need arise at
any point in the future; yet the Eritreans provide backing to other shadowy
Afar groups set against the Ethiopian government, including Ugugumo (‘rev-
 e p i log ue 255

olution’), and the Afar Revolutionary Democratic Front. Finally, we need to

note the degree to which—in time honoured fashion, it must be said—both
Eritrea and Ethiopia host a range of armed groups opposed to the other.
These are movements which are in the truest sense ‘stateless’, lacking even a
physical space of their own except that which is provided by host govern-
ments; it is on the whole a precarious and sorry existence, although not
necessarily entirely hopeless—for they all hope that one day their moment
will come, and that they too will march as an armed frontier on the sinister
centre. Ethiopia provides ‘hospitality’—in the loosest sense of that term—to
a range of Eritrean opposition movements of variable efficacy, strength, and
secrecy, including the ELF-Revolutionary Council, the Democratic
Movement for the Liberation of Eritrea, the National Alliance of Eritrean
Forces, the Eritrean Democratic Alliance, and the Eritrean Revolutionary
Democratic Front. The list goes on, somewhat tragically. For their part, the
Eritreans support the Ethiopian People’s Patriotic Front—reportedly a few
hundred Amhara, and perhaps some Oromo, soldiers operating around
Gondar—and the Tigrayan National Alliance for Democracy, which appears
to exploit provincial differences within Tigray itself. At the time of writing,
again, the opposition movement known as Ginbot 7 appeared to be attempt-
ing to draw together various armed groups in a common front against the
EPRDF, and reports were emerging that talks were taking place with the
Eritrean government, if only at the lower level. In August 2009, an Ethiopian
court found several key opposition leaders guilty of treason, and in December
they were sentenced to death, mostly in absentia.
While, in truth, we lack the space here to provide the doubtless deserved
level of detail on the various movements along the Ethiopia–Sudan border
and their dizzying recent histories of shifting allegiance, it is sufficient to note
here some key aspects of the story. At the ‘international’ level, many of the
ongoing conflicts remain extremely vulnerable to the interests of distant
actors—the Ethiopian, Eritrean, and Sudanese governments—willing to
intervene in the pursuit of larger agendas.At the local level, recurrent violence
frequently takes place around the issue of access to resources—increasingly,
land and water—while cattle-raiding is common. Local groups also compete
for access to the power represented by local administrations; but these are also,
and have long been, ill-governed regions as far as metropolitan centres are
concerned, far from the centres of power which purport to control them.This
is true, of course, until the state decides that certain regions are worth large-
scale economic—and thus political and military—investment, at which point
256 e p i log ue

the state presence becomes rather more visible, and invariably brutal, as in the
case of the Gibe dam project. At that point, hostility toward the state may
increase still further, when compared to local attitudes toward the arm’s-length
patronage—a rough-hewn version of colonial-era indirect rule—which pre-
ceded the new interventionism. Firearms, moreover, proliferate across the
borderlands, acquired from a myriad of sources, including the region’s gov-
ernments themselves, who often supply arms to groups in the interests of local
‘security’—the SPLM, and individuals in the adjacent Ethiopian and Sudanese
armies. It is hardly controversial to suggest that guns do much to heighten the
capacity for, and thus the desirability of, violent conflict. Thus, heavily armed
pastoralists fight for cattle and pasture; and in turn, it seems safe to say, cultures
of violence have been created, or at least enhanced, as resources become
scarcer, or as political exigencies shift. Of course, many of these conflicts date
back decades, and have their roots in the nineteenth century with the creation
of ‘Sudan’ on the one side, and the expansion of the habesha polity on the
other; it is impossible to say with any degree of certainty whether the frontier
violence we see today is any more intense than it was in the past, but what we
can say is that it represents a deeply destabilizing vortex, both for the indige-
nous peoples of the areas themselves as well as the adjacent states.
Even close watchers of the region know little about many of these indis-
tinct organizations, which seem to function in a curious political gloaming.
Yet they provide clues to the origins of the current political situation in
north-east Africa, and we may look on them in the way that astronomers
examine dust and dark matter for clues to the origins of the universe. They
echo the war cries, the rhetoric, and indeed the nomenclature of the move-
ments of the 1960s and 1970s, and as political organisms they remind us of
how it all began—in the ideological tribalism of people’s democratic revo-
lution, and in the violence necessary in the pursuit of regional, or ethnic, or
national liberation. They, too, are the products of fertile frontiers; but this
political harvest is less impressive, to date, in terms of the advance of the
cause, and it is for the historians of the future to assess the conversion of
potential to kinetic energy, and the march of these armed and vibrant fron-
tiers on decadent and dying centres.
 Endnotes

prologue
1. See for example K. Menkhaus, ‘Somalia: a country in peril, a policy nightmare’,
ENOUGH Strategy Paper (September, 2008) p.8.
2. S. Healy, Lost Opportunities in the Horn of Africa: How Conflicts Connect and Peace
Agreements Unravel (London, 2008).

chapter 1
1. The term habesha refers loosely to Amharic- and Tigrinya-speaking (i.e. Semitic)
peoples, and is commonly how these peoples describe themselves. In certain
contexts, notably those prior to the late nineteenth century, I consider it prefer-
able to terms such as ‘Ethiopia’ and certainly ‘Abyssinia’.
2. See for example C. Clapham, ‘Boundary and Territory in the Horn of Africa’,
in A. I. Asiwaju and P. Nugent (eds.), African Boundaries: Barriers, Conduits and
Opportunities (London, 1996) p.245.
3. J. Young, Armed Groups Along Sudan’s Eastern Frontier: An Overview and Analysis
(Geneva, 2007) pp.22, 26, 38.
4. Although see C. Ehret, Ethiopians and East Africans: The Probem of Contacts
(Nairobi, 1974).
5. Gadaa Melbaa, Oromia: An Introduction (Khartoum, 1988).
6. See also D. Crummey,‘Society, state and nationality in the recent historiography
of Ethiopia’, Journal of African History, 31:1 (1990).
7. D. L. Donham and W. James (eds.), The Southern Marches of Imperial Ethiopia
(Oxford, 2002; 1986).
8. J. Markakis, National and Class Conflict in the Horn of Africa (Cambridge, 1987).
9. B. K. Holcomb and Sisai Ibssa, The Invention of Ethiopia:The Making of a Dependent
Colonial State in Northeast Africa (Trenton, NJ, 1990).
10. J. Sorenson, Imagining Ethiopia: Struggles for History and Identity in the Horn of
Africa (New Brunswick, NJ, 1993).
11. W. James, D. Donham, E. Kurimoto, and A. Triulzi (eds.), Remapping Ethiopia:
Socialism and After (Oxford, 2002).
12. See for example D. Connell, Conversations with Eritrean Political Prisoners (Trenton,
NJ, 2005); Kidane Mengisteab and Okbazghi Yohannes, Anatomy of an African
258 e ndnote s

Tragedy: Political, Economic and Foreign Policy Crisis in Post-independence Eritrea

(Trenton, NJ, 2005); and my own ‘The politics of silence: interpreting apparent
stasis in contemporary Eritrea’, Review of African Political Economy, 36:120
(2009).
13. Ezekiel Gebissa (ed.), Contested Terrain: Essays on Oromo Studies, Ethiopianist
Discourse, and Politically Engaged Scholarship (Trenton, NJ, 2009), offers some
interesting insights into the rancour that has dogged the scholarly dimension of
the Oromo struggle.
14. D. Levine, Greater Ethiopia (Chicago, 2000; 1974) pp.36–9.
15. E. Gellner, Nations and Nationalism (Cambridge, 2005; 1983) p.85.
16. Bairu Tafla (ed.), Eritrean Studies Review, 5:1 (2007).
17. See for example P. R. Schmidt, M. C. Curtis and Zelalem Teka (eds.), The
Archaeology of Ancient Eritrea (Trenton, NJ, 2008).
18. I. M. Lewis, A Modern History of the Somali (Oxford, 2002).
19. D. Crummey, Land and Society in the Christian Kingdom of Ethiopia: From the
Thirteenth to the Twentieth Century (Oxford, 2000).
20. Donham and James, Southern Marches.
21. For example, Mohammed Hassen, The Oromo of Ethiopia: A History, 1570–1860
(Cambridge, 1990; Trenton, NJ, 1994); Asafa Jalata, Oromia and Ethiopia: State
Formation and Ethnonational Conflict, 1868–1992 (London and Boulder, CO,
1993); P. T. W. Baxter, J. Hultin, and A. Triulzi (eds.), Being and Becoming Oromo:
Historical and Anthropological Enquiries (Uppsala, 1996).
22. A partial (and largely unconvincing) exception is Semere Haile, ‘Historical
background to the Ethiopia–Eritrea conflict’, in B. Davidson and L. Cliffe
(eds.), The Long Struggle of Eritrea for Independence and Constructive Peace (Trenton,
NJ, 1988). Various other studies of the liberation struggle in the 1980s devoted
introductory chapters to antiquity.
23. K. Fukui and J. Markakis (eds.), Ethnicity and Conflict in the Horn of Africa
(London, 1994); E. Kurimoto and S. Simonse (eds.), Conflict, Age and Power in
North East Africa: Age Systems in Transition (Oxford, 1998).
24. P. Marsden, The Barefoot Emperor: An Ethiopian Tragedy (London, 2007).
25. For example J. C. McCann, People of the Plow: An Agricultural History of Ethiopia,
1800–1990 (London, 1995).
26. I. Orlowska, ‘Re-imagining empire: Ethiopian political culture under Yohannis
IV (1872–89)’, PhD thesis, School of Oriental and African Studies (London,
2006).
27. R. J. Reid, War in Pre-Colonial Eastern Africa (Oxford, 2007) pp.8ff.
28. For example, S. Rubenson, ‘Adwa 1896: the resounding protest’, in R. Rotberg
and A. Mazrui (eds.), Protest and Power in Black Africa (New York, 1970); for a
contemporary view, see G. F.-H. Berkeley, The Campaign of Adowa and the Rise
of Menelik (New York, 1969; first pub., 1902).
29. D. Gerhard, ‘The frontier in comparative view’, Comparative Studies in Society
and History, 1:3 (1959).
 e ndnote s 25 9

30. I. Kopytoff, ‘The internal African frontier: the making of African political cul-
ture’, in I. Kopytoff (ed.), The African Frontier: The Reproduction of Traditional
African Societies (Bloomington and Indianapolis, 1987).
31. Donham and James, Southern Marches; James et al., Remapping Ethiopia;
M. Leopold, Inside West Nile: Violence, History and Representation on an African
Frontier (Oxford, 2005); Nene Mburu, Bandits on the Border: The Last Frontier in
the Search for Somali Unity (Trenton, NJ, 2005);Young, Armed Groups; Clapham,
‘Boundary and territory’, passim.
32. See for example F. Barth (ed.), Ethnic Groups and Boundaries: The Social
Organisation of Culture Difference (London, 1969); and for a summation,
P. Lawrence, Nationalism: History and Theory (Harlow, 2005) p.187.

chapter 2
1. Schmidt et al., Ancient Eritrea.
2. Yosief Libsekal,‘Eritrea’, in International Council on Monuments and Sites (ICOMOS):
World Report 2001–02 on Monuments and Sites in Danger (Paris, 2002).
3. D. Phillipson, Ancient Ethiopia. Aksum: Its Antecedents and Successors (London,
1998) pp.111ff.
4. Taddesse Tamrat, Church and State in Ethiopia, 1270–1527 (Oxford, 1972) pp.54ff.
5. An affordable and accessible edition is M. Brooks (ed. and tr.), A Modern
Translation of the Kebra Negast (The Glory of the Kings) (Lawrenceville, NJ,
1995).
6. G. W. B. Huntingford (ed. and tr.), The Glorious Victories of Amda Seyon, King of
Ethiopia (London, 1965); Taddesse Tamrat, Church and State, chapters 3–5.
7. Levine, Greater Ethiopia.
8. See for example Huntingford, Glorious Victories;R. Pankhurst (ed.), The Ethiopian
Royal Chronicles (Addis Ababa, 1967).
9. Brooks, Modern Translation, p.127. See also Reid, War, and R. J. Reid, ‘War and
remembrance: orality, literacy and conflict in the Horn’, Journal of African
Cultural Studies, 18:1 (2006).
10. Lewis, Modern History, pp.20ff.
11. D. Robinson, Muslim Societies in African History (Cambridge, 2004) pp.114ff;
M. Abir, Ethiopia and the Red Sea: The Rise and Decline of the Solomonic Dynasty
and Muslim–European Rivalry in the Region (London, 1980) pp.10–13; and for a
useful contemporary account, see Arab Faqih (tr. P. L. Stenhouse), The Conquest
of Abyssinia (Hollywood, 2003).
12. See Crummey, Land and Society, pp.50–1 and passim.
13. Bahrey’s account is contained in C. F. Beckingham and G. W. B. Huntingford
(trs. and eds.), Some Records of Ethiopia 1593–1646: Being Extracts from ‘The History
of High Ethiopia or Abassia’, by Manoel de Almeida (London, 1954).
14. H. Salt, A Voyage to Abyssinia and Travels into the Interior of that Country (London,
1814) pp.299, 306.
260 e ndnote s

15. S. Gobat, Journal of Three Years’ Residence in Abyssinia (New York, 1969; first pub.,
1851) p.52.
16. H. Blanc, A Narrative of Captivity in Abyssinia (London, 1970; first pub., 1868)
p.290.
17. C. R. Markham, A History of the Abyssinian Expedition (London, 1869)
pp.39–40.
18. Although see for example Holcomb and Ibssa, The Invention of Ethiopia, and
Sorenson, Imagining Ethiopia. For more recent insights into current debates in
Oromo studies, see the collection of essays in Ezekiel Gebissa, Contested
Terrain.
19. Mohammed Hassan, The Oromo of Ethiopia, pp.18ff; M. Abir, ‘Ethiopia and the
Horn of Africa’, in R. Gray (ed.), Cambridge History of Africa: Vol. 4, from c.1600
to c.1790 (Cambridge, 1975) pp.537ff.
20. In particular, see the work of Asmarom Legesse: Gada: Three Approaches to the
Study of African Society (New York, 1973); and Oromo Democracy: An Indigenous
African Political System (Trenton, NJ, 2006).
21. Crummey, Land and Society, passim.
22. Levine, Greater Ethiopia, p.82.
23. H. S. Lewis, Jimma Abba Jifar, an Oromo monarchy: Ethiopia 1830–1932 (Lawrence-
ville, NJ, 2001); Mohammed Hassen, The Oromo of Ethiopia.
24. Crummey, Land and Society, pp.67ff.
25. E. A. Wallis Budge, A History of Ethiopia, Nubia and Abyssinia (London, 1928) II,
pp.445ff.
26. Pankhurst, Chronicles, pp.121ff.

chapter 3
1. The best detailed accounts are still S. Rubenson, ‘Ethiopia and the Horn’, in
J. E. Flint (ed.), Cambridge History of Africa: Vol. 5, c.1790–c.1870 (Cambridge,
1976); S. Rubenson, The Survival of Ethiopian Independence (London, 1976);
M.Abir, Ethiopia:The Era of the Princes (London, 1968). See also H.Weld Blundell,
The Royal Chronicle of Abyssinia, 1769–1840 (Cambridge, 1922).
2. Pankhurst, Chronicles, pp.140–2.
3. See for example UK National Archives Foreign Office Series (hereafter NA
FO) 1/1 Valentia and Salt: letters and documents. H. Salt, inclosure re. maps,
London, 22 August 1811.
4. J. Bruce, Travels to Discover the Source of the Nile, in the Years 1768, 1769, 1770, 1771,
1772, and 1773 (Edinburgh, 1790) Vol. 2, p.696.
5. Ibid., p.680.
6. M. Abir, ‘Ethiopia and the Horn of Africa’, in R. Gray (ed.), Cambridge History
of Africa: Vol. 4, from c.1600 to c.1790 (Cambridge, 1975) pp.571–7; Abir, Ethiopia:
The Era of the Princes, especially chapter 2.
7. Rubenson, ‘Ethiopia and the Horn’, pp.57ff.
 e ndnote s 2 61

8. See also Reid, War, pp.13–20 passim.

9. R. J. Reid, ‘Warfare and urbanisation: the relationship between town and con-
flict in pre-colonial eastern Africa’, in A. Burton (ed.), The Urban Experience in
Eastern Africa, c.1750–2000 (Nairobi, 2002).
10. Ibid., pp.288ff.
11. W. C. Plowden, Travels in Abyssinia and the Galla Country (London, 1868) p.74.
12. Blanc, Captivity, pp.294–5.
13. See for example Salt, Voyage, pp.486–95; Plowden, Travels, pp.75–7. For the early
use of firearms, see Merid Wolde Aregay, ‘A reappraisal of the impact of firearms
in the history of warfare in Ethiopia, c.1500–1800’, Journal of Ethiopian Studies,
14 (1980).
14. Gobat, Residence, p.441.
15. Plowden, Travels, esp. chapters III and IV.
16. Ibid., p.133.
17. M. Parkyns, Life in Abyssinia: Being Notes Collected During Three Years’ Residence
and Travels in that Country (London, 1966; first pub. 1853, 1868) p.xxi.
18. R. A. Brown, quoted in M. Howard, War in European History (Oxford, 2009)
p.1.
19. Plowden, Travels, p.38.
20. Quoted in Howard, War, p.23.
21. Plowden, Travels, p.45.
22. NA FO 1/1 Valentia and Salt: letters and documents. H. Salt, ‘Extracts from
original observations’, Chelicut, April 1810.
23. Among the best accounts are Mohammed Hassen, The Oromo of Ethiopia and
Lewis, Jimma Abba Jifar; see also M. Abir, ‘The emergence and consolidation of
the monarchies of Enarea and Jimma in the first half of the nineteenth century’,
Journal of African History, 6:2 (1965). For the unabashedly Oromo nationalist
interpretation—both of the nineteenth century and everything later, including
‘Abyssinian colonialism’ and the anti-colonial struggle—see for example Gadaa
Melbaa, Oromia.
24. Lewis, Jimma Abba Jifar, especially chapter 2.
25. Abir, ‘Emergence and consolidation’.
26. See Abir, Ethiopia:The Era of the Princes.
27. D. Crummey, ‘Society and ethnicity in the politics of Christian Ethiopia during
the Zamana Masfent’, International Journal of African Historical Studies, 8:2 (1975)
p.278.
28. A. Hoben, Land Tenure Among the Amhara: The Dynamics of Cognatic Descent
(Chicago, 1973); Levine, Greater Ethiopia; D. Crummey, ‘Abyssinian feudalism’,
Past and Present, 89 (1980); J. Goody, Technology, Tradition and the State in Africa
(London, 1971) pp.30–2.
29. A. Moore-Harell, ‘Economic and political aspects of the slave trade in Ethiopia
and the Sudan in second half of the nineteenth century’, International Journal of
African Historical Studies, 23:2–3 (1999).
262 e ndnote s

30. J. H. Arrowsmith-Brown (ed. and tr.), Prutky’s Travels in Ethiopia and Other
Countries (London, 1991) pp.25, 152, 179; see also Gobat, Residence, p.39.
31. Plowden, Travels, p.20.
32. Arrowsmith-Brown, Prutky’s Travels, pp.152, 179.
33. Parkyns, Life in Abyssinia, pp.155–6.
34. Ibid., p.441; H. Dufton, Narrative of a Journey Through Abyssinia in 1862–63
(Westport, 1970, 1st ed. 1867) pp.42–3.
35. Gobat, Residence, pp.401–2.
36. J. Bruce, Travels, III, p.88.
37. Salt, Voyage to Abyssinia, p.426.
38. H. Stern, Wanderings Among the Falashas in Abyssinia (London, 1968, 1st ed. 1862)
p.146.
39. A. B. Wylde, Modern Abyssinia (London, 1901) pp.71, 73.
40. Ibid., pp.125, 127.
41. For the basics, see S. Rubenson, King of Kings:Tewodros of Ethiopia (Addis Ababa
and Nairobi, 1966); D. Crummey, ‘Tewodros as reformer and moderniser’,
Journal of African History, 10:3 (1969); D. Crummey, ‘The violence of Tewodros’,
in B. A. Ogot (ed.), War and Society in Africa (London, 1972).
42. D. Crummey, ‘Banditry and resistance: noble and peasant in nineteenth-century
Ethiopia’, in D. Crummey, Banditry, Rebellion and Social Protest in Africa (London,
1986) p.133; in the same volume, see also T. Fernyhough, ‘Social mobility and
dissident elites in northern Ethiopia: the role of banditry, 1900–1969’. Much of
the analysis in the Crummey volume itself owes something to E. Hobsbawm,
Bandits (London, 2000).
43. Crummey, ‘Banditry and resistance’, p.135.
44. See the excellent account by R. Caulk, ‘Bad Men of the Borders: shum and
shifta in northern Ethiopia in the nineteenth century’, International Journal of
African Historical Studies, 17:2 (1984).
45. Stern, Wanderings, pp.64–6.
46. Rubenson, ‘Ethiopia and the Horn’, pp.65ff; Crummey, ‘The violence of
Tewodros’, pp.67ff.
47. Stern, Wanderings, pp.68–75; Dufton, Narrative, pp.122–31.
48. Dufton, Narrative, pp.113–14.
49. Reid, ‘War and remembrance’; Reid, War, especially chapter 2.
50. Stern, Wanderings, p.122.
51. Ibid., p.128.
52. Ibid., p.129.
53. Dufton, Narrative, p.105.
54. Markham, History, pp.84–5.
55. Blanc, Captivity, p.5.
56. NA FO 881/1609 Account of Mission to Abyssinia, by H. Rassam to Lord
Stanley, 1 September 1868, p.4.
57. Blanc, Captivity, pp.7, 8–9.
 e ndnote s 263

