AFRICANS IN PORTUGUESE COLONIAL ARMY

J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50

1 Paper presented at the Portuguese/African Encounters: An Interdisciplinary Congress, Brown
University, Providence MA, April 26-29, 2002. An earlier version focusing on the Mozambican case
was presented at the Second Congress of African Studies in the Iberian World, held in Madrid, Spain,
in 15-18 September 1999, and was published as João Paulo Borges Coelho, “Tropas negras na guerra
colonial: O caso de Moçambique,” in José Ramón Trujillo, ed., Africa hacia el siglo XXI (Madrid: Sial
Ediciones, Colección Casa de África 12, 2001).
African Troops in the Portuguese Colonial Army, 1961-1974:
Angola, Guinea-Bissau and Mozambique1
João Paulo Borges Coelho
Eduardo Mondlane University, Maputo, Mozambique
Abstract: The colonial powers systematically included Africans in the wars waged to
preserve their order. Portugal was not an exception in this respect. Since 1961, with the
beginning of the liberation wars in her colonies, Portugal incorporated Africans in her war
effort in Angola, Guinea-Bissau and Mozambique through a process enveloped in an
ideological discourse based on “multi-racialism” and on the preservation of the empire.
African engagement varied from marginal roles as servants and informers to more important
ones as highly operational combat units. By the end of the Portuguese colonial war, in 1974,
African participation had become crucial, representing about half of all operational colonial
troops. This paper explores in a comparative framework the three cases of Angola, Guinea-
Bissau and Mozambique, seeking the rationale behind the process and the shapes it took.
The abrupt end of the colonial war, triggered by a military coup in Portugal, paved the way
for the independence of the colonies, but left a legacy difficult to manage by the newly
independent countries. Shedding some light on the destiny of the former African
collaborators during this period, the paper suggests that they played a role in the postindependence
civil conflicts in Angola and Mozambique. © 2002 Portuguese Studies
Review. All rights reserved.
(...) if it isn’t to be a poor character with little utility, the European soldier
will cost us too much. It is therefore natural that to the African, more
adapted to the climate and much cheaper, the role will be reserved of
chair à canon (...)
Joaquim Mouzinho de Albuquerque, “A reorganização dos exércitos
ultramarinos,” Revista Militar 41 (7) (15 Apr. 1889)
Introduction
On 25 April 1974, a military coup in Lisbon paved the way to an abrupt end of
the long Portuguese colonial adventure and, more narrowly, of ten harsh years
during which colonial authority had been challenged by nationalist wars in
130 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
Angola, Guinea-Bissau and Mozambique. These three wars left profound marks
on the shape of the economies and societies of the three countries. One of these
marks was a legacy of thousands of Africans with a past of fighting side by side
with the Portuguese defence forces against independence. This study seeks to
discuss the historical rationale for such participation, as well as to shed some
light on the post-independence impacts it produced, which are still far from fully
understood.
It is almost a truism to say that the colonisation of the African continent
would have been impossible without local collaboration. The stereotyped picture
of immensely superior European forces defeating small, fragile and unarticulated
African resistances rarely corresponds to the historical truth. Much closer to
reality is the picture of European officials able to foster and manage internal
contradictions, attracting African forces into their orbit to make them fight other
African forces in order to install and preserve the colonial order.
In the two world wars of the last century, African troops fought in defence of
the colonial powers’ interests both in the African theatre and elsewhere.
Particularly after the 1950s, when nationalist movements began to fight for their
independence throughout the African continent, African participation in the
struggle to preserve the old colonial order acquired considerable importance.
Local collaboration was fundamental to guaranteeing the colonial project.
Portugal was not an exception in this respect, and often resorted to the
recruitment of Africans in her war effort, namely since the so-called
“Pacification Campaigns” in the late nineteenth century.