58. Ibid., pp.333–5.

59. Tewodros to Hormuzd Rassam, [28 January 1866], in S. Rubenson (ed.), Acta
Aethiopica II:Tewodros and his Contemporaries 1855–1868 (Addis Ababa, 1994).
60. NA FO 881/1609 Account of Mission to Abyssinia, by H. Rassam to Lord
Stanley, 1 September 1868, pp.2, 7.
61. Parkyns, Life in Abyssinia, p.xxiii.
62. For example, Reid, War, pp.147–8; Blanc, Captivity, p.315; Assegahhen to
d’Abbadie, 14 January 1866 and 28 November 1868, in Rubenson, Acta
Aethiopica II; NA FO 881/1493 Merewether to Stanley, 15 February 1867.
63. Pankhurst, Chronicles, p.151.
64. Crummey, ‘The violence of Tewodros’, p.71.
65. NA FO 881/1493 Merewether to Stanley, 15 February 1867.
66. Crummey, ‘The violence of Tewodros’, p.72.
67. The best accounts of the expedition itself remain D. Bates, The Abyssinian
Difficulty (Oxford, 1979), and more recently Marsden, The Barefoot Emperor.
There was a rush of publications through the 1860s, and especially around the
Abyssinian expedition, replete with breathless wonder at the crazed genius that
was Tewodros. A selection would include: Blanc, Captivity; Dufton, Narrative;
Markham, History; Plowden, Travels; H. Rassam, Narrative of the British Mission to
Theodore, King of Abyssinia (London, 1869); H. M. Stanley, Coomassie and Magdala:
the story of two British campaigns in Africa (London, 1874); Stern, Wanderings.
68. I. M. Lapidus, A History of Islamic Societies (Cambridge, 2002) pp.771–2.
69. Gellner, Nations, p.84.
70. K. S. Vikor, ‘Sufi brotherhoods in Africa’, in N. Levtzion and R. L. Pouwels
(eds.), The History of Islam in Africa (Oxford, 2000) p.468.
71. Iqbal Jhazbhay, ‘Islam and stability in Somaliland and the geo-politics of the war
on terror’, Journal of Muslim Minority Affairs, 28:2 (2008) p.179, drawing on the
work of Said Samatar among others. See also the collection of essays in
A. de Waal (ed.), Islamism and its Enemies in the Horn of Africa (London, 2004),
especially chapters 1 and 4 in this context.
72. A good overview can be found in L. Kaptjeins, ‘Ethiopia and the Horn of
Africa’, in Levtzion and Pouwels, History of Islam.
73. Abir, ‘Emergence and consolidation’, p.211.
74. See for example Hussein Ahmed, Islam in Nineteenth-Century Wallo, Ethiopia:
Revival, Reform and Reaction (Leiden, 2001).
75. Parkyns, Life in Abyssinia, p.276.
76. Blanc, Captivity, p.2.
77. Dufton, Narrative, pp.114–18.
78. Ibid., p.116.
79. NA FO 1/30 ‘Abyssinia ...’ Memo on Abyssinia by Sir E.Baring, to Granville,
February 1884.
80. For example, see NA FO 1/27B Yohannes to Victoria, 10 August 1872;Yohannes
to Alexander II, 19 June 1879, in S. Rubenson (ed.), Acta Aethiopica Vol. III:
2 64 e ndnote s

Internal Rivalries and Foreign Threats, 1869–1879 (Addis Ababa, 2000); also
A. Bulatovich (ed. and tr. R. Seltzer), Ethiopia Through Russian Eyes: Country in
Transition, 1896–98 (Lawrenceville, NJ, 2000) p.53.
81. Bulatovich, Ethiopia Through Russian Eyes, p.94.
82. See also the account in E. Paice, Tip and Run:The Untold Tragedy of the Great War
in Africa (London, 2007) pp.212ff.

chapter 4
1. See for example Gebru Tareke, Ethiopia: Power and Protest. Peasant Revolts in the
Twentieth Century (Lawrenceville, NJ, 1996) pp.109, 116, 118; J. Young, Peasant
Revolution in Ethiopia:The Tigray People’s Liberation Front, 1975–1991 (Cambridge,
1997) p.99; Alemseged Abbay, Identity Jilted or Re-Imagining Identity? The Divergent
Paths of the Eritrean and Tigrayan Nationalist Struggles (Lawrenceville, NJ, 1998)
passim; and, rather more disparagingly, the Eritrean nationalist viewpoint in
Jordan Gebre-Medhin, ‘Eritrea (Mereb-Melash) and Yohannes IV of Abyssinia’,
Eritrean Studies Review, 3:2 (1999).
2. NA FO 1/1 Valentia and Salt: letters and documents. ‘Observations on the trade
of the Red Sea’, by Valentia, p.9.
3. A good contemporary account of the early years of the nineteenth century is
provided in Salt, Voyage to Abyssinia, pp.269ff.
4. Gobat, Residence, pp.396–407.
5. The most exhaustive account remains Zewde Gabre-Selassie, Yohannes IV of
Ethiopia: A Political Biography (Oxford, 1975). See also Bairu Tafla (ed. and tr.),
A Chronicle of Emperor Yohannes IV (1872–89) (Wiesbaden, 1977).
6. By Donald Crummey, see ‘Orthodoxy and imperial reconstruction in Ethiopia,
1854–1878’, Journal of Theological Studies, 29:2 (1978); and ‘Imperial legitimacy
and the creation of neo-Solomonic ideology in 19th-century Ethiopia’, Cahiers
d’Etudes Africaines, 28 (1988).
7. Wylde, Modern Abyssinia, p.29.
8. Ibid., pp.33, 44; also NA FO 1/27B Abyssinia: Mission of General Kirkham.
Yohannes to Victoria, 10 August 1872.
9. This can be seen in the various correspondence from and about him contained
in Rubenson, Acta Aethiopica III.
10. R. J. Reid, ‘The challenge of the past: the quest for historical legitimacy in
independent Eritrea’, History in Africa, 28 (2001).
11. The best study of Massawa to date is J. Miran, Red Sea Citizens: Cosmopolitan
Society and Cultural Change in Massawa (Bloomington and Indianapolis,
2009).
12. H. Erlich, Ras Alula and the Scramble for Africa: A Political Biography. Ethiopia and
Eritrea, 1875–1897 (Lawrenceville, NJ, 1996) pp.ix, xiii.
13. Quoted in ibid., p.17.
14. G. H. Portal, My Mission to Abyssinia (London, 1892) p.81.
 e ndnote s 2 65

15. Salt, Voyage to Abyssinia, pp.307,488; NA FO 1/1 Salt to Cullen Smith, September
1811.
16. NA FO 1/1 Ras Welde Selassie to HM The King, 25 February 1811.
17. Gobat, Residence, p.39.
18. Ibid., pp.37–8.
19. Ibid., pp.389–90.
20. Plowden, Travels, pp.24–5, 27, 39.
21. Ibid., p.39.
22. Ibid., p.131.
23. Parkyns, Life in Abyssinia, pp.55, 98.
24. Plowden, Travels, p.22.
25. Salt, Voyage, p.213.
26. Ibid., p.200; Plowden, Travels, pp.25–6.
27. Salt, Voyage, pp.361–2.
28. NA FO 1/1 Salt to Cullen Smith, 4 March 1811.
29. NA FO 1/1 Salt, ‘Extracts’, April 1810.
30. Wube to Louis Philippe, 24 May 1845, in S.Rubenson (ed.), Acta Aethiopica
I: Correspondence and Treaties, 1800–1854 (Evanston, IL, and Addis Ababa,
1987).
31. Wube to Louis-Napoleon Bonaparte, 2 October [1849], in ibid.
32. Salt, Voyage, pp.305ff.
33. Crummey, ‘The violence of Tewodros’, pp.71–2, 75.
34. Birru Petros to Antoine d’Abbadie, 26 November 1858, in Rubenson, Acta
Aethiopica II.
35. See the Treaty between Simen-Tigray and France, 29 December 1859, in ibid.
36. Aregawi Subagadis to Theodore Gilbert, 30 December 1860, in ibid.
37. Blanc, Captivity, pp.8–9.
38. Ibid., p.50.
39. Ibid., pp.90–1.
40. Asseggahen to Antoine d’Abbadie, 14 January 1866 and 15 April 1867, in
Rubenson, Acta Aethiopica II.
41. Afe Werq and Welde Mesqel to Alemayyehu Tewodros, 21 August 1869, in
Rubenson, Acta Aethiopica III.
42. NA FO 1/27B Abyssinia: Mission of General Kirkham: Yohannes to Victoria,
4 June 1873.
43. NA FO 1/27B Yohannes to Granville, dated 1872.
44. NA FO 1/27B Kirkham’s Mission, statement for Granville, 31 October 1872.
45. NA FO 1/27B Kirkham to Granville, 13 May 1873.
46. Yohannes IV to Isma’il Ibrahim, [31 July 1872], in Rubenson, Acta Aethiopica III.
47. Yohannes IV to Victoria, 13 August 1872, in ibid.
48. Yohannes IV to E.Choquin de Sarzec, 24 March 1873, in ibid.
49. Yohannes IV/J.C.Kirkham to Granville, 13 May 1873, in ibid.
50. Yohannes IV to Granville, 15 May 1873, in ibid.
266 e ndnote s

51. For example,Yohannes IV to Victoria, 4 June 1873, in ibid.

52. NA FO 881/3058 Governor of Massawa to Khairy Pasha, 2 September 1875.
53. NA FO 881/3058 Ismail to Arendrup, 17 September 1875.
54. NA FO 881/3058 Stanton to Derby, 14 November 1875.
55. NA FO 881/3058 Stanton to Derby, 27 November 1875.
56. See for example Bahru Zewde, A History of Modern Ethiopia 1855–1991 (London,
2001) pp.51ff.
57. NA FO 881/1610 Cameron to Stanley, 28 September 1868.
58. Hailu Tewelde Medhin to Ismail Ibrahim, [14 November 1867], in Rubenson,
Acta Aethiopica III.
59. NA FO 881/1522 Munzinger’s ‘Notes and Route Observations’ (1867).
60. Welde Mikael Solomon to Napoleon III, 22 August 1869, in Rubenson, Acta
Aethiopica III.
61. Kasa Mircha to Munzinger, 17 November [1869], in ibid.
62. NA FO 1/29 Report by Lt Carter [1868?], appendix by Holdich.
63. Samuel Giyorgis et al. to E.Choquin de Sarzec, 13 March 1873, in Rubenson,
Acta Aethiopica III.
64. See also H. Marcus, The Life and Times of Menelik II: Ethiopia 1844–1913 (Oxford,
1975) pp.40, 42.
65. NA FO 881/3058 Stanton to Derby, 31 January 1876.
66. NA FO 881/3058 Extract from the Moniteur Egyptien, 3 December 1875.
67. NA FO 881/3058 Ismail to Governor of Massawa, 17 September 1875.
68. Yohannes IV to Hasan Ismail and Muhammad Ratib, 12 March 1876, in
Rubenson, Acta Aethiopica III.
69. Yohannes IV to Muhammad Ratib, [22 March 1876], in ibid.
70. See for example Welde Mikael Solomon to Muhammad Ratib et al., [March
1876], and Welde Mikael Solomon to Muhammad Ratib, [July 1876], in ibid.
See also Johannes Kolmodin, Traditions de Tsazzega et Hazzega (Uppsala, 1915).
71. For example,Welde Mikael Solomon to Muhammad Ratib, [September? 1876],
in Rubenson, Acta Aethiopica III.
72. The population of Se’azega to Charles Gordon, [March 1877], and Barya’u
Gebre Sadiq to Charles Gordon, 11 May 1878, in ibid.
73. Kasa Mircha to Munzinger, 17 November [1869], in ibid.
74. Asseggahen to Antoine d’Abbadie, [July–August?] 1866, in Rubenson, Acta
Aethiopica II.
75. NA FO 1/30 ‘Abyssinia ...’ Baker to Granville, 5 January 1884; also Marcus, Life
and Times, p.79.
76. For example, Salt, Voyage, pp.227–9.
77. Zekkariyas Tesfa Mikael to Guillaume Lejean, 5 July 1864, in Rubenson, Acta
Aethiopica II.
78. Portal, Mission, p.28.
79. Ibid., pp.7, 34.
80. NA FO 1/30 ‘Abyssinia ...’ Baker to Granville, 5 and 7 January 1884.
81. Ibid., 7 January 1884; see also Wylde, Modern Abyssinia, pp.30–1.
 e ndnote s 2 67

82. Portal, Mission, pp.5–6.

83. P. M. Holt, ‘Egypt and the Nile valley’, in J. E. Flint (ed.), Cambridge History of
Africa, Vol. 5: c.1790–c.1870 (Cambridge, 1976) pp.22ff; see also R. Gray,
A History of the Southern Sudan 1839–1889 (London, 1961).
84. A useful overview is in P. M. Holt and M. W. Daly, A History of the Sudan
(Harlow, 2000), chapter 4.
85. Wendy James’ work remains crucial. A selection would include: ’KWANIM
PA: The Making of the Uduk People. An Ethnographic Study of Survival in the
Sudan–Ethiopian Borderlands (Oxford, 1979); ‘War and “ethnic visibility”: the
Uduk on the Sudan–Ethiopian border’, in K. Fukui and J. Markakis (eds.),
Ethnicity and Conflict in the Horn of Africa (London, 1994); ‘Local centres on the
Western Frontier, 1974–97: a case study of Kurmuk’, in K. Fukui et al. (eds.),
Ethiopia in Broader Perspective: Papers of the XIIIth International Conference of
Ethiopian Studies Vol. II (Kyoto, 1997); ‘No place to hide: flag-waving on the
western frontier’, in W. James et al. (eds.), Remapping Ethiopia: Socialism and After
(Oxford, 2002).
86. A. Triulzi, Salt, Gold and Legitimacy: Prelude to the History of a No-man’s-land,
Bela Shangul,Wallagga, Ethiopia (ca.1800–1898) (Naples, 1981).
87. Sahle Dingil et al. to Louis Philippe, [June 1838], in Rubenson, Acta Aethiopica I.
88. Parkyns, Life in Abyssinia, p.177.
89. Plowden, Travels, p.20.
90. Parkyns, Life in Abyssinia, pp.182–5.
91. Ibid., p.187.
92. Plowden, Travels, p.19.
93. Ibid., pp.8–9.
94. NA FO 1/27B Stanton to Granville, 12 August 1872.
95. NA FO 1/27B Memo by Cherif Pasha, n.d., encl. in Stanton to Granville, 28
November 1872.
96. NA FO 1/31 Mason Bey to Nubar to Pasha, 7 May 1884.
97. NA FO 1/31 ‘Abyssinia ...’ Hewett to Secretary to the Admiralty, 22 June 1884,
and ‘Despatch from Hewett to Cairo’, 16 June 1884.
98. NA FO 1/30 ‘Abyssinia ...’ Memo on Abyssinia by Sir E.Baring, to Granville,
February 1884.
99. Portal, Mission, p.7.
100. Wylde, Modern Abyssinia, p.40.
101. See for example Berkeley, Campaign, p.379; Bahru Zewde, Modern Ethiopia,
p.58; Holt and Daly, History of the Sudan, pp.94–5.
102. Marcus, Life and Times, chapters 2 and 3.
103. Ibid., chapter 4.
104. Ibid., chapter 5; Rubenson, Survival, pp.384ff; G. N. Sanderson, ‘The Nile
basin and the eastern Horn, 1870–1908’, in R. O. Oliver and G. N. Sanderson
(eds.), Cambridge History of Africa, Vol. 6: from 1870 to 1905 (Cambridge, 1985)
pp.656ff.
105. Portal, Mission, p.248.
268 e ndnote s

106. Gobat, Residence, p.446.

107. Blanc, Captivity, pp.296–7.
108. Bulatovich, Russian Eyes, p.68.
109. Ibid., p.68.
110. Ibid., p.47.
111. Ibid., p.51.
112. Ibid., p.177.
113. Again, a great deal of rich contemporary detail on these campaigns is provided
by Bulatovich.
114. For a useful overview, see Bahru Zewde, Modern Ethiopia, pp.85ff; and on the
statuses of neftennya and gabbar in various local contexts, see Donham and
James, Southern Marches, passim.
115. Wylde, Modern Abyssinia, pp.56, 214–15.
116. Ibid., p.365.
117. Ibid., pp.174, 220–1.
118. Berkeley, Campaign, p.27.
119. R. Caulk, ‘Armies as predators: soldiers and peasants in Ethiopia, c.1850–1935’,
International Journal of African Historical Studies, 11:3 (1978); J. Dunn, ‘ “For God,
Emperor, and Country!” The evolution of Ethiopia’s nineteenth-century
army’, War in History, 1 (1994).
120. Reid, War, pp.79–106.
121. Berkeley, for example, suggests that an actual army might be rendered half as
large again by women and slaves: Berkeley, Campaign, p.9.
122. See also R. Caulk, ‘Firearms and princely power in Ethiopia in the nineteenth
century’, Journal of African History, 13:4 (1972) pp.609–10.
123. Wylde, Modern Abyssinia, p.5; see also NA FO 881/5530X ‘The Abyssinian
Army’, Lt Gleichen, 28 December 1887, pp.2–3.
124. Portal, Mission, pp.165–6.
125. Wylde, Modern Abyssinia, pp.7–9, 165.
126. NA FO 881/5530X ‘The Abyssinian Army’, Lt. Gleichen, 28 December 1887, p.1.
127. Bulatovich, Russian Eyes, p.100.

chapter 5
1. Rubenson, Survival, pp.384–99.
2. C. Giglio (tr. R. Caulk), ‘Article 17 of the Treaty of Uccialli’, Journal of African
History, 6:2 (1965).
3. R. Pateman, Eritrea: Even the Stones are Burning (Lawrenceville, NJ, 1990, 1998)
p.50.
4. See R. Pankhurst and D. Johnson, ‘The great drought and famine of 1888–92
in northeast Africa’, in D. Johnson and D. M. Anderson (eds.), The Ecology of
Survival: Case Studies from Northeast African History (London, 1988) pp.47–57;
McCann, People of the Plow, pp.89, 91–2.
 e ndnote s 269

5. NA FO 881/5530X ‘The Abyssinian Army’, by Lt. Gleichen.

6. Berkeley, Campaign, p.8.
7. Caulk, ‘Firearms and princely power’.
8. One of the best accounts remains Rubenson, ‘Adwa 1896’; and see also
Rubenson, Ethiopian Independence, pp.399–406.
9. See also R. J. Reid, ‘The Trans-Mereb Experience: perceptions of the historical
relationship between Eritrea and Ethiopia’, Journal of Eastern African Studies, 1:2
(2007) pp.240–6.
10. The idea recurs in virtually all of the key texts on modern Eritrean history. See
for example: Jordan Gebre-Medhin, Peasants and Nationalism in Eritrea (Trenton,
NJ, 1989) pp.56–69; Pateman, Eritrea, esp. pp.47–66; Ruth Iyob, The Eritrean
Struggle for Independence:Domination,Resistance,Nationalism 1941–1993 (Cambridge,
1995) p.4; Redie Bereketeab, Eritrea: the making of a nation, 1890–1991 (Trenton,
NJ, 2007) pp.75–132.
11. A good overview can be found in Redie Bereketeab, Eritrea, chapter 4.
12. Foreign Office (UK), Handbook [on Eritrea] prepared under the direction of the his-
torical section of the Foreign Office, No.126 (London, 1920).
13. Jordan Gebre-Medhin, Peasants and Nationalism, pp.56ff.
14. Foreign Office, Handbook.
15. Ibid.
16. See for example F. Locatelli, ‘Oziosi, vagaboni e pregiudicati: labour, law and
crime in colonial Asmara, 1890—1941’, International Journal of African Historical
Studies, 40:2 (2007); and her ‘Beyond the Campo Cintato: prostitutes, migrants
and “criminals” in colonial Asmara (Eritrea), 1890–1941’, in F. Locatelli and
P. Nugent (eds.), African Cities: Competing Claims on Urban Spaces (Leiden, 2009).
17. Uoldelul Chelati Dirar,‘Colonialism and the construction of national identities:
the case of Eritrea’, Journal of Eastern African Studies, 1:2 (2007).
18. For example, R. Caulk, ‘Ethiopia and the Horn’, in A. D. Roberts (ed.),
Cambridge History of Africa:Vol.VII (Cambridge, 1986) pp.724–5.
19. Ibid., p.725.
20. Ibid.
21. Foreign Office, Handbook.
22. Caulk, ‘Ethiopia and the Horn’, p.725.
23. Ibid., p.726; see also Pateman, Eritrea, pp.58ff.
24. See for example W. C. Young, ‘The Rashayida Arabs vs the State: the impact
of European colonialism on a small-scale society in Sudan and Eritrea’, Journal
of Colonialism and Colonial History, 9:2, Fall 2008.
25. Ibid.
26. Ibid.
27. In Tekeste Negash, No Medicine for the Bite of a White Snake (Uppsala, 1986),
pp.41–2; Kolmodin, Traditions, pp.281–3.
28. Tekeste Negash, No Medicine, p.44; Pateman, Eritrea, pp.51–2; Kolmodin,
Traditions, pp.284–5.
270 e ndnote s

29. See R. Caulk, ‘ ‘Black snake, white snake’: Bahta Hagos and his revolt against
Italian overrule in Eritrea, 1894’, in D. Crummey (ed.), Banditry, Rebellion and
Social Protest in Africa (London, 1986).
30. Pateman, Eritrea, p.52.
31. Tekeste Negash, No Medicine, p.45.
32. Redie Bereketeab, Eritrea, pp.120ff.
33. Foreign Office, Handbook.
34. Tekeste Negash, No Medicine, pp.45ff; also Tekeste Negash, Italian Colonialism in
Eritrea, 1882–1941: policies, praxis and impact (Uppsala, 1987) pp.127ff.
35. Tekeste Negash, Italian Colonialism, p.127.
36. Alemseged Tesfai, Aynefalale (Asmara, 2001) pp.267–8.
37. The best military account is A. J. Barker, Eritrea 1941 (London, 1966).
38. Ibid., p.109.
39. Quoted in S. Pankhurst, British Policy in Eritrea and Northern Ethiopia (Woodford
Green, 1946). ‘Benadir’ or ‘Benaadir’ was the term applied to southern Italian
Somaliland.
40. It is the opening gambit in M. Wrong, I Didn’t Do It For You: How the World
Betrayed a Small African Nation (London and New York, 2005).
41. Lewis, Modern History, chapter 1.
42. L.V. Cassanelli, The Shaping of Somali Society: Reconstructing the History of a Pastoral
People, 1600–1900 (Philadelphia, 1982) pp.201ff.
43. Ibid., pp.203–4.
44. R. Robinson and J. Gallagher, Africa and the Victorians: The Official Mind of
Imperialism (London, 1981) p.331.
45. Mburu, Bandits, pp.23ff; see also Markakis, National and Class Conflict, pp.43–4.
46. It is beyond the scope of this book to explore the clans in detail; suffice to say
here that the history of relations between the main groupings—Dir, Isaq,
Ogaden, Darod, Rahanweyn, Hawiye, and Digil—has been characterized by
both necessarily peaceful cooperation and sporadically violent competition
over material and political resources.
47. See Lewis, Modern History, pp.63–91; also R. L. Hess, ‘The poor man of God:
Muhammad Abdullah Hassan’, in N. R. Bennett (ed.), Leadership in Eastern
Africa (Boston, 1968).
48. To the British, naturally, he was the ‘Mad Mullah’. For a detailed account of the
military campaigns, see H. Moyse-Bartlett, The King’s African Rifles: A Study in
the Military History of East and Central Africa, 1890–1945 (Aldershot, 1956) II,
pp.160–94; II, 419–33.
49. Paice, Tip and Run, pp.212ff.
50. Said S. Samatar, Oral Poetry and Somali Nationalism:The Case of Sayyid Mahammad
‘Abdille Hasan (Cambridge, 1982).
51. L. Farago, Abyssinia on the Eve (London, 1935) pp.258ff.
52. Record of the British Military Administration in Eritrea and Somalia, The First
to be Freed (London, 1944) pp.49–53.
 e ndnote s 2 71

53. Quoted in ibid., p.50.

54. C. Barnes, ‘The Somali Youth League, Ethiopian Somalis and the greater
Somalia idea, c.1946–48’, Journal of Eastern African Studies, 1:2 (2007).
55. Lewis, Modern History, pp.161ff.
56. Two accessible collections of documents from the British perspective, both
edited by Peter Woodward, are: British Documents on Foreign Affairs: Reports and
Papers from the Foreign Office Confidential Print. Part III, from 1940 through 1945.
Series G, Africa (Bethesda, MD, 1998); and British Documents on Foreign Affairs:
Reports and Papers from the Foreign Office Confidential Print. Part IV, from 1945
through 1950. Series G, Africa (Bethesda, MD, 1999—).
57. Pateman, Eritrea, pp.67ff; Jordan Gebre-Medhin, Peasants and Nationalism, chap-
ter 5; Redie Bereketeab, Eritrea, chapter 5, passim.
58. See also the range of opinion expressed in contemporary British analysis:
S. Pankhurst, Eritrea on the Eve: The Past and Future of Italy’s ‘First-born’ Colony,
Ethiopia’s Ancient Sea Province (Woodford Green, 1952) p.59; S. Longrigg, A Short
History of Eritrea (Oxford, 1945) pp.169–70; S. F. Nadel, Races and Tribes of Eritrea
(Asmara, 1944) pp.71, 78; S. Longrigg, ‘The future of Eritrea’, African Affairs
45(180) (1946) pp.122, 126; M. Perham, The Government of Ethiopia (London,
1948) pp.434–5; G. K. N. Trevaskis, Eritrea: A Colony in Transition, 1941–52
(London, 1960) p.130.
59. L. Ellingson, ‘The emergence of political parties in Eritrea, 1941–1950’, Journal
of African History, 18:2 (1977); Ruth Iyob, Eritrean Struggle, chapter 4.
60. See for example Alazar Tesfa Michael, Eritrea To-Day: Fascist Oppression Under the
Nose of British Military [sic] (Woodford Green, c.1946) p.5.
61. See Pankhurst, British Policy; Pankhurst, Eritrea on the Eve; and with her son
R. Pankhurst, Ethiopia and Eritrea:The Last Phase of the Reunion Struggle, 1941–1952
(Woodford Green, 1953). For a breezily refreshing commentary on Sylvia
Pankhurst’s obsession with Ethiopia, see W. B. Carnochan, Golden Legends:
Images of Abyssinia, Samuel Johnson to Bob Marley (Stanford, 2008).
62. See by Alemseged Abbay, ‘The trans-Mereb past in the present’, Journal of
Modern African Studies, 35:2 (1997); and Identity Jilted, passim.
63. Markakis, National and Class Conflict, p.64; see also Longrigg, ‘The future of
Eritrea’, p.126.
64. This was the short-lived Bevin-Sforza plan, which rested on the idea that the
highlands of Eritrea might belong to Ethiopia, but the western lowlands should
really be annexed to Sudan. The southern Afar lowlands were also to go to
Ethiopia.
65. Longrigg, ‘The future of Eritrea’, p.126.
66. Okbazghi Yohannes, Eritrea: A Pawn in World Politics (Gainesville, FL, 1991)
chapters 4 and 5.
67. Ellingson, ‘Emergence’.
68. Jordan Gebre-Medhin, Peasants and Nationalism, p.80.
69. Ruth Iyob, Eritrean Struggle, pp.65–6.
2 72 e ndnote s

70. Fernyhough, ‘Social mobility’, p.165 and passim.

71. Jordan Gebre-Medhin, Peasants and Nationalism, pp.108ff.
72. Ruth Iyob, Eritrean Struggle, chapter 4, passim.
73. D. Cumming, ‘The UN disposal of Eritrea’, African Affairs, 52:207 (1953) p.131.
74. Much information comes from Trevaskis, both in Eritrea, especially pp.103ff; and
his report on the issue, ‘A Study of the Development of the Present Shifta
Problem and the Means Whereby it can be Remedied’ (June 1950), Research
and Documentation Centre (Asmara), Box/293 File SH/20 Vol. II Acc 13406.
75. Plowden, Travels, chapter 2.
76. Trevaskis provides considerable detail on this in ‘Shifta Problem’, appendix B,
pp.12ff.
77. Ibid.
78. See also Jordan Gebre-Medhin, Peasants and Nationalism, chapter 6, passim.
79. See the ‘sitreps’ (situation reports) contained in the UK National Archives War
Office (WO) 230 Series, containing details of anti-shifta activities. Also
N. Mburu, ‘Patriots or bandits? Britain’s strategy for policing Eritrea, 1941–
1952’, Nordic Journal of African Studies, 9:2 (2000).
80. See for example RDC Box / 287 File 70 / B / 9 Acc.13339. The Amnesty—a
progress report. Enclosed letter to Senior Divisional Officers, 30 July 1951.
81. Trevaskis, Eritrea, pp.70–1.
82. J.-B.Gewald, ‘Making tribes: social engineering in the Western province of
British-administered Eritrea, 1941–1952’, Journal of Colonialism and Colonial
History, 1:2 (2000).
83. RDC Box / 287 File 70 / B / 9 Acc.13339, Address by His Excellency the Chief
Administrator, 16 June 1951.
84. Okbazghi Yohannes, Eritrea, provides an in-depth examination. See also Ruth
Iyob, Eritrean Struggle, pp.73ff; and Bereket Habte Selassie, ‘From British rule to
federation and annexation’, in B. Davidson, L. Cliffe, and Bereket Habte Selassie
(eds.), Behind the War in Eritrea (Nottingham, 1980).
85. The full text is in Tekeste Negash, Eritrea and Ethiopia: The Federal Experience
(Uppsala, 1997) pp.188–208.

chapter 6
1. For an excellent overview, see Caulk, ‘Ethiopia and the Horn’, pp.707ff.
2. Bahru Zewde, Modern Ethiopia, p.121.
3. There are a number of detailed accounts of these events. See for example the excel-
lent essay by R. Caulk, ‘Ethiopia and the Horn’, in A. D. Roberts (ed.), Cambridge
History of Africa,Vol. 7: from 1905 to 1940 (Cambridge, 1986); Marcus, History, pp.116ff;
P.Henze, Layers of Time: A History of Ethiopia (London, 2000), pp.188ff; Bahru
Zewde, Modern Ethiopia, pp.111ff. See also Haile Selassie’s own account of his early
life in E. Ullendorff (ed. and tr.), The Autobiography of Emperor Haile Selassie I:
‘My Life and Ethiopia’s Progress’, 1892–1937 (London, 1976).
 e ndnote s 273

4. Marcus, History, p.120.

5. Abdusammad H. Ahmad, ‘Trading in slaves in Bela-Shangul and Gumuz,
Ethiopia: border enclaves in history, 1897–1938’, Journal of African History,
40:3 (1999).
6. See also R. Pankhurst, Economic History of Ethiopia, 1800–1935 (Addis Ababa,
1968) pp.108ff.
7. See Bahru Zewde, Pioneers of Change in Ethiopia:The Reformist Intellectuals of the
Early Twentieth Century (Oxford, 2002); and, for a useful summary, Bahru Zewde,
Modern Ethiopia, pp.103–11.
8. Ullendorff, Autobiography, pp.156ff; Henze, Layers of Time, pp.202ff.
9. W. Thesiger, The Danakil Diary: Journeys Through Abyssinia, 1930–34 (London,
1998) p.3.
10. Lewis, Jimma Abba Jifar, p.46.
11. Bahru Zewde, Modern Ethiopia, p.91.
12. Thesiger, Danakil Diary, pp.21, 99, and passim.
13. A. H. M. Jones and E. Monroe, A History of Ethiopia (Oxford, 1935) p.173.
14. Bahru Zewde, Modern Ethiopia, p.137.
15. Evelyn Waugh, Waugh in Abyssinia (Baton Rouge, 2007; first pub. 1936) p.24.
16. Ibid., pp.222–3, 239.
17. See for example Farago, Abyssinia.
18. Jones and Monroe, Ethiopia, pp.173–4; also Tibebe Ashete, ‘Towards a history
of the incorporation of the Ogaden: 1887–1935’, Journal of Ethiopian Studies,
27:2 (1994).
19. U. Almagor, ‘Institutionalising a fringe periphery: Dassanetch–Amhara rela-
tions’, in Donham and James, Southern Marches.
20. Sir P. Mitchell, African Afterthoughts (London, 1954) p.271.
21. Moyse-Bartlett, King’s African Rifles, II, p.209.
22. D. A. Low, ‘Uganda: the establishment of the Protectorate, 1894–1919’, in
V. Harlow and E. M. Chilver (eds.), History of East Africa (Oxford, 1965) II,
pp.106–7.
23. D. A. Low, ‘British East Africa: the establishment of British rule, 1895–1912’, in
Harlow and Chilver, History of East Africa, II, pp.30–1.
24. Moyse-Bartlett, King’s African Rifles, II, pp.447–9.
25. Ibid., p.467.
26. For a study of these shifts in the recent past, see for example G. Schlee, ‘Brothers
of the Boran once again: on the fading popularity of certain Somali identities
in northern Kenya’, Journal of Eastern African Studies, 1:3 (2007).
27. Sir H. Macmichael, The Anglo-Egyptian Sudan (London, 1934) p.97.
28. D. Johnson, ‘On the Nilotic frontier: imperial Ethiopia in the southern Sudan,
1898–1936’, in D. Donham and W. James (eds.), The Southern Marches of Imperial
Ethiopia (Oxford, 2002, 1986) p.240.
29. Ibid.
30. Macmichael, Anglo-Egyptian Sudan, pp.177–9, 182–6.
2 74 e ndnote s

31. Ibid., p.178; P.Garretson, ‘Vicious cycles: ivory, slaves and arms on the new Maji
frontier’, in Donham and James, Southern Marches.
32. Macmichael, Anglo-Egyptian Sudan, pp.182–3.
33. For example T. M. Coffey, Lion by the Tail: The Story of the Italian–Ethiopian War
(London, 1974); A. Mockler, Haile Selassie’s War:The Italian–Ethiopian Campaign,
1935–1941 (London, 1984).
34. Coffey, Lion by the Tail, pp.3ff.
35. See for example A. Hilton, The Ethiopian Patriots: Forgotten Voices of the Italo-
Abyssinian War 1935–41 (Stroud, 2007).
36. C. Zoli, ‘The organisation of Italy’s East African Empire’, Foreign Affairs, 16:1
(1937) pp.81–2.
37. E. M. Robertson, Mussolini as Empire-Builder: Europe and Africa 1932–36 (London,
1977) p.9.
38. H. Erlich, ‘Tigrean politics, 1930–35, and the approaching Italo-Ethiopian war’,
in H. Erlich, Ethiopia and the Challenge of Independence (Boulder, CO, 1986)
p.141.
39. For example Bahru Zewde, Modern Ethiopia, p.147.
40. See for example Howe to Eden, 5 July 1943, in P. Woodward (ed.), British
Documents on Foreign Affairs: Reports and Papers from the Foreign Office Confidential
Print. Part III, from 1940 through 1945. Series G, Africa (Bethesda, MD, 1998) (here-
after BDFA, III) Vol. III (Africa, April–December 1943).
41. Zoli, ‘Italy’s East African Empire’, pp.83–4.
42. Erlich, Challenge of Independence, especially pp.129–34, 135–65.
43. A. Triulzi, ‘Italian colonialism and Ethiopia’, Journal of African History, 23:2 (1982).
44. A. Sbacchi, Ethiopia under Mussolini: Fascism and the Colonial Experience (London,
1985); Haile Larebo, The Building of an Empire: Italian Land Policy and Practice in
Ethiopia (Trenton, NJ, 2006; first pub. 1994).
45. Zoli, ‘Italy’s East African Empire’, pp.80–1.
46. Ibid., p.84.
47. Ibid., pp.81–2.
48. Ibid., p.86.
49. Ibid., p.87.
50. Waugh, Waugh in Abyssinia, pp.234–5.
51. Henze, Layers of Time, p.226.
52. Waugh, Waugh in Abyssinia, pp.237–8.
53. Bahru Zewde, Pioneers, pp.42, 46, 86–8.
54. ‘Ethiopian personalities’, enclosure in Howe to Eden, 11 December 1942, in
BDFA, III,Vol. II, (Africa, January 1942–March 1943).
55. Mitchell, Afterthoughts, pp.202, 204.
56. H. Marcus, The Politics of Empire: Ethiopia, Great Britain and the United States,
1941–1974 (Lawrenceville, NJ, 1995; first pub. 1983) pp.8ff.
57. Foreign Office, ‘Policy for Ethiopia’, 20 January 1942, in BDFA, III, Vol. II
(Africa, January 1942–March 1943.
 e ndnote s 2 75

58. Young, Peasant Revolution, pp.51ff; Gebru Tareke, ‘Peasant resistance in Ethiopia:
the case of Weyane’, Journal of African History, 25:1 (1984); Gebru Tareke, Ethiopia:
Power and Protest, pp.89–124; H. Erlich, ‘ ‘Tigrean nationalism’, British involve-
ment, and Haile Selassie’s emerging absolutism: northern Ethiopia, 1941–1943’,
Asian and African Studies, 15:2 (1981).
59. Young, Peasant Revolution, pp.53–4; J. Hammond, Fire From the Ashes: A Chronicle
of Revolution in Tigray, Ethiopia, 1975–1991 (Lawrenceville, NJ, 1999) pp.165,
244–5.
60. Gebru Tareke, ‘Peasant resistance’, p.77.
61. Ibid., pp.79, 81–7.
62. ‘Ethiopia: Political Review for 1943’, enclosure in Howe to Eden, 13 June 1944,
in BDFA, III,Vol. IV (Africa, 1944).
63. E. A. Chapman-Andrews, ‘Political memorandum on Eritrea’, 27 October 1940,
in BDFA, III,Vol. I (Africa, 1940–1).
64. Marcus, Politics of Empire.
65. These can be followed in minute detail in the five volumes which make up
P.Woodward (ed.), British Documents on Foreign Affairs: Reports and Papers from the
Foreign Office Confidential Print. Part IV, from 1945 through 1950. Series G, Africa
(Bethesda, MD, 1999) (hereafter BDFA, IV). See also Tibebe Eshete, ‘The root
causes of political problems in the Ogaden, 1942–1960’, Northeast African Studies,
13:1 (1991).
66. R. Greenfield, Ethiopia: A New Political History (London, 1965) pp.306ff.
67. Bahru Zewde, Pioneers, offers the best account of this.
68. For example, ‘Ethiopia: Annual Review for 1948’, enclosure in Lascelles to
Bevin, 9 February 1949, in BDFA, IV,Vol. IV (Africa, January 1948–December
1949).
69. ‘Ethiopia: Annual review for 1949’, enclosure in Lascelles to Bevin, 8 May 1950,
in BDFA, IV,Vol.V (Africa, January–December 1950).
70. See J. Iliffe, The African Poor: A History (Cambridge, 1987) p.157.
71. See for example H. Erlich, ‘The Ethiopian Army and the 1974 Revolution’, in
M. Janowitz (ed.), Armed Forces and Society (Chicago, 1983).
72. Messay Kebede, Radicalism and Cultural Dislocation in Ethiopia, 1960–1974
(Rochester, NY, 2008).
73. Arguably the finest account of these events is still the late Richard Greenfield’s
Ethiopia.
74. For example, Addis Hiwet, Ethiopia: From Autocracy to Revolution (London,
1975).
75. H. Erlich, ‘The Eritrean autonomy, 1952–1962: its failure and its contribution to
further escalation’, in Y. Dinstein (ed.), Models of Autonomy (New York, 1981).
76. Jordan Gebre-Medhin, Peasants and Nationalism, pp.144ff.
77. J. Markakis, ‘The nationalist revolution in Eritrea’, Journal of Modern African
Studies, 26:1 (1988) p.54.
78. Okbazghi Yohannes, Eritrea, pp.189ff.
276 e ndnote s

79. T. Killion, ‘Eritrean workers’ organisation and early nationalist mobilisation:

1948–1958’, Eritrean Studies Review, 2:1 (1997).
80. Markakis, ‘Nationalist revolution’, p.54.
81. See also Ruth Iyob, Eritrean Struggle, pp.98ff.
82. Othman Saleh Sabbe (tr. Muhamad Fawaz al-Azem), The History of Eritrea
(Beirut, n.d. [c.1974?]) p.249.
83. Erlich, ‘Eritrean autonomy’, pp.178–9.
84. D. Johnson, The Root Causes of Sudan’s Civil Wars (Oxford, 2003) pp.29ff.
85. Markakis,‘Nationalist revolution’, pp.55; D.Pool, From Guerrillas to Government:
the Eritrean People’s Liberation Front (Oxford, 2001) pp.36–7.
86. Pool, Guerrillas, pp.49ff.
87. Gaim Kibreab, ‘Eritrean–Sudanese relations in historical perspective’, in
R. J. Reid (ed.), Eritrea’s External Relations: Understanding its Regional Role and
Foreign Policy (London, 2009) pp.72ff.
88. Wolde-Yesus Ammar, ‘The role of Asmara students in the Eritrean nationalist
movement, 1958–68’, Eritrean Studies Review, 2:1 (1997).
89. Pool, Guerrillas, p.52.
90. See the engrossing account given by Haile Wold’ensae to Dan Connell in the
latter’s Conversations, pp.25ff.
91. Markakis, ‘Nationalist revolution’, pp.56–61; Pool, Guerrillas, pp.49–58; Ruth
Iyob, Eritrean Struggle, pp.109–17.
92. Pool, Guerrillas, p.52
93. Lewis, Modern History, pp.178ff.
94. Markakis, National and Class Conflict, pp.175–80.
95. The best of a very small pool of work on this is Mburu, Bandits; see also Lewis,
Modern History, pp.183ff.
96. I am grateful to Hannah Whittaker, a doctoral research student at the School
of Oriental and African Studies in London, for this analysis.
97. Mburu, Bandits, pp.182, 202–3.
98. Ibid., p.78.
99. Gebru Tareke, Power and Protest, pp.125ff.
100. Ibid., pp.160ff.
101. See the wonderfully thoughtful account of the Ethiopian situation in Messay
Kebede, Radicalism and Cultural Dislocation.