The present study looks at this collaboration in the context of the wars for
independence in the former Portuguese colonies—a collaboration that has to be
perceived at various levels, since its nature, importance and intensity varied
throughout the period in which the wars were fought. Wars tend evidently to
involve everybody within the area they cover. Here it will be useful, however, to
narrow the focus and look at the involvement which directly derived from the
colonial strategy—the “African involvement as strategy” or, as it was called in
those days, the Africanisation of the war effort.
This approach requires an historical perspective. Participation by African
troops was indeed uneven throughout the thirteen years of the Portuguese
colonial military campaigns. It began on the margins, limited to secondary roles
or, at the war fronts, to population control, intelligence gathering and
reconnaissance by informers and scouts. African troops became increasingly
important, however, and on the eve of the military coup of 25 April 1974
Africans accounted for more than 50 percent of the contingent fighting the war.
What follows is an attempt to discuss this Africanisation and the impact it had
on ending the war. We shall also examine briefly the profound traces the process
left as a heritage passed on to the newly independent African countries.
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 131
2 Malyn Newitt, A History of Mozambique (London: C. Hurst & Co., 1995), 21.
3 Valentim Alexandre and Jill Dias, coords., O Império Africano, 1825-1890, Vol. 10 of Nova
história da expansão Portuguesa, dir. by Joel Serrão and A.H. de Oliveira Marques (Lisbon: Editorial
Estampa, 1998), 140-1.

The Tradition of African Participation in the Colonial Army
Until the second half of the nineteenth century Portugal was far from controlling
the territories of what would become her colonies. Her sea supremacy, based on
innovative boat-and-cannon technology, certainly permitted a fairly resolute
coastal occupation.2 However, incursions into the hinterland were occasional,
carried out by individual entrepreneurs looking for stocks of food and smallscale
trade, or for women. Strategic expeditions such as the one lead by Paulo
Dias in 1575 to control the silver mines in Angola, or by Francisco Barreto to
control the gold mines of Mwenemutapa, were rare. Both men and the
environment in the interior were harshly hostile, and Portugal did not have a
structural reason to undertake such expeditions: Brazil was far more important
than Africa, the latter being of interest only in the sense of promising gold and
furnishing slaves to the New World. As a result, in the nineteenth century
Portugal could claim some presence and sovereignty only in coastal strips in
Angola and Mozambique, and in certain isolated posts connected through rivers
in Guinea.3 The vast hinterland remained only visited by traders.
In Angola, although traders within the Portuguese orbit (pombeiros, aviados,
feirantes, funantes, etc.) could generally develop their activity without serious
competition, the possibility of contestation by local African chiefs justified the
presence of military forces, whose primary role was, therefore, to keep open the
trade routes to the hinterland. During the nineteenth century, the military were,
according to their nature, roughly composed of three different forces. Firstly, the
“first-line” army consisting of full-time soldiers recruited locally, expeditionary
corps sent from Lisbon, and deportees to the overseas territories. Secondly, the
“second-line” forces, mainly local part-time volunteers, including many
“civilised” mulattoes and Africans, who joined mobile companies headed by
loyal African chiefs, traders, or farmers with honorary military ranks. Their role
was to complement the activity of the army in times of war, or to assure
administrative tasks such as tax-collection or porter control. Finally, when times
were troubled, the authorities could convoke loyal African chiefs, who would
come with their private armies to form what was usually known as the Guerra
Preta (black war). This third contingent, parts of which were known as
empacaceiros (or hunters of a wild buffalo called pacaça), could reach several
thousand. The advantages these fighters brought to the wars waged by the
Portuguese were obvious, from the fact that no expenses were involved (the war
132 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
4 René Pélissier, História das campanhas de Angola. Resistência e revoltas, 1845-1941, 2 vols.
(Lisbon: Editorial Estampa, 1986), 1: 36-40. The scheme was also developed in other Portuguese
colonies. In Timor, for example, second line “Companies of Dwellers” were formed on the basis of
traditional militias (Moradores and Arraiais), with the mission of controlling terrestrial and maritime
borders, policing and protecting public buildings. By 1974, 53 of these companies were operating. See
Eurico António Sales Grade, “Timor: O corpo militar de segunda linha,” Revista Militar 26 (4-5) (1974):
205.