chapter 7
1. Messay Kebede, Radicalism and Cultural Dislocation.
2. For a remarkably sympathetic contemporary assessment, see P. Schwab, Ethiopia:
Politics, Economics and Society (London, 1985). For more critical, in-depth analy-
ses from the 1980s—including a spate of studies published in 1988, virtually on
the eve of the regime’s collapse—see F. Halliday and M. Molyneux, The
Ethiopian Revolution (London, 1981); C. Clapham, Transformation and Continuity in
 e ndnote s 2 77

Revolutionary Ethiopia (Cambridge, 1988); J. W. Harbeson, The Ethiopian

Transformation: The Quest for the Post Imperial State (Boulder and London, 1988);
and E. J. Keller, Revolutionary Ethiopia: From Empire to People’s Republic
(Bloomington and Indianapolis, 1988). Arguably the best single account—and
certainly the best by an Ethiopian scholar—is Andargachew Tiruneh, The
Ethiopian Revolution, 1974–1987: A Transformation from an Aristocratic to a Totalitarian
Autocracy (Cambridge, 1993).
3. S. Decalo, Coups and Army Rule in Africa: Motivations and Constraints (New Haven
and London, 1990) pp.33–4, 293–4.
4. See for example A. Gavshon, Crisis in Africa: Battleground of East and West (New
York, 1981) pp.258ff.
5. C. Legum and B. Lee, The Horn of Africa in Continuing Crisis (New York and
London, 1979).
6. Quoted in Bahru Zewde, Modern Ethiopia, p.255.
7. Gebru Tareke, The Ethiopian Revolution: War in the Horn of Africa (New Haven
and London, 2009) chapter 4.
8. Andargachew Tiruneh, Ethiopian Revolution, pp.345–6.
9. Ibid., pp.211–12; also M. Dines, ‘The Ethiopian ‘Red Terror’’, in Davidson et al.,
Behind the War, pp.60–1.
10. Markakis, National and Class Conflict, pp.237ff; Andargachew Tiruneh, Ethiopian
Revolution, chapter 3, passim.
11. Dines, ‘The Ethiopian “Red Terror’’’, p.61; R. Pateman, ‘Drought, famine and
development’, in L. Cliffe and B. Davidson (eds.), The Long Struggle of Eritrea for
Independence and Constructive Peace (Trenton, NJ, 1988) pp.169ff.
12. G. Hancock, Ethiopia: The Challenge of Hunger (London, 1985); R. D. Kaplan,
Surrender or Starve: Travels in Ethiopia, Sudan, Somalia and Eritrea, 2nd edn. (New
York, 2003) pp.20ff.
13. For an excellent overview, see Gaim Kibreab, ‘Eritrean–Sudanese relations’.
14. See Young, Armed Groups, pp.18ff.
15. Ibid., pp.21–2.
16. The EPDM later became part of the EPRDF coalition, and changed its name
to the Amhara National Democratic Movement (ANDM).
17. Again see Gaim Kibreab’s work on Eritrean refugees in Sudan, notably People
on the Edge in the Horn (Oxford, 1996); and Critical Reflections on the Eritrean War
of Independence (Trenton, NJ, 2008).
18. Again,Young, Armed Groups, provides the best overview; see also the illuminat-
ing piece by Eisei Kurimoto, ‘Fear and Anger: female versus male narratives
among the Anywaa’, in James et al., Remapping Ethiopia.
19. James, ‘No place to hide’.
20. Ibid., pp.263–4.
21. Holt and Daly, History of the Sudan, pp.187ff.
22. Gebru Tareke,‘The Ethiopia–Somalia war of 1977 revisited’, International Journal
of African Historical Studies, 33:3 (2000); Lewis, Modern History, pp.231–48.
278 e ndnote s

23. Gebru Tareke, Ethiopian Revolution, chapter 6 passim; Markakis, National and
Class Conflict, pp.225ff.
24. Legum and Lee, Horn of Africa, pp.68 ff.
25. Lewis, Modern History, pp.239–48.
26. A. Triulzi, ‘Competing views of national identity in Ethiopia’, in I. M. Lewis
(ed.), Nationalism and Self-Determination in the Horn of Africa (London, 1983).
27. M. Dines, ‘Ethiopian violation of human rights in Eritrea’, in Cliffe and
Davidson, Long Struggle, p.148.
28. C. Clapham (ed.), African Guerrillas (Oxford, 1998).
29. Sorenson, Imagining Ethiopia, p.62.
30. Mohammed Hassan, ‘Conquest, tyranny and ethnocide against the Oromo: a
historical assessment of human rights conditions in Ethiopia, ca.1880s–2002’, in
Ezekiel Gebissa, Contested Terrain, pp.30–1.
31. See H. S. Lewis, ‘The development of Oromo political consciousness from 1958
to 1994’, in P. T. W. Baxter et al., Being and Becoming Oromo for an insightful
overview. In the same collection, see also Gemetchu Megerssa, ‘Oromumma:
tradition, consciousness and identity’; and Mohammed Hassan, ‘The develop-
ment of Oromo nationalism’; as well as P. Baxter,‘The creation and constitution
of Oromo nationality’, in K. Fukui and J. Markakis (eds.), Ethnicity and Conflict
in the Horn of Africa (London, 1994).
32. Hassan, ‘Conquest’, pp.31–2.
33. Asafa Jalata, Oromia and Ethiopia, pp.179–80.
34. Hassan, ‘Conquest’, pp.33–6.
35. Levine, Greater Ethiopia.
36. The point is well made in Clapham, ‘Boundary and territory’, p.245.
37. As Gunther Schlee has shown, however, even this can be complicated, as identi-
ties and senses of belonging shift from time to time between ‘Somali’ and ‘Oromo’:
see for example Schlee, ‘Brothers of the Boran’; also his Identities on the Move:
Clanship and Pastoralism in Northern Kenya (Manchester, 1989), and ‘Gada systems
on the meta-ethnic level: Gabbra / Boran / Garre interactions in the Kenyan /
Ethiopian borderland’, in Kurimoto and Simonse, Conflict, Age and Power.
38. Asafa Jalata, Oromia and Ethiopia, p.193.
39. Ibid., pp.193–5.
40. Mohammed Hassan, ‘The development of nationalism’, p.69.
41. Bahru Zewde, Pioneers of Change, p.133.
42. Young, Peasant Revolution, pp.44–9, 99.
43. J.Young, ‘The Tigray and Eritrean Peoples’ Liberation Fronts: a history of ten-
sions and pragmatism’, Journal of Modern African Studies, 34:1 (1996) p.106.
44. The term ‘Agame’ as a pejorative for Tigrayan came to have widespread cur-
rency, even among Amhara: see Alemseged Abbay, Identity Jilted, pp.140, 142;
Young, Peasant Revolution, p.69.The antiquity of the concept is unclear. In 1901,
Wylde noted the ‘saying’ that ‘nothing ever good came out of Agame’: Wylde,
Modern Abyssinia, p.200.
 e ndnote s 279

45. Young, Peasant Revolution, pp.87, 92ff; Markakis, National and Class Conflict,
pp.253ff; Gebru Tareke, Ethiopian Revolution, chapter 3.
46. Young, Peasant Revolution, chapter 4 passim.
47. A vivid first-hand account of this later period is provided by Hammond, Fire
from the Ashes.
48. Gebru Tareke, Ethiopian Revolution, chapter 9.
49. Pool, Guerrillas, pp.87ff; Ruth Iyob, Eritrean Struggle, pp.123ff.
50. The history of the ELF–EPLF civil war remains shrouded, and it will be some
time before a fuller picture of it becomes clear; but see for example Gaim
Kibreab, Critical Reflections.
51. Pool, Guerrillas, chapter 3 passim.
52. Ibid., pp.67ff; Gaim Kibreab, Critical Reflections, chapter 5 passim: Markakis,
National and Class Conflict, pp.131ff.
53. Bereket Habte Selassie, The Crown and the Pen: The Memoirs of a Lawyer Turned
Rebel (Trenton, NJ, 2007) pp.299ff.
54. Pool, Guerrillas, pp.76ff; Gaim Kibreab, Critical Reflections, chapter 7; Kidane
Mengisteab and Okbazghi Yohannes, Anatomy, pp.46ff.
55. Andargachew Tiruneh, Ethiopian Revolution, pp.77–8.
56. For example, see Dines, ‘Ethiopian violation’.
57. Ibid., pp.149–52.
58. Interesting contemporary analysis is offered by R. Sherman, Eritrea: The
Unfinished Revolution (New York, 1980).
59. Awet Weldemichael, ‘The Eritrean long march: the strategic withdrawal of the
Eritrean People’s Liberation Front (EPLF), 1978–79’, Journal of Military History,
73:4 (2009).
60. For a rare published memoir dealing with these events, see Tekeste Fekadu,
Journey from Nakfa to Nakfa: Back to Square One, 1976–1979 (Asmara, 2002).
61. Gebru Tareke, Ethiopian Revolution, chapter 7; for a personal account of military
action in this period, see Alemseged Tesfai, Two Weeks in the Trenches: Reminiscences
of Childhood and War in Eritrea (Lawrenceville, NJ, 2002) pp.43ff.
62. Pool, Guerrillas, Pateman, Eritrea; and D. Connell, Against All Odds: A Chronicle of
the Eritrean Revolution (Lawrenceville, NJ, 1997). See also D. Pool,‘Eritrean nation-
alism’, in Lewis, Nationalism and Self-Determination; and J. Harding, Small Wars,
Small Mercies: Journeys in Africa’s Disputed Nations (London, 1993) chapter 6.
63. Press Department, Ethiopian Ministry of Information, Historical Truth About
Eritrea (Addis Ababa, 1988).
64. Tekeste Fekadu, The Tenacity and Resilience of Eritrea, 1979–1983 (Asmara, 2008);
Connell, Against All Odds, passim; Pool, Guerrillas, part II passim.
65. Pateman, Eritrea, chapter 9.
66. T. Redeker Hepner, Soldiers, Martyrs,Traitors, and Exiles: Political Conflict in Eritrea
and the Diaspora (Philadelphia, 2009).
67. For a useful summary, see D. Pool, ‘The Eritrean People’s Liberation Front’, in
Clapham, African Guerrillas.
280 e ndnote s

68. The ‘National Democratic Programme of the EPLF’, (January 1977), in

Davidson et al., Behind the War, pp.143–50.
69. The key source is the excellent piece by Dan Connell, ‘Inside the EPLF: the
origins of the “People’s Party” and its role in the liberation of Eritrea’, Review
of African Political Economy, 89 (September 2001). It is also appended in Connell’s
Conversations, and indeed it is based on information derived from those
interviews.
70. R. Leonard, ‘Popular participation in liberation and revolution’, in Cliffe and
Davidson, Long Struggle.
71. See R. J. Reid, ‘Old Problems in New Conflicts: some observations on Eritrea
and its relations with Tigray, from liberation struggle to inter-state war’, Africa,
73:3 (2003); and R. J. Reid, ‘ “Ethiopians believe in God, Sha’abiya believe in
mountains”: the EPLF and the 1998–2000 war in historical perspective’, in
D. Jacquin-Berdal and M. Plaut (eds.), Unfinished Business: Ethiopia and Eritrea at
War (Lawrenceville, NJ, 2005). See also Young,‘The Tigray and Eritrean Peoples’
Liberation Fronts’.
72. Exceptions, again, are Young, ‘The Tigray and Eritrean Peoples’ Liberation
Fronts’; Alemseged Abbay, Identity Jilted; and, from an earlier era, the prescient
analysis in P. Gilkes, ‘Centralism and the PMAC’, in Lewis, Nationalism and Self-
Determination in the Horn of Africa.
73. Most markedly, perhaps, Roy Pateman, who in the second edition of his major
book on the Eritrean struggle in 1997–8 unwisely declared that there was no
military threat to Eritrea from Ethiopia as long as the EPRDF remained in
power: Pateman, Eritrea, p.264.
74. Tekeste Negash and K. Tronvoll, Brothers at War: Making Sense of the Eritrean–
Ethiopian War (Oxford, 2000).
75. Reid, ‘Old problems’, pp.376ff; Alemseged Tesfai, ‘“The March of Folly” re-
enacted: a personal view’, Eritrean Studies Review, 3:2 (1999) p.216; D. Connell,
‘Against more odds: the second siege of Eritrea’, Eritrean Studies Review, 3:2
(1999) p.197.
76. Young, ‘The Tigray and Eritrean Peoples’ Liberation Fronts’, p.105.
77. Publications of the EPLF: ‘The TPLF and the development of its relations with
the EPLF’ (c.1984). Research and Documentation Centre, Asmara (hereafter
RDC), Acc. No. 05062/Rela/3 p.4.
78. Publications of the TPLF: ‘The Eritrean struggle, from where to where? An
assessment’ (1985). RDC Acc. No. Rela/10359 pp.42–3.
79. RDC Acc. No. 05062/Rela/3 p.20.
80. Ibid., pp.19–20, 27.
81. RDC Acc. No. Rela/10359 pp.161–2.
82. RDC Acc. No. 05062/Rela/3 p.8; Young, ‘The Tigray and Eritrean peoples’
liberation fronts’, p.106.
83. RDC Acc. No. Rela/10359 p.51.
84. RDC Acc. No. 05062/Rela/3 pp.4, 13.
 e ndnote s 2 81

85. Ibid., p.18.

86. RDC Acc. No. Rela/10359 pp.81–2.
87. Foreign Relations Bureau of the TPLF: ‘On our Differences with the EPLF’
(1986). RDC Acc. No. 2399 p.3.
88. Clapham, Transformation and Continuity, p.212.
89. ‘National Democratic Programme, Eritrean People’s Liberation Front’ (March
1987), in Cliffe and Davidson, Long Struggle, pp.205–13.
90. Tekeste Fekadu, Tenacity and Resilience, pp.207ff.
91. Gebru Tareke, Ethiopian Revolution, chapter 7; Alemseged Tesfai, Two Weeks,
pp.99ff.
92. Quoted in Connell, Against All Odds, p.228.
93. Gebru Tareke, Ethiopian Revolution, chapter 10.

chapter 8
1. Ruth Iyob, Eritrean Struggle, pp.137ff.
2. Ibid., pp.138–40; Pool, Guerrillas, pp.161ff.
3. Ruth Iyob, Eritrean Struggle, pp.141–2.
4. Quoted in ibid., p.143.
5. The spirit of this period is captured in Connell, Against All Odds, pp.263ff and
279ff; Pateman, Eritrea, chapter 11; and Ruth Iyob, Eritrean Struggle, chapters 8
and 9.
6. Connell, Against All Odds, pp.287ff.
7. In ibid., p.251.
8. The Constitutional Commission of Eritrea, Information on Strategy, Plans and
Activities (Asmara, October 1995).
9. Gaim Kibreab, Eritrea: a dream deferred (Woodbridge, 2009) chapter 2 passim.
10. Pateman, Eritrea, pp.248, 250, 260–1; Connell, Against All Odds, pp.294ff.
11. The single best analysis to date is D. Turton (ed.), Ethnic Federalism: the Ethiopian
experience in comparative perspective (Oxford, 2006). See also, for example, J. Abbink,
‘Breaking and making the state: the dynamics of ethnic democracy in Ethiopia’,
Journal of Contemporary African Studies, 13:2 (1995); J. Abbink, ‘Ethnicity and con-
stitutionalism in contemporary Ethiopia’, Journal of African Law, 41 (1997);
J. M. Cohen,‘“Ethnic federalism” in Ethiopia’, Northeast African Studies, 2:2 (1995);
J. Young, ‘Ethnicity and power in Ethiopia’, Review of African Political Economy,
23:70 (1996); Kidane Mengisteab,‘New approaches to state building in Africa: the
case of Ethiopia’s ethnic-based federalism’, African Studies Review, 40:3 (1997).
12. For example, Henze, Layers of Time.
13. Marcus, History, pp.231ff.
14. Ibid., p.240.
15. J. Prendergast and M. Duffield, ‘Liberation politics in Ethiopia and Eritrea’, in
T. M. Ali and R. O. Matthews (eds.), Civil Wars in Africa: Roots and Resolution
(Montreal and Kingston, 1999) p.49.
282 e ndnote s

16. D. Donham, ‘Introduction’, in James et al., Remapping Ethiopia, p.6.

17. For example, Merera Gudina, ‘Contradictory interpretations of Ethiopian his-
tory: the need for a new consensus’, in Turton, Ethnic Federalism; various contri-
butions to James et al., Remapping Ethiopia; K.Tronvoll,‘Human rights violations
in federal Ethiopia: when ethnic identity is a political stigma’, International
Journal on Minority and Group Rights, 15 (2008).
18. K. Tronvoll, ‘Ambiguous identities: the notion of war and “significant others”
among the Tigreans of Ethiopia’, in V. Broch-Due (ed.), Violence and Belonging:
The Quest for Identity in Post-colonial Africa (London, 2005); and his War and the
Politics of Identity in Ethiopia:The Making of Enemies and Allies in the Horn of Africa
(Oxford, 2009).
19. S. Pausewang, K.Tronvoll, and L. Aalen (eds.), Ethiopia since the Derg: A Decade of
Democratic Pretensions and Performance (London, 2002).
20. Marcus, History, p.244;Young, Peasant Revolution, pp.211ff.
21. The author was told by General Tsadkan Gebretensae that he felt he was denied
the opportunity to complete the mission in Eritrea: interview (Addis Ababa, 14
September 2005). The split, indeed, divided Tigrayans more generally, and gave
rise to new opposition groups in exile. An example is the ‘Tigrayan International
Solidarity for Justice and Democracy’, which has accused the Ethiopian gov-
ernment of giving away Ethiopian territory to the ‘tyrant’ Isaias.
22. See for example Human Rights Watch, ‘“Why am I still here?”The 2007 Horn
of Africa renditions and the fate of those still missing’ (1 October 2008); and
A. Mitchell, ‘US agents interrogating terror suspects held in Ethiopian prisons’,
International Herald Tribune (4 April 2007); and A. Mitchell, ‘US agents visit secret
Ethiopian jails’, Mail and Guardian Online (4 April 2007).
23. R. Paz, ‘The youth are older: the Iraqization of the Somali Mujahidin Youth
Movement’, The Project for the Research of Islamist Movements: Occasional Papers,
6:2 (2008); R. Marchal, ‘Islamic political dynamics in the Somali civil war’, in
de Waal, Islamism.
24. R. J. Reid, ‘A fierce race’, History Today, 50:6 (2000); R. J. Reid, ‘Revisiting
primitive war: perceptions of violence and race in history’, War and Society, 26:2
(2007); Reid, War, esp. chapter 1.
25. Most obviously, Tekeste Negash and Tronvoll, Brothers at War.
26. P. Gilkes and M. Plaut, War in the Horn: The Conflict between Eritrea and Ethiopia
(London, 1999); D. Jacquin-Berdal and M. Plaut (eds.), Unfinished Business:
Ethiopia and Eritrea at War (Lawrenceville, NJ, 2005);Tronvoll, War and the Politics
of Identity; Ruth Iyob, ‘The Ethiopian–Eritrean conflict: diasporic vs. hegem-
onic states in the Horn of Africa, 1991–2000’, Journal of Modern African Studies,
38:4 (2000).
27. See Pateman, Eritrea, pp.235ff; Ruth Iyob, Eritrean Struggle, pp.136–7.
28. See appendices 1 and 2 in Tekeste Negash and Tronvoll, Brothers at War; Amare
Tekle, ‘The basis of Eritrean–Ethiopian cooperation’, in Amare Tekle (ed.),
Eritrea and Ethiopia: From Conflict to Cooperation (Lawrenceville, NJ, 1994).
 e ndnote s 2 83

29. D. Styan,‘Twisting Ethio-Eritrean economic ties: misperceptions of war and the

misplaced priorities of peace, 1997–2002’, in Jacquin-Berdal and Plaut, Unfinished
Business.
30. Tekeste Negash and Tronvoll, Brothers at War, pp.23–9.
31. Appendix 3 in ibid.
32. More detailed accounts can be found in Tekeste Negash and Tronvoll, Brothers
at War; and Gilkes and Plaut, War in the Horn.
33. Interview with General Tsadkan Gebretensae (Addis Ababa, 14 September
2005).
34. In 1999, the author conducted extensive research among Ethiopians living in
Asmara as part of a report for the International Labour Organisation in
Geneva.
35. Much of this analysis is based on personal observation, as the author was resi-
dent in Eritrea throughout this period.
36. This extensive document can be viewed at http://www.un.org/NewLinks/
eebcarbitration/EEBC-Decision.pdf .
37. A number of useful interim reports have been produced, including International
Crisis Group, Ethiopia and Eritrea: Preventing War, Africa Report No.101 (22
December 2005); and S. Hally and M. Plaut, Ethiopia and Eritrea: Allergic to
Persuasion (London, 2007).
38. P. Gilkes, ‘Violence and identity along the Eritrean–Ethiopian border’, in
Jacquin-Berdal and Plaut, Unfinished Business; Tronvoll, ‘Ambiguous identities’;
Tronvoll, War and the Politics of Identity.
39. Tekeste Negash, Eritrea and Ethiopia.
40. Alemseged Abbay, Identity Jilted.
41. Addis Birhan, Eritrea: A Problem Child of Ethiopia (Addis Ababa, 1998).
42. This is the basic purpose behind the ostensibly ‘independent’Walta Information
Centre’s publication of Chronology of the Ethio-Eritrean Conflict and Basic
Documents (Addis Ababa, 2001).
43. This was repeated to me on many occasions by Tigrayan and Amhara inform-
ants in Addis Ababa during trips in 2005 and 2006.
44. Medhane Tadesse, The Eritrean–Ethiopian War: Retrospect and Prospects (Addis
Ababa, 1999).
45. Tronvoll, War and the Politics of Identity, pp.139ff.
46. A. Triulzi, ‘The past as contested terrain: commemorating new sites of memory
in war-torn Ethiopia’, in P. Kaarsholm (ed.), Violence, Political Culture and
Development in Africa (Oxford, 2006). But for a somewhat quirkier, more polem-
ical interpretation, see Mesfin Araya, ‘Contemporary Ethiopia in the context of
the battle of Adwa, 1896’, in Paulos Milkias and Getachew Metaferia (eds.), The
Battle of Adwa (New York, 2005).
47. Hilton, Ethiopian Patriots, pp.68–70, 185; also Tronvoll, War and the Politics of
Identity, chapters 5 and 6, passim.
48. Tronvoll, War and the Politics of Identity, pp.61–84.
284 e ndnote s

49. Reid, ‘Old problems’.

50. Tekie Fessehatzion, ‘Eritrea and Ethiopia: from conflict to cooperation to con-
flict’ and Jordan Gebre-Medhin, ‘Eritrea (Mereb-Melash)’, both in Eritrean
Studies Review, 3:2 (1999).
51. Alemseged Tesfai, ‘“The March of Folly”’ and D.Connell, ‘Against more odds’.
52. These views, gathered during extensive periods of fieldwork in Eritrea, are
presented in Reid, ‘Old problems’.
53. For example, Gaim Kibreab, ‘Mass expulsion of Eritreans and Ethiopians of
Eritrean origin from Ethiopia and human rights violations’, Eritrean Studies
Review, 3:2 (1999).
54. The full text is in Connell, Conversations.
55. Gaim Kibreab, Eritrea, pp.36ff.
56. This is well documented, for example, in Kidane Mengisteab and Okbazghi
Yohannes, Anatomy; and, again, the rich material in Connell, Conversations.
57. Another high profile victim, indeed, was the Orthodox Abuna, the head of the
Church, arrested and detained in 2006. See for example United States
Commission on International Religious Freedom, Annual Report (Washington,
DC, May 2009) pp.28ff.
58. See Human Rights Watch, Service for Life: State Repression and Indefinite
Conscription in Eritrea (New York, April 2009); and a number of contributions in
D. O’Kane and T. Redeker Hepner (eds.), Biopolitics, Militarism and Development:
Eritrea in the Twenty-First Century (New York and Oxford, 2009).
59. Gaim Kibreab, Eritrea, pp.310ff.
60. ‘Eritrea’s controversial push to feed itself ’, BBC News, 24 December 2009.
61. I explore this in more detail in my ‘Caught in the headlights of history: Eritrea,
the EPLF and the post-war nation-state’, Journal of Modern African Studies, 43:3
(2005), and in a follow-up to that piece, ‘The politics of silence’.
62. See for example T. Muller, ‘Human resource development and the state: higher
education in post- revolutionary Eritrea’, and M. Treiber, ‘Trapped in adoles-
cence: the post-war urban generation’, both in O’Kane and Redeker Hepner,
Biopolitics.
63. Reid, ‘The politics of silence’, passim.
64. Reid, ‘Caught in the headlights’, p.478.
65. See the various contributions in Reid, Eritrea’s External Relations for an in-
depth exploration of these issues.
66. See also R. J. Reid, ‘Eritrea’s regional role and foreign relations: past and present
perspectives’, in ibid.
67. Quoted in Connell, Against All Odds, pp.282–3.
68. Gaim Kibreab, ‘Eritrean-Sudanese relations’, passim.
69. For a summary, see Kidane Mengisteab and Okbazghi Yohannes, Anatomy,
pp.216ff.
70. Institute for Security Studies Situation Report, The Eritrea–Djibouti Border
Dispute (15 September 2008).
 e ndnote s 2 85

71. D. Connell, ‘Eritrea and the United States: towards a new US policy’, in Reid,
Eritrea’s External Relations, pp.135ff.
72. Foremost among these critiques are: Gaim Kibreab, Critical Reflections and his
sequel to that volume, Eritrea; Kidane Mengisteab and Okbazghi Yohannes,
Anatomy; and Yohannes Gebremedhin, The Challenges of a Society in Transition:
Legal Development in Eritrea (Trenton, NJ, 2004).
73. Readers familiar with it will perhaps observe that I owe much in the way of
inspiration here to John Lonsdale’s work on Mau Mau: see, in particular,
B. Berman and J. Lonsdale, Unhappy Valley: Conflict in Kenya and Africa. Book Two:
Violence and Ethnicity (London, 1992).
74. Again, see in particular Sorenson, Imagining Ethiopia; Holcomb and Ibssa, Invention
of Ethiopia; Gebissa, Contested Terrain; Baxter et al., Being and Becoming Oromo.
75. The case is made, for example, in Holcomb and Ibssa, Invention of Ethiopia. The
rather more complex interaction between Amhara, Tigrinya, and Oromo over
the preceding 300 years was of less interest to nationalists.
76. See for example ‘Oromo talks’, Africa Confidential, 38(21) (October 1997).
77. A useful overview can be found in International Crisis Group, Ethiopia: ethnic
federalism and its discontents, Africa Report No.153 (4 September 2009).
78. For example, Amnesty International, ‘Ethiopia’, in Annual Report 2003
(London, 2003). See also Mohammed Hassan, ‘Conquest’; Asafa Jalata, Oromia
and Ethiopia, pp.229ff.
79. Oromo Forum for Dialogue and Reconciliation, A Proposal for the Consolidation
of Oromo Forces for Liberation, Peace and Prosperity in the Horn of Africa (Melbourne,
2008).
80. Ibid., p.3.
81. The author interviewed Oromo prisoners of war near Nakfa, northern Eritrea,
in September 2000.
82. This was evident on several trips made by the author to the former front line
areas in July and August 2000.
83. Tronvoll, War and the Politics of Identity, pp.189ff.
84. A useful overview of the 1990s and early 2000s is in Lewis, Modern History,
chapter 11; see also M. Bradbury, Becoming Somaliland (Oxford, 2008), and of
course the excellent work continually produced by Ken Menkhaus, including
Country in Peril, and shorter pieces in the Horn of Africa Bulletin, notably January
2008 and August 2008.
85. K. Menkhaus, ‘Somalia and Somaliland: terrorism, political Islam and state
collapse’, in R. I. Rotberg (ed.), BattlingTerrorism in the Horn of Africa (Washington,
DC, 2005); Paz, ‘The youth are older’.
86. An excellent overview of a complicated situation is provided in Healy, Lost
Opportunities, pp.20ff.
87. See for example T. Hagmann, ‘Beyond clannishness and colonialism: under-
standing political disorder in Ethiopia’s Somali region, 1991–2004’, Journal of
Modern African Studies, 43:4 (2005).
286 e ndnote s

88. Human Rights Watch, Collective Punishment: War Crimes and Crimes against
Humanity in the Ogaden Area of Ethiopia’s Somali Region (New York, June 2008).
89. For example, Human Rights Watch, ‘Ethiopia’s Dirty War’ (5 August 2007), and
‘Somalia: war crimes devastate population’ (8 December 2008); Amnesty
International, Routinely Targeted: Attacks on Civilians in Somalia (London, May
2008). See also Menkhaus, Country in Peril.
90. Human Rights Watch, ‘Ethiopia’, January 2009, available at: <http://www.hrw.
org/sites/default/files/related_material/Ethiopia.pdf>.
91. D. Connell, ‘Eritrea: on a slow fuse’, in Rotberg, Battling Terrorism; D. Connell,
‘The EPLF/PFDJ Experience: how it shapes Eritrea’s regional strategy’, in
Reid, Eritrea’s External Relations.
92. A. McGregor, Who’s Who in the Somali Insurgency:A Reference Guide (Washington,
DC: The Jamestown Foundation, September 2009).
93. Among the best accounts to date are R. Middleton, Piracy in Somalia: Threatening
Global Trade, Feeding Local Wars. Chatham House Briefing Paper (London: Chatham
House, October 2008); and K. Menkhaus, ‘Dangerous waters’, Survival, 51:1 (2009).
94. K. Menkhaus, J. Prendergast and C.Thomas-Jensen, ‘Beyond Piracy: Next Steps
to Stabilise Somalia’, Enough! (Washington, DC, May 2009).
95. The best account remains Bradbury, Becoming Somaliland.
96. Iqbal Jhazbhay, ‘Islam and stability’.

epilogue
1. Many such communities are examined in much greater detail than can be
offered here in: Donham and James, Southern Marches and its successor volume,
James et al., Remapping Ethiopia; and Fukui and Markakis, Ethnicity and Conflict.
2. See for example R. Love, Economic Drivers of Conflict and Cooperation in the Horn
of Africa: A Regional Perspective and Overview, Chatham House Briefing Paper
(London, December 2009).
3. See a number of the contributions to A.de Waal (ed.), War in Darfur and the
Search for Peace (London, 2007); and G. Prunier, Armed Movements in Sudan,
Chad, CAR, Somalia, Eritrea and Ethiopia, Center for International Peace
Operations (Berlin, 2008).
4. The best recent survey is provided by Young, Armed Groups.
5. See also J.Young, ‘Along Ethiopia’s western frontier: Gambella and Benishangul
in transition’, Journal of Modern African Studies, 37:2 (1999).
6. Human Rights Watch, Targeting the Anuak: Human Rights Violations and Crimes
Against Humanity in Ethiopia’s Gambella Region (New York, March 2005).
7. J. Markakis, ‘Ethnic conflict and the state in the Horn of Africa’, in Fukui and
Markakis, Ethnicity and Conflict, p.221.
8. Amnesty International, Annual Report 2003; ‘ “Deaths” in Ethiopian demonstra-
tion’, BBC News (Africa, 24 May 2002).
9. ‘Ethiopian Officials Held Over Violence’, by Nita Bhalla, BBC News (Africa,
21 August 2002).
 Bibliography

a note on oral informants and personal information

Much of the analysis in this book, particularly that in Part IV, is based on numerous
conversations and interviews—informal and otherwise—with a range of groups
and individuals in both Eritrea and Ethiopia since 1997. It is neither possible nor
politic to list them here; the majority, indeed—especially in Eritrea—must remain
anonymous. Where appropriate I have noted oral sources in endnotes.

i archival collections
National Archives of the United Kingdom, London, UK:
Foreign Office (FO) 1 and 881 series.
War Office (WO) 230 series.
Research and Documentation Centre, Asmara, Eritrea:
Box/293 File SH/20
Box/287 File 70
Rela/10359
05062/Rela/3
Box/2399

ii published primary sources

Alazar Tesfa Michael. Eritrea To-Day: Fascist Oppression Under the Nose of British
Military (Woodford Green: New Times Book Dept, c.1946).
Alemseged Tesfai. Two Weeks in the Trenches: Reminiscences of Childhood and War in
Eritrea (Lawrenceville, NJ: Red Sea Press, 2002).
Amnesty International. ‘Ethiopia’, Annual Report 2003 (London: Amnesty
International, 2003).
Amnesty International. Routinely Targeted: Attacks on Civilians in Somalia (London:
Amnesty International, May 2008).
Arab Faqih. The Conquest of Abyssinia (tr. P. L. Stenhouse) (Hollywood, CA: Tsehai,
2003).
Arrowsmith-Brown, J. H. (ed. and tr.). Prutky’s Travels in Ethiopia and Other Countries
(London: The Hakluyt Society, 1991).
2 88 b i bl i og raphy

Bairu Tafla (ed. and tr.). A Chronicle of Emperor Yohannes IV (1872–89) (Wisbaden:
Franz Steiner Verlag GMBH, 1977).
Beckingham, C. F. and G. W. B. Huntingford (eds. and trs.). Some Records of Ethiopia
1593–1646: Being Extracts From ‘The History of High Ethiopia or Abassia’, by Manoel
de Almeida (London: Hakluyt Society, 1954).
Bereket Habte Selassie. The Crown and the Pen:The Memoirs of a Lawyer Turned Rebel
(Trenton: Red Sea Press, 2007).
Berkeley, G. F.-H. The Campaign of Adowa and the Rise of Menelik, 2nd edn. (New
York: Negro Universities Press, 1969).
Blanc, H. A Narrative of Captivity in Abyssinia, 2nd edn. (London: Frank Cass, 1970).
Brooks, M. (ed. and tr.). A Modern Translation of the Kebra Negast (The Glory of the
Kings) (Lawrenceville: Red Sea Press, 1995).
Bruce, J. Travels to Discover the Source of the Nile, in the Years 1768, 1769, 1770, 1771, 1772,
and 1773, 5 vols. (London: G. G. J. and J. Robinson, 1790).
Budge, E. A. Wallis. A History of Ethiopia, Nubia and Abyssinia, 2 vols. (London:
Methuen, 1928).
Burton, R. First Footsteps in East Africa, Or, An Exploration of Harar (New York:
Dover, 1987).
Bulatovich, A. Ethiopia Through Russian Eyes: Country in Transition 1896–1898 (ed. and
tr. R. Seltzer) (Lawrenceville: Red Sea Press, 2000).
Connell, D. Against All Odds: A Chronicle of the Eritrean Revolution, 2nd edn.
(Lawrenceville: Red Sea Press, 1997).
Connell, D. Conversations with Eritrean Political Prisoners (Trenton: Red Sea Press,
2005).
Constitutional Commission of Eritrea. Information on Strategy, Plans and Activities
(Asmara, October 1995).
Cumming, D. ‘The UN disposal of Eritrea’, African Affairs, 52:207 (1953).
Dufton, H. Narrative of a Journey Through Abyssinia in 1862–63, 2nd edn. (Westport:
Negro Universities Press, 1970).
Eritrean People’s Liberation Front. ‘National democratic programme of the EPLF’,
January 1977, in B. Davidson et al. (eds.), Behind the War in Eritrea (Nottingham:
Spokesman, 1980).
Eritrean People’s Liberation Front. ‘National democratic programme’, March 1987,
in L. Cliffe and B. Davidson (eds.), The Long Struggle of Eritrea for Independence and
Constructive Peace (Trenton: Red Sea Press, 1988).
Farago, L. Abyssinia on the Eve (London: Putnam, 1935).
Foreign Office (UK). Handbook [on Eritrea] Prepared Under the Direction of the
Historical Section of the Foreign Office, No.126 (London: H.M. Stationery Office,
1920).
Gobat, S. Journal of Three Years’ Residence in Abyssinia, 2nd edn. (New York: Negro
Universities Press, 1969).
Hammond, J. Fire From the Ashes:A Chronicle of Revolution in Tigray, Ethiopia, 1975–1991
(Lawrenceville: Red Sea Press, 1999).
 b i bl i og raphy 289

Hancock, G. Ethiopia:The Challenge of Hunger (London: Gollancz, 1985).

Harding, J. Small Wars, Small Mercies: Journeys in Africa’s Disputed Nations (London:
Viking, 1993).
Harmsworth, G. Abyssinian Adventure (London: Hutchinson, 1935).
Hilton, A. (ed.). The Ethiopian Patriots: Forgotten Voices of the Italo-Abyssinian War,
1935–41 (Stroud: Spellmount, 2007).
Human Rights Watch. Targeting the Anuak: Human Rights Violations and Crimes
Against Humanity in Ethiopia’s Gambella Region (New York, March 2005).
Human Rights Watch. ‘Ethiopia’s dirty war’ (5 August 2007).
Human Rights Watch. Collective Punishment:War Crimes and Crimes Against Humanity
in the Ogaden Area of Ethiopia’s Somali Region (New York, June 2008).
Human Rights Watch. ‘Somalia: war crimes devastate population’ (8 December 2008).
Human Rights Watch. Service for Life: State Repression and Indefinite Conscription in
Eritrea (New York, April 2009).
Huntingford, G. W. B. (ed. and tr.). The Glorious Victories of Amda Seyon, King of
Ethiopia (London: Oxford University Press, 1965).
Kaplan, R. D. Surrender or Starve: Travels in Ethiopia, Sudan, Somalia and Eritrea, 2nd
edn. (New York:Vintage, 2003).
Kolmodin, J. (tr.). Traditions de Tsazzega et Hazzega (Uppsala: Archives d’Etudes
Orientales, 1915).
Longrigg, S. H. A Short History of Eritrea (Oxford: Clarendon Press, 1945).
Longrigg, S. H. ‘The future of Eritrea’, African Affairs, 45(180) (1946).
Macmichael, Sir H. The Anglo-Egyptian Sudan (London: Faber and Faber, 1934).
Markham, C. R. A History of the Abyssinian Expedition (London: Macmillan, 1869).
Ministry of Information, UK. The First to be Freed: The Record of British Military
Administration in Eritrea and Somalia, 1941–1943 (London: H.M. Stationery Office,
1944).
Ministry of Information, Ethiopia. Historical Truth About Eritrea (Addis Ababa:
Ethiopian Ministry of Information, 1988).
Mitchell, Sir P. African Afterthoughts (London: Hutchinson, 1954).
Nadel, S. F. Races and Tribes of Eritrea (Asmara: British Military Administration, 1944).
Oromo Forum for Dialogue and Reconciliation. A Proposal for the Consolidation of
Oromo Forces for Liberation, Peace and Prosperity in the Horn of Africa (Melbourne,
August 2008).
Othman Saleh Sabbe. The History of Eritrea (tr. Muhamad Fawaz al-Azem) (Beirut:
n.p., c.1974?)
Pankhurst, R. (ed.). The Ethiopian Royal Chronicles (Addis Ababa: Oxford University
Press, 1967).
Pankhurst, S. British Policy in Eritrea and Northern Ethiopia (Woodford Green: n.p.,
1946).
Pankhurst, S. Eritrea on the Eve: The Past and Future of Italy’s ‘First-born’ Colony,
Ethiopia’s Ancient Sea Province (Woodford Green: New Times and Ethiopia News
Books, 1952).
290 b i bl i og raphy

Pankhurst, S. and R. Pankhurst. Ethiopia and Eritrea: The Last Phase of the Reunion
Struggle, 1941–1952 (Woodford Green: Lalibela House, 1953).
Parkyns, M. Life in Abyssinia: Being Notes Collected During Three Years’ Residence and
Travels in that Country, 3rd edn. (London: Frank Cass, 1966).
Perham, M. The Government of Ethiopia (London: Faber and Faber, 1948).
Plowden,W. C. Travels in Abyssinia and the Galla Country (ed.T. C. Plowden) (London:
Longmans, Green, 1868).
Portal, G. H. My Mission to Abyssinia, 2nd edn. (New York: Negro Universities Press,
1969).
Rassam, H. Narrative of the British Mission to Theodore, King of Abyssinia, 2 vols.
(London: John Murray, 1869).
Rubenson, S. et al. (eds. and trs.). Acta Aethiopica I: Correspondence and Treaties, 1800–
1854 (Evanston: Northwestern University Press, 1987).
Rubenson, S. et al. (eds. and trs.). Acta Aethiopica II: Tewodros and his Contemporaries,
1855–1868 (Addis Ababa: Addis Ababa University Press, 1994).
Rubenson, S. et al. (eds. and trs). Act Aethiopica III: Internal Rivalries and Foreign
Threats, 1869–1879 (Addis Ababa: Addis Ababa University Press, 2000).
Salt, H. A Voyage to Abyssinia and Travels into the Interior of that Country (London: F. C.
and J. Rivington, 1814).
Stanley, H. M. Coomassie and Magdala: The Story of Two British Campaigns in Africa
(New York: Harper and Brothers, 1874).
Stern, H. Wanderings Among the Falashas in Abyssinia, 2nd edn. (London: Frank Cass,
1968).
Tekeste Fekadu. Journey from Nakfa to Nakfa: Back to Square One, 1976–1979 (Asmara:
n.p., 2002).
Tekeste Fekadu. The Tenacity and Resilience of Eritrea, 1979–1983 (Asmara: Hidri,
2008).
Thesiger, W. The Danakil Diary: Journeys Through Abyssinia (London: Flamingo,
1998).
Thomson, B. Ethiopia: The Country that Cut Off Its Head. A Diary of the Revolution
(London: Robson Books, 1975).
Trevaskis, G. K. N. Eritrea: A Colony in Transition (London: Oxford University Press,
1960).
Ullendorff, E. (ed. and tr.). The Autobiography of Emperor Haile Selassie I: ‘My Life and
Ethiopia’s Progress’, 1892–1937 (London: Oxford University Press, 1976).
United States Commission on International Religious Freedom. Annual Report
(Washington, May 2009).
Walta Information Centre. Chronology of the Ethio-Eritrean Conflict and Basic
Documents (Addis Ababa: Walta Information Centre, 2001).
Waugh, E. Waugh in Abyssinia (Baton Rouge: Louisiana State University Press,
2007).
Weld Blundell, H. The Royal Chronicle of Abyssinia, 1769–1840 (Cambridge: Cambridge
University Press, 1922).
 b i bl i og raphy 2 91