5 Valentim Alexandre, Velho Brasil novas Áfricas. Portugal e o Império (1808-1975) (Porto: Edições
Afrontamento, 2000), 233-4.
6 Pélissier, Campanhas, 1: 35 and passim.
7 In what had became a tradition already in those times. For a vivid example, see Alexandre Lobato,
Os Austríacos em Lourenço Marques (Maputo: Arquivo Histórico de Moçambique, 2000), 294 and
passim, who describes the sending of a naval force from Goa, through Mozambique Island, to surprise
and expel the Austrian trade post in Delagoa Bay, in 1781.
8 Allen Isaacman, A tradição de resistência em Moçambique. O vale do Zambeze, 1850-1921 (Porto:
Edições Afrontamento, 1979), 81.
booty being their reward), to avoiding regular troops’ casualties and
humiliation.4
Things started to change during the nineteenth century, in particular as Brazil
ceased to import slaves and as Portugal, in domestic political terms, entered the
new and more stable phase of “Regeneration”, which offered greater
opportunities to look after the colonial territories.5 This was a time when, in
Mozambique, a genuine transition was already under way, based on a rather
sharp decline of the mercantile dynamic, gradually replaced by a new one that
would be based on the management of the labour force internally and for export,
as well as on establishing administrative structures down to the bottom level. In
Angola, where mercantile exchanges with the hinterland preserved some vigour
(alcohol, weapons, slaves, cloth, ivory), this process would only take place
somewhat later.6
In any case, these transformations demanded territorial control and the
breaking of local resistance in order to assure Portuguese authority. In view of
the serious problems Portugal had to face from the very beginnings of her
presence in Africa, particularly the long distances from Lisbon and the shortage
of men, conquest was achieved to a large extent through the use of local forces,
complemented by the dispatch of expeditionary troops from Lisbon and India7
whenever the situation was deemed serious.
Even so, the conquest undertaken by the expeditionary forces would not have
been possible without the vital participation of local contingents, either recruited
directly on an ad hoc basis, or through political agreements with African
potentates. According to Isaacman,8 during the critical “pacification” campaign
in the Zambezi valley in the last quarter of the ninteenth century, more than 90
percent of the soldiers on the colonial side were Africans.
Such contingents were also reinforced by units recruited in other colonial
territories, as the Portuguese avoided having troops fight in their home areas and
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 133
9 António Enes, Moçambique. Relatório apresentado ao Governo (Lisboa: Agência Geral do
Ultramar, 1946 (originallly published 1893), 115.
10 See E. A. Azambuja Martins, O soldado Africano de Moçambique (Lisboa: Agência Geral das
Colónias, 1936), 34, for a table of all Mozambican units dispatched to Angola, Guinea, India, Macau
and Timor between 1904 and 1932. See also Pedro Pezarat Correia, “A participação local no
desenvolvimento das campanhas,” in Estudos sobre as Campanhas de África, 1961-1974 (Estoril: Ed.
Atena, 2000), 144.
11 Azambuja Martins, Soldado Africano, 9-10.
12 Azambuja Martins, Soldado Africano, 12.
13 John P. Cann, Contra-insurreição em África, 1961-1974. O modo português de fazer a guerra.
Estoril: Edições Atena, 1998), 132. The German experience with locally recruited forces, particularly
the Ascaris, was considered in Mozambique as highly positive. In 1915, the German force in Tanganyika
was composed of 2,200 European troops, 11,100 regular African troops and 3,200 irregulars, in 24
companies, each one roughly with a dozen Europeans and 300 Africans. See Mário Costa, “É o inimigo
que fala. Subsídios inéditos para o estudo da Campanha da África Oriental, 1914-1918,” Lourenço
Marques (1928), Arquivo Histórico Militar, 2ª Divisão, 7ª Secção, Caixa 151, Nº 6.