Woodward, P. (ed.). British Documents on Foreign Affairs: Reports and Papers from the
Foreign Office Confidential Print. Part III, from 1940 through 1945. Series G, Africa. 5
vols. (Bethesda: University Publications of America, 1998).
Woodward, P. (ed.). British Documents on Foreign Affairs: Reports and Papers from the
Foreign Office Confidential Print. Part IV, from 1945 through 1950. Series G, Africa. 5
vols. (Bethesda: University Publications of America, 1999–).
Wylde, A. B. Modern Abyssinia, 2nd edn. (Westport: Negro Universities Press, 1970).
Zoli, C. ‘The organisation of Italy’s East African empire’, Foreign Affairs, 16:1
(1937).

iii other sources

Abbink, J. ‘Breaking and making the state: the dynamics of ethnic democracy in
Ethiopia’, Journal of Contemporary African Studies, 13:2 (1995).
Abbink, J. ‘Ethnicity and constitutionalism in contemporary Ethiopia’, Journal of
African Law, 41 (1997).
Abir, M.‘The emergence and consolidation of the monarchies of Enarea and Jimma
in the first half of the nineteenth century’, Journal of African History, 6:2 (1965).
Abir, M. Ethiopia: the era of the princes (London: Longmans, 1968).
Abir, M. ‘Ethiopia and the Horn of Africa’, in R. Gray (ed.), Cambridge History of
Africa:Vol. 4, from c.1600 to c.1790 (Cambridge: Cambridge University Press, 1975).
Abir, M. Ethiopia and the Red Sea: The Rise and Decline of the Solomonic Dynasty and
Muslim–European Rivalry in the Region (London: Frank Cass, 1980).
Addis Birhan. Eritrea: A Problem Child of Ethiopia (Addis Ababa: Marran Books, 1998).
Addis Hiwet. Ethiopia: From Autocracy to Revolution (London: Review of African
Political Economy, 1975).
Ahmad, Abdussamad H. ‘Trading in slaves in Bela-Shangul and Gumuz, Ethiopia:
border enclaves in history, 1897–1938’, Journal of African History, 40:3 (1999).
Ahmed, H. Islam in Nineteenth-Century Wallo, Ethiopia: Revival, Reform and Reaction
(Leiden: Brill, 2001).
Alemseged Abbay. ‘The trans-Mereb past in the present’, Journal of Modern African
Studies, 35:2 (1997).
Alemseged Abbay. Identity Jilted or Re-Imagining Identity? The Divergent Paths of the
Eritrean and Tigrayan Nationalist Struggles (Lawrenceville: Red Sea Press, 1998).
Alemseged Tesfai. ‘“The March of Folly” re-enacted: a personal view’, Eritrean
Studies Review, 3:2 (1999).
Alemseged Tesfai. Aynefalale (Asmara: Hidri Publishers, 2001).
Almagor, U. ‘Institutionalising a fringe periphery: Dassanetch-Amhara relations’, in
D. Donham and W. James (eds.), The Southern Marches of Imperial Ethiopia, 2nd edn.
(Oxford: James Currey, 2002).
Amare Tekle. ‘The basis of Eritrean–Ethiopian cooperation’, in Amare Tekle (ed.),
Eritrea and Ethiopia: From Conflict to Cooperation (Lawrenceville: Red Sea Press,
1994).
292 b i bl i og raphy

Andargachew Tiruneh. The Ethiopian Revolution, 1974–1987: A Transformation from an

Aristocratic to a Totalitarian Autocracy (Cambridge: Cambridge University Press: 1993).
Anonymous. ‘Oromo talks’, Africa Confidential, 38(21) (October 1997).
Asafa Jalata. Oromia and Ethiopia: State Formation and Ethnonational Conflict, 1868–2004,
2nd edn. (Trenton: Red Sea Press, 2005).
Asmarom Legesse. Gada: Three Approaches to the Study of African Society (New York:
Free Press, 1973).
Asmarom Legesse. Oromo Democracy: An Indigenous African Political System (Trenton:
Red Sea Press, 2006).
Awet Weldemichael. ‘The Eritrean Long March: the strategic withdrawal of the
Eritrean People’s Liberation Front (EPLF), 1978–79’, Journal of Military History,
73:4 (2009).
Bahru Zewde. A History of Modern Ethiopia, 1855–1991 (Oxford: James Currey, 2001).
Bahru Zewde. Pioneers of Change in Ethiopia: The Reformist Intellectuals of the Early
Twentieth Century (Oxford: James Currey, 2002).
Bairu Tafla (ed.). ‘Eritrea on the Eve of Colonial Rule’, special issue of Eritrean
Studies Review, 5:1 (2007).
Barker, A. J. Eritrea 1941 (London: Faber and Faber, 1966).
Barnes, C. ‘The Somali Youth League, Ethiopian Somalis and the Greater Somalia
idea, c.1946–48’, Journal of Eastern African Studies, 1:2 (2007).
Bates, D. The Abyssinian Difficulty: the Emperor Theodorus and the Magdala Campaign,
1867–68 (Oxford: Oxford University Press, 1979).
Baxter, P. ‘The creation and constitution of Oromo nationality’, in K. Fukui and
J. Markakis (eds.), Ethnicity and Conflict in the Horn of Africa (London: James Currey,
1994).
Baxter, P., J. Hultin and A. Triulzi (eds.). Being and Becoming Oromo: Historical and
Anthropological Enquiries (Uppsala: Nordiska Afrikainstitutet, 1996).
Bereket Habte Selassie. ‘From British rule to federation and annexation’, in
B. Davidson, L. Cliffe and Bereket Habte Selassie (eds.), Behind the War in Eritrea
(Nottingham: Spokesman, 1980).
Berman, B. and J. Lonsdale. Unhappy Valley: Conflict in Kenya and Africa. Book Two:
Violence and Ethnicity (London: James Currey, 1992).
Bradbury, M. Becoming Somaliland (Oxford: James Currey, 2008).
Carnochan, W. B. Golden Legends: Images of Abyssinia, Samuel Johnson to Bob Marley
(Stanford: Stanford University Press, 2008).
Cassanelli, L. V. The Shaping of Somali Society: Reconstructing the History of a Pastoral
People, 1600–1900 (Philadelphia: University of Pennsylvania Press, 1982).
Caulk, R. ‘Firearms and princely power in Ethiopia in the nineteenth century’,
Journal of African History, 13:4 (1972).
Caulk, R. ‘Armies as predators: soldiers and peasants in Ethiopia, c.1850–1935’,
International Journal of African Historical Studies, 11:3 (1978).
Caulk, R. ‘Bad men of the borders: shum and shifta in northern Ethiopia in the
nineteenth century’, International Journal of African Historical Studies, 17:2 (1984).
 b i bl i og raphy 2 93

Caulk, R. ‘Ethiopia and the Horn’, in A. D. Roberts (ed.), Cambridge History of

Africa,Vol. 7: 1905–1940 (Cambridge: Cambridge University Press, 1986).
Caulk, R. ‘“Black snake, white snake”: Bahta Hagos and his revolt against Italian
overrule in Eritrea, 1894’, in D. Crummey (ed.), Banditry, Rebellion and Social
Protest in Africa (London: James Currey, 1986).
Clapham, C. Transformation and Continuity in Revolutionary Ethiopia (Cambridge:
Cambridge University Press, 1988).
Clapham, C. ‘Boundary and territory in the Horn of Africa’, in A. I. Asiwaju and
P. Nugent (eds.), African Boundaries: Barriers, Conduits and Opportunities (London:
Pinter, 1996).
Clapham, C. Africa and the International System:The Politics of State Survival (Cambridge:
Cambridge University Press, 1996).
Clapham, C. (ed.). African Guerrillas (Oxford: James Currey, 1998).
Cliffe, L. and B. Davidson (eds.). The Long Struggle of Eritrea for Independence and
Constructive Peace (Trenton: Red Sea Press, 1988).
Coffey, T. M. Lion by the Tail:The Story of the Italian–Ethiopian War (London: Hamish
Hamilton, 1974).
Cohen, J. M. ‘“Ethnic federalism” in Ethiopia’, Northeast African Studies, 2:2 (1995).
Connell, D. ‘Against more odds: the second siege of Eritrea’, Eritrean Studies Review,
3:2 (1999).
Connell, D. ‘Inside the EPLF: the origins of the “People’s Party” and its role in the
liberation of Eritrea’, Review of African Political Economy, 89 (2001).
Connell, D. ‘Eritrea: on a slow fuse’, in R.I.Rotberg (ed.), Battling Terrorism in the
Horn of Africa (Cambridge and Washington:World Peace Foundation / Brookings
Institution Press, 2005).
Connell, D. ‘The EPLF / PFDJ Experience: how it shapes Eritrea’s regional
strategy’, in R. J. Reid (ed.), Eritrea’s External Relations: Understanding its
Regional Role and Foreign Policy (London and Washington: Chatham House,
2009).
Crummey, D. ‘Tewodros as reformer and moderniser’, Journal of African History, 10:3
(1969).
Crummey, D. ‘The violence of Tewodros’, in B. A. Ogot (ed.), War and Society in
Africa (London: Frank Cass, 1972).
Crummey, D. ‘Society and ethnicity in the politics of Christian Ethiopia during the
Zamana Masfent’, International Journal of African Historical Studies, 8:2 (1975).
Crummey, D. ‘Orthodoxy and imperial reconstruction in Ethiopia, 1854–1878’,
Journal of Theological Studies, 29:2 (1978).
Crummey, D. ‘Abyssinian feudalism’, Past and Present, 89 (1980).
Crummey, D. ‘Banditry and resistance: noble and peasant in nineteenth-century
Ethiopia’, in D. Crummey (ed.), Banditry, Rebellion and Social Protest in Africa
(London: James Currey, 1986).
Crummey, D. (ed.). Banditry, Rebellion and Social Protest in Africa (London: James
Currey, 1986).
294 b i bl i og raphy

Crummey, D. ‘Imperial legitimacy and the creation of Neo-Solomonic ideology in

19th-century Ethiopia’, Cahiers d’Etudes Africaines, 28 (1988).
Crummey, D.‘Society, state and nationality in the recent historiography of Ethiopia’,
Journal of African History, 31:1 (1990).
Crummey, D. Land and Society in the Christian Kingdom of Ethiopia: From the Thirteenth
to the Twentieth Century (Oxford: James Currey, 2000).
Davidson, B., L. Cliffe and Bereket Habte Selassie (eds.). Behind the War in Eritrea
(Nottingham: Spokesman, 1980).
de Waal, A. (ed.). Islamism and its Enemies in the Horn of Africa (London: Hurst,
2004).
de Waal, A. ‘Darfur’s elusive peace’, in A. de Waal (ed.), War in Darfur and the Search
for Peace (London: Justice Africa/Global Equity Initiative, Harvard University,
2007).
Decalo, S. Coups and Army Rule in Africa: Motivations and Constraints (New Haven
and London:Yale University Press, 1990).
Dines, M. ‘The Ethiopian “Red Terror”’, in B. Davidson, L. Cliffe and Bereket
Habte Selassie (eds.), Behind the War in Eritrea (Nottingham: Spokesman, 1980).
Dines, M. ‘Ethiopian violation of human rights in Eritrea’, in L. Cliffe and
B. Davidson (eds.), The Long Struggle of Eritrea for Independence and Constructive
Peace (Trenton NJ: Red Sea Press, 1988).
Donham, D. and W. James (eds.). The Southern Marches of Imperial Ethiopia, 2nd edn.
(Oxford: James Currey, 2002).
Dunn, J. ‘“For God, Emperor, and Country!” The evolution of Ethiopia’s
nineteenth-century army’, War in History, 1 (1994).
Ehret, C. Ethiopians and East Africans: The Problem of Contacts (Nairobi: East African
Publishing House, 1974).
Ellingson, L. ‘The emergence of political parties in Eritrea, 1941–1950’, Journal of
African History, 18:2 (1977).
Erlich, H. ‘ “Tigrean nationalism”, British involvement, and Haile Selassie’s emerg-
ing absolutism: northern Ethiopia, 1941–1943’, Asian and African Studies, 15:2
(1981).
Erlich, H. ‘The Eritrean autonomy, 1952–1962: its failure and its contribution to
further escalation’, in Y. Dinstein (ed.), Models of Autonomy (New York:Transaction
Books, 1981).
Erlich, H. ‘The Ethiopian Army and the 1974 Revolution’, in M. Janowitz (ed.),
Armed Forces and Society (Chicago: University of Chicago Press, 1983).
Erlich, H. ‘Tigrean politics, 1930–35, and the approaching Italo-Ethiopian war’, in
H. Erlich, Ethiopia and the Challenge of Independence (Boulder: Lynne Reinner,
1986).
Erlich, H. Ras Alula and the Scramble for Africa: A Political Biography. Ethiopia and
Eritrea, 1875–1897 (Lawrenceville: Red Sea Press, 1996).
Ezekiel Gebissa (ed.). Contested Terrain: Essays on Oromo Studies, Ethiopianist Discourse,
and Politically Engaged Scholarship (Trenton: Red Sea Press, 2009).
 b i bl i og raphy 2 95

Fernyhough,T. ‘Social mobility and dissident elites in northern Ethiopia: the role of
banditry, 1900–1969’, in D. Crummey (ed.), Banditry, Rebellion and Social Protest in
Africa (London: James Currey, 1986).
Firebrace, J. [et al.]. Never Kneel Down: Drought, Development and Liberation in Eritrea
(Trenton: Red Sea Press, 1985).
Fukui, K. and J. Markakis (eds.). Ethnicity and Conflict in the Horn of Africa (London:
James Currey, 1994).
Gadaa Melbaa. Oromia: An Introduction (Khartoum: n.p., 1988).
Gaim Kibreab. People on the Edge in the Horn: Displacement, Land Use and the
Environment in the Gedaref Region, Sudan (Oxford: James Currey, 1996).
Gaim Kibreab. ‘Mass expulsion of Eritreans and Ethiopians of Eritrean origin from
Ethiopia and human rights violations’, Eritrean Studies Review, 3:2 (1999).
Gaim Kibreab. Critical Reflections on the Eritrean War of Independence (Trenton: Red
Sea Press, 2008).
Gaim Kibreab. Eritrea: A Dream Deferred (Woodbridge: James Currey, 2009).
Gaim Kibreab. ‘Eritrean–Sudanese relations in historical perspective’, in R. J. Reid
(ed.), Eritrea’s External Relations: Understanding its Regional Role and Foreign Policy
(London and Washington: Chatham House/Brookings Institution Press, 2009).
Garretson, P. ‘Vicious cycles: ivory, slaves and arms on the new Maji frontier’, in
D. Donham and W. James (eds.), The Southern Marches of Imperial Ethiopia, 2nd edn.
(Oxford: James Currey, 2002).
Gavshon, A. Crisis in Africa: Battleground of East and West (New York: Pelican, 1981).
Gebru Tareke. ‘Peasant resistance in Ethiopia: the case of Weyane’, Journal of African
History, 25:1 (1984).
Gebru Tareke. Ethiopia: Power and Protest. Peasant Revolts in the Twentieth Century
(Cambridge: Cambridge University Press, 1991).
Gebru Tareke. ‘The Ethiopia–Somalia war of 1977 revisited’, International Journal of
African Historical Studies, 33:3 (2000).
Gebru Tareke. The Ethiopian Revolution: War in the Horn of Africa (New Haven and
London:Yale University Press, 2009).
Gellner, E. Nations and Nationalism (Oxford: Blackwell, 1983).
Gemetchu Megerssa. ‘Oromumma: tradition, consciousness and identity’, in P. Baxter
et al. (eds.), Being and Becoming Oromo: Historical and Anthropological Enquiries
(Uppsala: Nordiska Afrikainstitutet, 1996).
Gerhard, D. ‘The frontier in comparative view’, Comparative Studies in Society and
History, 1:3 (1959).
Gewald, J.-B. ‘Making tribes: social engineering in the Western province of British-
administered Eritrea, 1941–1952’, Journal of Colonialism and Colonial History, 1:2
(2000).
Giglio, C.‘Article 17 of the Treaty of Uccialli’ (tr. R. Caulk), Journal of African History,
6:2 (1965).
Gilkes, P. ‘Centralism and the PMAC’, in I. M. Lewis (ed.), Nationalism and Self-
determination in the Horn of Africa (London: Ithaca Press, 1983).
296 b i bl i og raphy

Gilkes, P. ‘Eritrea: historiography and mythology’, African Affairs, 90:361 (1991).

Gilkes, P.‘Violence and identity along the Eritrean–Ethiopian border’, in D. Jacquin-
Berdal and M. Plaut (eds.), Unfinished Business: Ethiopia and Eritrea at War
(Lawrenceville: Red Sea Press, 2005).
Gilkes, P. and M. Plaut. War in the Horn: The Conflict Between Eritrea and Ethiopia
(London: Royal Institute of International Affairs, 1999).
Goody, J. Technology,Tradition and the State in Africa (London: Cambridge University
Press, 1971).
Gray, R. A History of the Southern Sudan 1839–1889 (London: Oxford University
Press, 1961).
Greenfield, R. Ethiopia: A New Political History (London: Pall Mall Press, 1965).
Hagmann,T. ‘Beyond clannishness and colonialism: understanding political disorder
in Ethiopia’s Somali region, 1991–2004’, Journal of Modern African Studies, 43:4
(2005).
Haile Larebo. The Building of an Empire: Italian Land Policy and Practice in Ethiopia, 2nd
edn. (Trenton: Red Sea Press, 2006).
Halliday, F. and M. Molyneux. The Ethiopian Revolution (London:Verso, 1981).
Harbeson, J. W. The Ethiopian Transformation: The Quest for the Post Imperial State
(Boulder and London: Westview Press, 1988).
Healy, S. and M. Plaut. Ethiopia and Eritrea: Allergic to Persuasion (London: Chatham
House, 2007).
Healy, S. Lost Opportunities in the Horn of Africa: How Conflicts Connect and Peace
Agreements Unravel (London: Chatham House, 2008).
Henze, P. Layers of Time: A History of Ethiopia (London: Hurst, 2000).
Hess, R. L.‘The Poor Man of God: Muhammad Abdullah Hassan’, in N. R. Bennett
(ed.), Leadership in Eastern Africa (Boston: Boston University, 1968).
Hoben, A. Land Tenure Among the Amhara:The Dynamics of Cognatic Descent (Chicago:
University of Chicago Press, 1973).
Hobsbawm, E. Bandits (London: Abacus, 2000).
Holcomb, B. K. and Sisai Ibbsa. The Invention of Ethiopia: The Making of Dependent
Colonial State in Northeast Africa (Trenton: Red Sea Press, 1990).
Holt, P. M. ‘Egypt and the Nile valley’, in J. E. Flint (ed.), Cambridge History of Africa,
Vol. 5: c.1790–c.1870 (Cambridge: Cambridge University Press, 1976).
Holt, P. M. and M. W. Daly. A History of the Sudan (Harlow: Longmans, 2000).
Howard, M. War in European History, 2nd edn. (Oxford: Oxford University Press,
2009).
Iliffe, J. The African Poor: A History (Cambridge: Cambridge University Press, 1987).
Institute for Security Studies Situation Report. The Eritrea–Djibouti Border Dispute
(15 September 2008).
International Crisis Group. Ethiopia and Eritrea: Preventing War,Africa Report No.101
(22 December 2005).
International Crisis Group. Ethiopia: Ethnic Federalism and Its Discontents, Africa
Report No.153 (4 September 2009).
 b i bl i og raphy 2 97

Jacquin-Berdal, D. and M. Plaut (eds.). Unfinished Business: Ethiopia and Eritrea at War
(Lawrenceville: Red Sea Press, 2005).
James, W. ’KWANIM PA: The Making of the Uduk People. An Ethnographic Study of
Survival in the Sudan–Ethiopian Borderlands (Oxford: Clarendon Press, 1979).
James, W. ‘War and “ethnic visibility”: the Uduk on the Sudan–Ethiopia border’, in
K. Fukui and J. Markakis (eds.), Ethnicity and Conflict in the Horn of Africa (London:
James Currey, 1994).
James, W. ‘Local centres on the Western Frontier, 1974–97: a case study of Kurmuk’,
in K. Fukui et al. (eds.), Ethiopia in Broader Perspective: Papers of the XIIIth
International Conference of Ethiopian Studies,Vol. II (Kyoto: Shokado Book Sellers,
1997).
James, W., D. Donham, E. Kurimoto, and A. Triulzi (eds.). Remapping Ethiopia:
Socialism and After (Oxford: James Currey, 2002).
James, W. ‘No place to hide: flag-waving on the western frontier’, in W. James et al.
(eds.), Remapping Ethiopia: Socialism and After (Oxford: James Currey, 2002).
Jhazbhay, I. ‘Islam and stability in Somaliland and the geo-politics of the war on
terror’, Journal of Muslim Minority Affairs, 28:2 (2008).
Johnson, D. ‘On the Nilotic frontier: imperial Ethiopia in the southern Sudan,
1898–1936’, in D. Donham and W. James (eds.), The Southern Marches of Imperial
Ethiopia, 2nd edn. (Oxford: James Currey, 2002).
Johnson, D. The Root Causes of Sudan’s Civil Wars (Oxford: James Currey, 2003).
Johnson, D. and D. M. Anderson (eds.). The Ecology of Survival: Case Studies from
Northeast African History (London: Lester Crook, 1988).
Jones, A. H. M. and E. Monroe. A History of Ethiopia (Oxford: Clarendon Press,
1935).
Jordan Gebre-Medhin. Peasants and Nationalism in Eritrea: A Critique of Ethiopian
Studies (Trenton: Red Sea Press, 1989).
Jordan Gebre-Medhin. ‘Eritrea (Mereb-Melash) and Yohannes IV of Abyssinia’,
Eritrean Studies Review, 3:2 (1999).
Kaptjeins, L. ‘Ethiopia and the Horn of Africa’, in N. Levtzion and R. L. Pouwels
(eds.), The History of Islam in Africa (Oxford: James Currey, 2000).
Keller, E. J. Revolutionary Ethiopia: From Empire to People’s Republic (Bloomington
and Indianapolis: Indiana University Press, 1988).
Kidane Mengisteab. ‘New approaches to state building in Africa: the case of
Ethiopia’s ethnic-based federalism’, African Studies Review, 40:3 (1997).
Kidane Mengisteab and Okbazghi Yohannes. Anatomy of an African Tragedy: Political,
Economic and Foreign Policy Crisis in Post-independence Eritrea (Trenton: Red Sea
Press, 2005).
Killion,T. ‘Eritrean workers’ organisation and early nationalist mobilisation: 1948–1958’,
Eritrean Studies Review, 2:1 (1997).
Kopytoff, I. ‘The internal African frontier: the making of African political culture’,
in I. Kopytoff (ed.), The African Frontier: The Reproduction of Traditional African
Societies (Bloomington and Indianapolis: Indiana University Press, 1987).
298 b i bl i og raphy

Kurimoto, E. ‘Fear and Anger: female versus male narratives among the Anywaa’, in
W.James et al. (eds.), Remapping Ethiopia: Socialism and After (Oxford: James Currey,
2002).
Kurimoto, E. and S. Simonse (eds.). Conflict, Age and Power in North East Africa: Age
Systems in Transition (Oxford: James Currey, 1998).
Lapidus, I. A History of Islamic Societies (Cambridge: Cambridge University Press, 2002).
Legum, C. and B. Lee. The Horn of Africa in Continuing Crisis (New York and London:
Africana Publishing Co., 1979).
Leonard, R. ‘Popular participation in liberation and revolution’, in L. Cliffe and
B. Davidson (eds.), The Long Struggle of Eritrea for Independence and Constructive
Peace (Trenton: Red Sea Press, 1988).
Leopold, M. Inside West Nile:Violence, History and Representation on an African Frontier
(Oxford: James Currey, 2005).
Levine, D. Greater Ethiopia: The Evolution of a Multi-ethnic Society, 2nd edn. (Chicago:
University of Chicago Press, 2000).
Lewis, H. S. ‘The development of Oromo political consciousness from 1958 to
1994’, in P. Baxter et al. (eds.), Being and Becoming Oromo: Historical and
Anthropological Enquiries (Uppsala: Nordiska Afrikainstitutet, 1996).
Lewis, H. S. Jimma Abba Jifar, an Oromo monarchy: Ethiopia 1830–1932, 2nd edn.
(Lawrenceville: Red Sea Press, 2001).
Lewis, I. M. (ed.). Nationalism and Self-Determination in the Horn of Africa (London:
Ithaca Press, 1983).
Lewis, I. M. A Modern History of the Somali, 4th edn. (Oxford: James Currey, 2002).
Locatelli, F. ‘Oziosi, vagabondi e pregiudicati: labour, law and crime in colonial
Asmara, 1890–1941’, International Journal of African Historical Studies, 40:2 (2007).
Locatelli, F. ‘Beyond the Campo Cintato: prostitutes, migrants and “criminals” in
colonial Asmara (Eritrea), 1890–1941’, in F. Locatelli and P. Nugent (eds.), African
Cities: Competing Claims on Urban Spaces (Leiden: Brill, 2009).
Love, R. Economic Drivers of Conflict and Cooperation in the Horn of Africa: A Regional
Perspective and Overview, Chatham House Briefing Paper (London, December
2009).
Low, D. A. ‘Uganda: the establishment of the Protectorate, 1894–1919’, in V. Harlow
and E. M. Chilver (eds.), History of East Africa, Vol. II (Oxford: Clarendon Press,
1965).
Low, D.A. ‘British East Africa: the establishment of British rule, 1895–1912’, in
V. Harlow and E. M. Chilver (eds.), History of East Africa,Vol. II (Oxford: Clarendon
Press, 1965).
Marchal, R. ‘Islamic political dynamics in the Somali civil war’, in A. de Waal (ed.),
Islamism and its Enemies in the Horn of Africa (London: Hurst, 2004).
Marcus, H. The Life and Times of Menelik II: Ethiopia, 1844–1913 (Oxford: Clarendon
Press, 1975).
Marcus, H. The Politics of Empire: Ethiopia, Great Britain and the United States, 1941–1974
(Berkeley: University of California Press, 1983).
 b i bl i og raphy 299

Marcus, H. A History of Ethiopia, 2nd edn. (Berkeley: University of California Press,

2002).
Markakis, J. National and Class Conflict in the Horn of Africa (Cambridge: Cambridge
University Press, 1987).
Markakis, J. ‘The nationalist revolution in Eritrea’, Journal of Modern African Studies,
26:1 (1988).
Markakis, J. ‘Ethnic conflict and the state in the Horn of Africa’, in K. Fukui and
J. Markakis (eds.), Ethnicity and Conflict in the Horn of Africa (London: James Currey,
1994).
Marsden, P. The Barefoot Emperor: An Ethiopian Tragedy (London: Harper, 2007).
Mburu, N. ‘Patriots or Bandits? Britain’s strategy for policing Eritrea, 1941–1952’,
Nordic Journal of African Studies, 9:2 (2000).
Mburu, N. Bandits on the Border: The Last Frontier in the Search for Somali Unity
(Trenton: Red Sea Press, 2005).
McCann, J. C. People of the Plow: An Agricultural History of Ethiopia, 1800–1990
(London: University of Wisconsin Press, 2005).
McGregor, A. Who’s Who in the Somali Insurgency: A Reference Guide (Washington,
DC: The Jamestown Foundation, September 2009).
Medhane Tadesse. The Eritrean-Ethiopian War: Retrospect and Prospects (Addis Ababa:
n.p., 1999).
Menkhaus, K.‘Somalia and Somaliland: terrorism, political Islam, and state collapse’,
in R. I. Rotberg (ed.), Battling Terrorism in the Horn of Africa (Cambridge/
Washington, DC: World Peace Foundation / Brookings Institution Press, 2005).
Menkhaus, K. ‘The crisis in Somalia: tragedy in five acts’, African Affairs, 106:424
(2007).
Menkhaus, K. ‘Somalia: a country in peril, a policy nightmare’, Enough! Strategy
Paper (September 2008).
Menkhaus, K. ‘Dangerous waters’, Survival, 51:1 (2009).
Menkhaus, K., J. Prendergast, and C. Thomas-Jensen. ‘Beyond piracy: next steps to
stabilise Somalia’, Enough! (May 2009).
Merera Gudina. ‘Contradictory interpretations of Ethiopian history: the need for a
new consensus’, in D. Turton (ed.), Ethnic Federalism: The Ethiopian Experience in
Comparative Perspective (Oxford: James Currey, 2006).
Merid Wolde Aregay. ‘A reappraisal of the impact of firearms in the history of war-
fare in Ethiopia, c.1500–1800’, Journal of Ethiopian Studies, 14 (1980).
Mesfin Araya. ‘Contemporary Ethiopia in the context of the battle of Adwa, 1896’,
in Paulos Milkias and Getachew Metaferia (eds.), The Battle of Adwa: Reflections
on Ethiopia’s Historic Victory Against European Colonialism (New York: Algora
Publishing, 2005).
Messay Kebede. Radicalism and Cultural Dislocation in Ethiopia, 1960–1974 (Rochester:
University of Rochester Press, 2008).
Middleton, R. Piracy in Somalia:Threatening Global Trade, Feeding Local Wars. Chatham
House Briefing Paper (London: Chatham House, October 2008).
30 0 b i bl i og raphy

Miran, J. Red Sea Citizens: cosmopolitan society and cultural change in Massawa
(Bloomington and Indianapolis: Indiana University Press, 2009).
Mockler, A. Haile Selassie’s War: The Italian–Ethiopian Campaign, 1935–1941 (London:
Random House, 1984).
Mohammed Hassan. The Oromo of Ethiopia: A History, 1570–1860 (Cambridge:
Cambridge University Press, 1990).
Mohammed Hassan. ‘The development of Oromo nationalism’, in P. Baxter et al.
(eds.), Being and Becoming Oromo: Historical and Anthropological Enquiries (Uppsala:
Nordiska Afrikainstitutet, 1996).
Mohammed Hassan. ‘Conquest, tyranny and ethnocide against the Oromo: a his-
torical assessment of human rights conditions in Ethiopia, ca.1880s–2002’, in
Ezekiel Gebissa (ed.), Contested Terrain: Essays on Oromo Studies, Ethiopianist
Discourse, and Politically Engaged Scholarship (Trenton: Red Sea Press, 2009).
Moore-Harell, A. ‘Economic and political aspects of the slave trade in Ethiopia and
the Sudan in second half of the nineteenth century’, International Journal of African
Historical Studies, 23:2–3 (1999).
Moyse-Bartlett, H. The King’s African Rifles: A Study in the Military History of East and
Central Africa, 1880–1945, 2 vols. (Aldershot: Gale and Polden, 1956).
Muller, T. ‘Human resource development and the state: higher education in post-
revolutionary Eritrea’, in D. O’Kane and T. Redeker Hepner (eds.), Biopolitics,
Militarism and Development: Eritrea in the Twenty-First Century (New York and
Oxford: Berghahn, 2009).
O’Kane, D. and T. Redeker Hepner (eds.). Biopolitics, Militarism and Development:
Eritrea in the Twenty-First Century (New York and Oxford: Berghahn, 2009).
Okbazghi Yohannes. Eritrea: A Pawn in World Politics (Gainesville: University of
Florida Press, 1991).
Orlowska, I. ‘Re-Imagining Empire: Ethiopian political culture under Yohannis IV
(1872–89)’, PhD thesis (London, 2006).
Paice, E. Tip and Run:The Untold Tragedy of the Great War in Africa (London: Phoenix,
2007).
Pankhurst, R. Economic History of Ethiopia, 1800–1935 (Addis Ababa: Haile Sellassie I
University, 1968).
Pankhurst, R. and D. Johnson. ‘The great drought and famine of 1888–92 in north-
east Africa’, in D. Johnson and D. M. Anderson (eds.), The Ecology of Survival: Case
Studies from Northeast African History (London: Lester Crook, 1988).
Patemam, R. ‘Drought, famine and development’, in L. Cliffe and B. Davidson
(eds.), The Long Struggle of Eritrea for Independence and Constructive Peace (Trenton,
NJ: Red Sea Press, 1988).
Pateman, R. Eritrea: Even the Stones are Burning, 2nd edn. (Lawrenceville: Red Sea
Press, 1998).
Paulos Milkias and Getachew Metaferia (eds.). The Battle of Adwa: Reflections on
Ethiopia’s Historic Victory Against European Colonialism (New York: Algora
Publishing, 2005).
 b i bl i og raphy 301

Pausewang, S., K. Tronvoll, and L. Aalen (eds.). Ethiopia Since the Derg: A Decade of
Democratic Pretensions and Performance (London: Zed Books, 2002).
Paz, R. ‘The Youth are Older: the Iraqization of the Somali Mujahidin Youth
Movement’, The Project for the Research of Islamist Movements: Occasional Papers, 6:2
(2008).
Phillipson, D. Ancient Ethiopia. Aksum: Its Antecedents and Successors (London: The
British Museum Press, 1998).
Pool, D. ‘Eritrean nationalism’, in I. M. Lewis (ed.), Nationalism and Self-determination
in the Horn of Africa (London: Ithaca Press, 1983).
Pool, D. ‘The Eritrean People’s Liberation Front’, in C. Clapham (ed.), African
Guerrillas (Oxford: James Currey, 1998).
Pool, D. From Guerrillas to Government:The Eritrean People’s Liberation Front (Oxford:
James Currey, 2001).
Prendergast, J. and M. Duffield. ‘Liberation politics in Ethiopia and Eritrea’, in T. M.
Ali and R. O. Matthews (eds.), Civil Wars in Africa: Roots and Resolution (Montreal
and Kingston: McGill-Queen’s University Press, 1999).
Prunier, G. Armed Movements in Sudan, Chad, CAR, Somalia, Eritrea and Ethiopia
(Berlin: Center for International Peace Operations, 2008).
Redeker Hepner,T. Soldiers, Martyrs,Traitors, and Exiles: Political Conflict in Eritrea and
the Diaspora (Philadelphia: University of Pennsylvania Press, 2009).
Redie Bereketeab. Eritrea: The Making of a Nation, 1890–1991 (Trenton: Red Sea
Press, 2007).
Reid, R. J. ‘A Fierce Race’, History Today, 50:6 (2000).
Reid, R. J. ‘The challenge of the past: the quest for historical legitimacy in inde-
pendent Eritrea’, History in Africa, 28 (2001).
Reid, R. J. ‘Warfare and urbanisation: the relationship between town and conflict in
pre-colonial eastern Africa’, in A. Burton (ed.), The Urban Experience in Eastern
Africa, c.1750–2000 (British Institute in Eastern Africa: Nairobi, 2002).
Reid, R. J. ‘Old problems in new conflicts: some observations on Eritrea and its
relations with Tigray, from liberation struggle to inter-state war’, Africa, 73:3
(2003).
Reid, R. J. ‘Caught in the headlights of history: Eritrea, the EPLF and the post-war
nation-state’, Journal of Modern African Studies, 43:3 (2005).
Reid, R. J. ‘ “Ethiopians believe in God, Sha’abiya believe in mountains”: the EPLF
and the 1998–2000 war in historical perspective’, in D. Jacquin-Berdal and
M. Plaut (eds.), Unfinished Business: Ethiopia and Eritrea at War (Lawrenceville:
Red Sea Press, 2005).
Reid, R. J. ‘War and remembrance: orality, literacy and conflict in the Horn of
Africa’, Journal of African Cultural Studies, 18:1 (2006).
Reid, R. J. War in Pre-colonial Eastern Africa: The Patterns and Meanings of State-level
Conflict in the Nineteenth Century (Oxford: James Currey, 2007).
Reid, R. J. ‘The Trans-Mereb experience: perceptions of the historical relationship
between Eritrea and Ethiopia’, Journal of Eastern African Studies, 1:2 (2007).
30 2 b i bl i og raphy

Reid, R. J. ‘Revisiting primitive war: perceptions of violence and race in history’,

War and Society, 26:2 (2007).
Reid, R. J. ‘The politics of silence: interpreting apparent stasis in contemporary
Eritrea’, Review of African Political Economy, 36:120 (2009).
Reid, R. J. (ed.). Eritrea’s External Relations: Understanding its Regional Role and Foreign
Policy (London: Chatham House, 2009).
Robertson, E. M. Mussolini as Empire-builder: Europe and Africa 1932–36 (London:
Macmillan, 1977).
Robinson, D. Muslim Societies in African History (Cambridge: Cambridge University
Press, 2004).
Robinson, R. and J. Gallagher. Africa and the Victorians (London: Macmillan, 1961).
Rotberg, R. I. (ed.). Battling Terrorism in the Horn of Africa (Cambridge / Washington,
DC: World Peace Foundation / Brookings Institution Press, 2005).
Rubenson, S. King of Kings: Tewodros of Ethiopia (Addis Ababa and Nairobi: Oxford
University Press, 1966).
Rubenson, S. ‘Adwa 1896: the resounding protest’, in R. Rotberg and A. Mazrui
(eds.), Protest and Power in Black Africa (New York: Oxford University Press,
1970).
Rubenson, S. ‘Ethiopia and the Horn’, in J. E. Flint (ed.), Cambridge History of Africa:
Vol. 5, c.1790–c.1870 (Cambridge: Cambridge University Press, 1976).
Rubenson, S. The Survival of Ethiopian Independence (London: Heinemann, 1976).
Ruth Iyob. The Eritrean Struggle for Independence: Domination, Resistance, Nationalism
1941–1993 (Cambridge: Cambridge University Press, 1995).
Ruth Iyob. ‘The Ethiopian–Eritrean conflict: diasporic vs. hegemonic states in the
Horn of Africa, 1991–2000’, Journal of Modern African Studies, 38:4 (2000).
Samatar, Said S. Oral Poetry and Somali Nationalism: The Case of Sayyid Mahammad
‘Abdille Hasan (Cambridge: Cambridge University Press, 1982).
Sanderson, G. N. ‘The Nile basin and the eastern Horn, 1870–1908’, in R. O. Oliver
and G. N. Sanderson (eds.), Cambridge History of Africa, Vol. 6: from 1870 to 1905
(Cambridge: Cambridge University Press, 1985).
Sbacchi, A. Ethiopia under Mussolini: Fascism and the Colonial Experience (London:
Zeb Books, 1985).
Schlee, G. Identities on the Move: Clanship and Pastoralism in Northern Kenya
(Manchester: Manchester University Press, 1989).
Schlee, G. ‘Gada systems on the meta-ethnic level: Gabbra / Boran / Garre interac-
tions in the Kenyan / Ethiopian borderland’, in E. Kurimoto and S. Simonse
(eds.), Conflict, Age and Power in North East Africa: Age Systems in Transition (Oxford:
James Currey, 1998).
Schlee, G. ‘Brothers of the Boran once again: on the fading popularity of certain
Somali identities in northern Kenya’, Journal of Eastern African Studies, 1:3
(2007).
Schmidt, P., M. Curtis, and Zelalem Teka (eds.). The Archaeology of Ancient Eritrea
(Trenton: Red Sea Press, 2008).
 b i bl i og raphy 303

Schwab, P. Ethiopia: Politics, Economics and Society (London: Frances Pinter, 1985).
Semere Haile. ‘Historical background to the Ethiopia–Eritrea conflict’, in
B. Davidson and L. Cliffe (eds.), The Long Struggle of Eritrea for Independence and
Constructive Peace (Trenton: Red Sea Press, 1988).
Sherman, R. Eritrea:The Unfinished Revolution (New York: Praeger, 1980).
Sorenson, J. Imagining Ethiopia: Struggles for History and Identity in the Horn of Africa
(New Brunswick: Rutgers University Press, 1993).
Styan, D. ‘Twisting Ethio-Eritrean economic ties: misperceptions of war and the
misplaced priorities of peace, 1997–2002’, in D. Jacquin-Berdal and M. Plaut
(eds.), Unfinished Business: Ethiopia and Eritrea at War (Lawrenceville: Red Sea
Press, 2005).
Taddesse Tamrat. Church and State in Ethiopia, 1270–1527 (Oxford: Clarendon Press,
1972).
Tekeste Negash. No Medicine for the Bite of a White Snake: Notes on Nationalism and
Resistance in Eritrea, 1890–1940 (Uppsala: University of Uppsala, 1986).
Tekeste Negash. Italian Colonialism in Eritrea, 1882–1941: Policies, Praxis and Impact
(Uppsala: University of Uppsala, 1987).
Tekeste Negash. Eritrea and Ethiopia: The Federal Experience (Uppsala: Nordiska
Afrikainstitutet, 1997).
Tekeste Negash and K.Tronvoll. Brothers at War: making sense of the Eritrean-Ethiopian
war (Oxford: James Currey, 2000).
Tekie Fessehatzion. ‘Eritrea and Ethiopia: from conflict to cooperation to conflict’,
Eritrean Studies Review, 3:2 (1999).
Tibebe Eshete. ‘The root causes of political problems in the Ogaden, 1942–1960’,
Northeast African Studies, 13:1 (1991).
Tibebe Eshete. ‘Towards a history of the incorporation of the Ogaden: 1887–1935’,
Journal of Ethiopian Studies, 27:2 (1994).
Treiber, M. ‘Trapped in adolescence: the post-war urban generation’, in D. O’Kane
and T. Redeker Hepner (eds.), Biopolitics, Militarism and Development: Eritrea in the
Twenty-First Century (New York and Oxford: Berghahn, 2009).
Triulzi, A. Salt, Gold and Legitimacy: Prelude to the History of a No-man’s-Land, Bela
Shangul,Wallagga, Ethiopia (ca.1800–1898) (Naples: Istituto Universitario Orientale,
1981).
Triulzi, A. ‘Italian colonialism and Ethiopia’, Journal of African History, 23:2 (1982).
Triulzi, A. ‘Competing views of national identity in Ethiopia’, in I. M. Lewis (ed.),
Nationalism and Self-determination in the Horn of Africa (London: Ithaca Press,
1983).
Triulzi, A. ‘The past as contested terrain: commemorating new sites of memory in
war-torn Ethiopia’, in P. Kaarsholm (ed.), Violence, Political Culture and Development
in Africa (Oxford: James Currey, 2006).
Tronvoll, K. ‘Ambiguous identities: the notion of war and “significant others”
among the Tigreans of Ethiopia’, in V. Broch-Due (ed.), Violence and Belonging:
The Quest for Identity in Post-colonial Africa (London: Routledge, 2005).
304 b i bl i og raphy

Tronvoll, K. ‘Human rights violations in federal Ethiopia: when ethnic identity is a

political stigma’, International Journal on Minority and Group Rights, 15 (2008).
Tronvoll, K. War and the Politics of Identity in Ethiopia: The Making of Enemies and
Allies in the Horn of Africa (Oxford: James Currey, 2009).
Turton, D. (ed.). Ethnic Federalism:The Ethiopian Experience in Comparative Perspective
(Oxford: James Currey, 2006).
Uoldelul Chelati Dirar. ‘Colonialism and the construction of national identities: the
case of Eritrea’, Journal of Eastern African Studies, 1:2 (2007).
Vikor, K. S. ‘Sufi brotherhoods in Africa’, in N. Levtzion and R. L. Pouwels (eds.),
The History of Islam in Africa (Oxford: James Currey, 2000).
Wolde-Yesus Ammar.‘The role of Asmara students in the Eritrean nationalist move-
ment, 1958–68’, Eritrean Studies Review, 2:1 (1997).
Wrong, M. I Didn’t Do It For You: How the World Betrayed a Small African Nation
(London and New York: Fourth Estate, 2005).
Yohannes Gebremedhin. The Challenges of a Society in Transition: Legal Development
in Eritrea (Trenton: Red Sea Press, 2004).
Yosief Libsekal. ‘Eritrea’, in International Council on Monuments and Sites (ICOMOS):
World Report 2001–02 on Monuments and Sites in Danger (Paris: ICOMOS, 2002).
Young, J. ‘The Tigray and Eritrean Peoples’ Liberation Fronts: a history of tensions
and pragmatism’, Journal of Modern African Studies, 34:1 (1996).
Young, J. ‘Ethnicity and power in Ethiopia’, Review of African Political Economy, 23:70
(1996).
Young, J. Peasant Revolution in Ethiopia:The Tigray People’s Liberation Front, 1975–1991
(Cambridge: Cambridge University Press, 1997).
Young, J. ‘Along Ethiopia’s western frontier: Gambella and Benishangul in transi-
tion’, Journal of Modern African Studies, 37:2 (1999).
Young, J. Armed Groups Along Sudan’s Eastern Frontier: An Overview and Analysis.
HSBA Working Paper 9 (Geneva: Small Arms Survey, 2007).
Young, W. C. ‘The Rashayida Arabs vs. the State: the impact of European colonial-
ism on a small-scale society in Sudan and Eritrea’, Journal of Colonialism and
Colonial History, 9:2 (2008).
Zewde Gabre-Selassie. Yohannes IV of Ethiopia: a political biography (Oxford: Clarendon
Press, 1975).
 Index

Abba Bagibo 46 tensions with other groups 98, 186,

Abba Jifar I 46 188, 239, 253
Abeba Aregai 145 Anuak 11, 138, 180, 253
Aberra, Dejazmach 105 Arsi 88
Acholi 137 ascari 97, 102, 108, 119, 122, 144, 158
Adal 28–9, 33 Asfeha Woldemichael, Dejazmach 155,
Adwa 49, 192 156
battle of (1896) 89, 90, 92, 98, 105, Aussa sultanate 135, 160–1
223–4, 248 Axum 24–5
Afar 5, 12, 29, 61, 71, 86, 89, 110, 135, Azebo 89, 143, 149
143, 160, 254–5
Agame 73, 189 Bahta Hagos 100, 105, 107
Agaw 11, 25 Bale 154, 165–6, 177
Ahmed ibn Ibrahim (‘Gran’) 29–30, banda 146, 149, 152
63–4 baria 11, 48, 81–2, 174
Akele Guzay 49, 71, 72, 73, 75, 76, 77, Begemeder 44, 134, 142, 145
105, 195 Beja 11–12, 179
Al-Beshir, Omar 181, 251 Beni Amer 11, 83, 104, 125, 156, 160,
Algiers Agreement 1 161, 196
Ali, Ras 45 Benishangul 11, 179, 180, 181,
All Amhara People’s Organisation 252–3
(AAPO) 213 Benishangul People’s Liberation
Alula, Ras 70, 75, 78, 83, 85, 96, 202 Movement (BPLM) 14, 252–3
Aman Andom 175, 197 Berta 11, 139, 180, 253
Amda Tsion 27–8 Britain 56, 67, 107–8, 112–13,
Amhara 3, 5, 11, 134, 136, 137, 138, 143, 146–8, 149, 150–1
216, 234 see also Eritrea, British Military
activism in 1990s 213, 214 Administration (BMA); Northern
defining features of 13–14 Frontier District, Kenya (NFD);
during zemene mesafint 42, 43 Somaliland, British
expansion under Menelik II 84–6, 87
political role in twentieth-century Christianity, Orthodox 13, 25–30, 52,
Ethiopia 106, 115, 117, 131, 133, 59–60, 62–5, 68–9, 119, 131, 154,
135, 146, 147, 154, 160, 210, 215, 161–2
224–5 Cushitic culture and people 11–12, 14
30 6 i nde x

Darfur 2, 232, 235, 246, 251, 252 Front (ELF); Eritrean Liberation
Dassanetch 136 Movement (ELM); Eritrean People’s
Derg 55 Liberation Front (EPLF)
internal wars 183–207 Eritrean-Ethiopian war (1998–2000) 1–2,
origins and organisation 173–7 194, 207, 216, 217–26, 238–9, 240,
and Somalia 181–3 246, 249
and Sudan 178–81 Eritrean Liberation Front (ELF) 126,
Djibouti 1, 99, 244 155, 158–9, 160–3, 168, 178, 190,
drought 10–11, 96, 111, 176–7 195–8, 203, 206
Eritrean Liberation Movement
economy 9, 21–2, 33–4, 211, 249 (ELM) 155, 158–9, 161, 168
colonial Eritrean 100–1, 102–5 Eritrean People’s Liberation Front
as driver of nationalist violence 115, (EPLF) 58, 162, 163–4, 175, 178,
122, 125–6, 126–7, 152, 156–7, 166 180, 185, 189, 190, 192, 241, 247
and Eritrean-Ethiopian war ethos and organisation of armed
(1998–2000) 219, 221 struggle 193–207
modern Eritrean 228–9 independent Eritrea 209–12, 226–35
and nineteenth-century war 43–4, 48, relations with TPLF 191, 193, 194,
53–4, 72, 85, 90, 96–7, 111, 126–7 200–5, 248; see also Eritrean-
overview of region 10–13 Ethiopian war (1998–2000)
and state-level violence under the Ethiopia
Derg 176–7, 187 claims to Eritrea 69–70, 74–8, 98–9,
and territorial expansion 42, 86–9, 106, 108, 115, 150–1, 214
133, 135 EPRDF rule 3, 181, 213–17, 236–40
see also drought; famine; trade Italian invasion and occupation 102,
Egypt 49, 51, 54, 64, 70, 73–83 135–6, 137, 139–46
Emeru Haile Selassie, Ras 145 see also Amhara; Derg; Eritrean-
Enarea 46, 61 Ethiopian war (1998–2000); Haile
Eritrea Selassie; Menelik II; Solomonic
British Military Administration ideology; Tewodros II; Yohannes IV;
(BMA) in 114–28 zemene mesafint
federated with Ethiopia 115, 126, Ethiopian Democratic Union
127–8, 154–9, 168–9 (EDU) 179, 190
as frontier zone in nineteenth- Ethiopian People’s Revolutionary
century 69–78, 78–84 Democratic Front (EPRDF) 3, 181,
growth of nationalism in 106, 107, 192, 206, 209, 210, 213–14, 217,
114–28, 154, 155–9 236, 239, 247, 253, 254
as independent state 3–4, 206, Ethiopian People’s Revolutionary Party
208–12, 226–35 (EPRP) 179, 180, 190
as Italian colony 99–108, 134 ethnic federalism, see Ethiopia, EPRDF
and liberation struggle 121, 123, 155, rule
158–9, 160–4, 168–9, 175–6, 180, ethnicity 11–14, 15–20, 40–1, 47–8
183–4, 193–207 armed liberation struggle 179–81, 183–8
and new historiography 15–20 in Ethiopia since 1990s 215–16; see
Somalia and 233, 241–2, 243–4, 251 also Ethiopia, under EPRDF rule
see also Eritrean-Ethiopian war see also under individual ethnic groups
(1998–2000); Eritrean Liberation and territories
 i nde x 307

famine 10–11, 85, 88, 96–7, 152, 166–7, Intergovernmental Authority on

173, 176–7, 199 Development (IGAD) 233, 241
federation, see Eritrea, federated with Isaias Afeworki 2, 195, 200, 206, 210,
Ethiopia 211, 214, 220, 222, 223, 224, 225,
frontier, role and concept of 4, 20–3, 58, 226–8, 229, 230, 232, 233, 245, 247
248–9, 250–1 Islam 28–30, 237, 248, 251–2
in Eritrea 105, 117–18, 126, 155–6,
Gambella People’s Liberation Front/ 157, 160–2, 195
Movement (GPLF/M) 14, 180, as external threat 73–4, 80, 232,
252–3 240–5, 251–2
Gash-Barka 79–84, 159 and Solomonic Ethiopia 59–65, 68–9,
geography 9–11 130–1, 151
Gojjam 13, 34, 42, 43, 44, 48, 50, 53, 56, and the Somali 109, 111–12, 217, 237,
68, 85, 86, 96, 134, 142, 145, 154, 240–5
166, 192 Ismail, Khedive 70, 75, 109
Gondar 13, 34–5, 40–1, 44, 48, 55, 192 Italian colonialism
‘Greater Ethiopia’ concept 14–15, 70 Eritrea 95–108, 140, 142
Gugsa, Ras 134, 135 Somalia 108–13, 140
Gurage 11, 86, 216 see also Ethiopia, Italian invasion and
occupation
Habte Giorgis, Fitaurari 131, 132, 134 Iyasu, Lij 64–5, 112, 129–31, 133, 134
Hadendowa 12, 79–80, 104, 125
Hadiya 11 Jimma Abba Jifar 46, 86, 87, 134–5, 146
Haile Mariam Redda 149
Haile Selassie 65, 106, 127, 174 Kaffa 11, 86, 88, 131, 133
authoritarianism of 151–2 Kambata 11
attempted coup d’etat against Kassa, Ras 135
(1960) 153, 167 Kebre Negast 26–8, 63
and crises of later reign 159–69 Konso 12, 174
early reign and Italian occupation 129–46 Kunama 11, 123, 125, 161
and Eritrean federation 154–9 Kwara 50, 52, 79
opposition to 152, 153
post-1941 regime of 146–51 Lasta 43, 67, 73, 96
Haile Selassie Gugsa 142 Leqa Nekemte 86, 87
Hailu Tewelde Medhin 73, 75
Hamasien 49, 71, 73, 74, 75–7, 78, 83, Mahdists (Sudan) 49, 64, 79, 83–4, 96
96, 123, 125 Meles Zenawi 3, 191, 192, 209, 210, 213–14,
Hamed Idris Awate 125, 126, 159 216, 217, 220, 225, 241, 245, 247
Harar 13, 46–7, 61, 88, 109, 131, 135, Menelik II 42, 49, 64–5, 68, 84–92,
143, 152, 177 96–9,105, 109–10, 129–30, 133, 236
Haud 109–10, 113, 151 Mengistu Haile Mariam 173, 174, 175, 176,
Hazzega 76–7 178, 179, 180, 181, 182, 184, 187, 192
historiography 14–20 Mereb River, see Eritrea; Tigray; Tigrinya
Mikael Sehul, Ras 35, 40–1, 51, 67
Ibrahim Sultan 118, 120, 157 Mikhail, Ras 129–31
Idris Muhammad Adam 156, 157 militarism and military culture
Ifat 28 and Eritrean shifta 122, 126–7
30 8 i nde x

militarism and military culture (cont.) and Ethiopia under EPRDF 215, 216,
and foundations of Solomonic 236–40
state 27–8 historiography of 15–20
and guerrilla insurgency 184–5, gada organisation among 32–3
201–2, 208–9, 247–8 and Gibe monarchies 46–7, 86–7
and modern Eritrea 3–4, 196–7, growth of nationalism among 154,
206–7, 229–35, 247–8 160, 164, 183, 185–7, 236–40
in nineteenth-century Ethiopia 43–7, as soldiers 43–4, 238–9
48, 56–8, 89–92 see also Oromo Liberation Front
in twentieth-century Ethiopia 132, (OLF); Oromo People’s Democratic
141, 153, 167–8, 174, 223–4, 247–8 Organisation (OPDO)
see also warfare, role and organisation Oromo Liberation Front (OLF) 3, 181,
of 187–8, 192, 213, 236–9, 241
Muhammad Abdille Hassan 111–12, 131 Oromo People’s Democratic
Muhammad Ali 79–81 Organisation (OPDO) 237
Muhammad Shafi, Shaikh 61–2
‘Patriots’ 145–6, 152
Nara 11, 125, 159, 161, 195 People’s Front for Democracy and Justice
nationalism 15–20, 183–5 (PFDJ) 212, 227
see also Eritrea; Oromo; Somali; Tigray proxy war, as regional phenomenon 1–4,
Nimeiri, Jaafar 178–9, 181, 251 246, 248–9, 255
Northern Frontier District, Kenya
(NFD) 110, 136–7, 164–5, 250 Rashaida 104
Nuer 138, 253 Raya 143, 149
religious conflict, see Christianity; Islam
Ogaden 61, 111–13, 130–1, 136, 140, Rendile 137
143, 246
insurgency in 3, 152, 164–5, 181–3, Saho 12, 71, 76, 77–8, 125, 193
241–2, 246 Samburu 137
Eritrea and 241–2 Sebagadis, Ras 66–7, 71, 72
Ethiopian claims to and occupation Semien 52, 56, 142
of 88, 109–10, 147, 148, 151 Semitic culture and people 5, 11, 13, 25
and Ethiopia-Somalia conflicts 164–5, Serae 71, 72, 73, 75, 77, 123
176, 181–3, 197 shangalla 11, 48, 81, 139
and Muhammad Abdille Hassan shifta 53, 58, 67, 77, 101, 102, 107,
111–13 136–9, 161, 163, 166, 179, 235
Ogaden National Liberation Front in Eritrea during 1940s and 1950s 114,
(ONLF) 3, 241–2 119, 120–7, 155, 158–9
Omotic peoples 12, 14, 33, 46 in Ethiopia during Italian
Oromo 1, 3, 5, 11, 110, 136, 137, 138, occupation 144–6
246, 249, 251, 253 general definition of 50–1, 120–1
‘Amharization’ of 12, 13, 186 see also Tigray
during zemene mesafint 40–1, 42, 48, ‘Shifta War’ (c.1963–8) 164–5
66–7 Shire 71–2, 192
early migrations of 30–5 Shoa 13, 26, 33–4, 42, 43–4, 48, 52, 53,
and Ethiopian empire 89, 98, 142, 68, 96, 130, 134, 141–3, 145, 152,
145, 146, 147, 152, 165 166, 169
 i nde x 30 9

Siad Barre 175, 181–2, 209, 241, 246 Taka 79–80, 82

Sidama 5, 11, 33, 253–4 Takle Haymanot 68, 85, 86
Sidama People’s Democratic Tedla Bairu 117, 120, 155, 156
Organisation (SPDO) 254 Teroa 125
slave trade 48–9, 138–9 Tewodros II 49–58, 60, 62–3, 70, 72–3,
slavery 133 77, 173, 174
Solomonic ideology 26–8, 40–1, 52, 58, Tigray 3, 35, 40–1, 42, 43, 52, 53, 56,
62, 63, 67–8 81–3, 96, 97, 98, 100, 130, 136, 143,
Somali 1, 2, 3, 5, 11, 12, 29, 86, 88, 109, 183, 213, 215, 216, 218, 239, 253
111, 131, 137, 142, 152, 216, 233, growth of nationalism in 66, 148, 154,
240–5, 249 185, 188–90
nationalism 110–11, 113, 152, 164–5, 181 links with colonial Eritrea 102, 104, 189
Somalia 2, 137, 175 marginalisation of 89–90, 188
collapse and civil war 183, 209, 233, nineteenth-century resurgence
235, 240–5, 246–7 of 66–9, 70–3
Eritrea and 233, 241–2, 243–4, 251 opposition to Haile Selassie in 134,
Ethiopian intervention in 135, 141–3, 146
(2006–9) 217, 240, 242–5 and shifta 123–5, 149–50, 152, 190
independence of 113 and Tigray-Tigrinya idea 117, 201,
wars with Ethiopia (1963–4) 154, 160, 215, 223, 225
164, 181; (1977–8) 175, 176, 181–3, Woyane revolt (1943) in 117, 148–50
197 Tigray People’s Liberation Front
Somaliland, British 109, 111–12 (TPLF) 58, 148, 150, 179, 180,
Somaliland, Italian 108–13 206–7, 241
Somaliland, Republic of 245 ethos and organisation of armed
South Sudan 2, 178–9, 251 struggle 188–92
Soviet Union 175, 176, 178, 181–2, 198 in power in Ethiopia 209–10, 214,
Sudan 2, 51, 54, 101, 103, 117, 198, 205, 215, 216–17, 234
232, 246, 249 relations with EPLF 191, 193, 194,
Egyptian expansion into 78–80 200–5, 248; see also Eritrean-
and Ethiopian borderlands 78–84, Ethiopian war (1998–2000)
138–9, 179–81, 249, 252–3 Tigre 5, 12, 160, 193
and frontier with colonial Tigrinya 5, 11, 12, 13, 43, 48, 100, 114,
Eritrea 104–5 115, 117, 125, 150, 156, 160, 169,
National Islamic Front in 181, 232, 188, 193, 195, 218
251, 252, 253 trade 28, 32, 41, 42, 46, 48–9, 60–1, 66,
nationalism in 158, 178 77–8, 79, 84–5, 103, 137–9, 219
relations with Eritrea 209, 212, 232, Tsazegga 76–7
235, 248, 251 Tsenadegle 125
relations with Ethiopia 178–81, Turkana 137
252–3
Sudan People’s Liberation Movement/ Uganda 137, 138
Army (SPLM/A) 179, 180, 248, Union of Islamic Courts (UIC) 2, 241,
251–2, 253 242, 243, 244
unionism, Eritrean 115, 116–27, 155, 169
Tafari, Ras, see Haile Selassie United States 127, 150–1, 174–5, 179,
Taitu 130 182, 192, 217, 229, 232–3, 240, 242
310 i nde x

warfare, role and organisation of Welde Mikael 75, 76, 77

during zemene mesafint 43–7, 48, 51–2 Welde Selassie, Ras 45–6, 66, 71, 72
in early history 27–8 Wellega 87, 135, 236
during Eritrean-Ethiopian war Western Somali Liberation Front
(1998–2000) 220–1 (WSLF) 181, 241
and Eritrean liberation struggle 160–6 Wichale, Treaty of 96, 98
and Ethiopian invasion of Somalia Woldeab Woldemariam 117, 120, 157
(2006–9) 242–3 Wollo 13, 25, 34–5, 41, 43, 52, 53,
historiography of 19–20 56, 61–2, 64, 129, 130, 135,
under Menelik 84–5, 86–9, 97–8 152, 166
and armed insurgency 124–5, 184–5, Woyane, see Tigray
187–8, 191–2, 197–200, 205–6 Wube, Ras 42, 52, 67, 72, 81–2
and slaving campaigns 133
under Tewodros 54–5, 56–8 Yemen 212, 232
see also Adwa, battle of; Eritrean- Yohannes IV 51, 56, 60, 63–4, 66–9, 70,
Ethiopian war (1998–2000); 73–84, 85, 148, 189, 202, 225
militarism and military culture;
shifta; Somalia, wars with Ethiopia; Zagwe 25–6
zemene mesafint zemene mesafint 3, 39–49, 50–2, 55,
Warrashek 41–2, 52 57–8, 67, 213
Welayta 11, 88 Zewditu 131, 132