understood the advantages of moving locally recruited forces from one colonial
territory to another, in the struggle to impose their order. Enes mentions Indian
soldiers in Mozambique,9 but most significant was the participation of Angolans
in the Zambezi valley campaign, as well as, a little later, of “Landim” units from
Mozambique in Angola and Guinea. Units from Mozambique and Angola also
went to serve in distant colonies such as Goa, Macau and Timor.10
Initially, the recruitment of local forces was unsystematic and arbitrary,
mirroring other forced labour requisitions for public works or plantation
undertakings. Azambuja Martins describes raids in Angola to capture “idlers”,
who were tied to a rope and taken to military quarters, a procedure that made the
victims of this recruitment method widely and sarcastically known as “the
volunteers of the rope”.11 However, as Portuguese authority and administration
was extended to the bottom level, local recruitment underwent noticeable
transformations and, in particular, was facilitated by the population census. This
evolution is clearly evident, for example, in the Mozambican legislation on the
subject. The decree of 14 November 1901, in its article 61, prescribed that every
colony should have its own recruitment regulations. In 1904, a regulation
stipulated that recruitment should be undertaken with the local régulo as
intermediary, who would receive a gratification for each man recruited. In 1906
a more aggressive recruitment scheme based on the use of agents was
implemented, using, as Azambuja Martins eloquently writes, “the same methods
used for recruiting labour for the mines in Transvaal.”12 In the face of the
decreasing resistance of locals to military service, the 1914 regulation prescribed
the creation of a military reserve, which permitted the engagement of 25,000
Africans, or 44 percent of the total force, in the struggle against the German
invasion of Northern Mozambique during the First World War.13
From this point on, service in the armed forces began to be perceived as an
important way of “nationalising” the African population of the colonies. The
regulation of June 1933 stipulated distinct service branches for “common
134 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
14 Cited in Azambuja Martins, Soldado Africano, 17. For more on the civilising role of the army, “as
a beacon that spreads the light of our civilizing action,” see Mário C. Moutinho, O indígena no
pensamento colonial Português, 1895-1961 (Lisboa: Edições Universitárias Lusófonas, 2000), 108-119.
15 See Secretaria Geral da Defesa Nacional, Pº 324.10, 18 de Novembro de 1961, Arquivo da
Secretaria Geral de Defesa Nacional (further SGDN) (Forte de S. Julião da Barra), Caixa 6690.5.
16 See Comando Militar de Moçambique, Quartel General, nº 53546, Lourenço Marques, 4 de
Dezembro de 1962, to Chefe do Estado-Maior das Forças Armadas, “Proposta da R.M.M. sobre
incorporações de recrutas indígenas. Passagem à disponibilidade das Praças I que terminaram o período
de obrigação normal de serviço”, Arquivo da SGDN (Forte de S. Julião da Barra), Caixa 6690.5.
blacks” and for “non indigenous blacks,” the latter being enrolled in the same
service branches as Europeans born in the colonies. Africans were to be
registered in the ranks under a Portuguese Christian name, and it was expected
that military service would act as a powerful “civilising” mechanism or, in the
words of General Norton de Matos, as “one of the most effective mechanisms
for opening a breach in the tenebrous primitive civilisations.”14
Since the mid-1930s, the program of building the colonial forces seems to
have suffered from tensions between a vision in which Europeans born in the
colonies would lead as officers an increasingly African army, and the practical
need to promote Africans to lead local units, owing to the limited numbers of
Europeans. On a broader level, the tension was between maintaining separate
armies in Portugal and her colonies within the framework of an organisational
schema that dated back to 1893, or to unify the armies in a single multicontinental
Portuguese army. The doubts would persist until the eve of the
independence wars.
Africans in the Colonial Defence Forces at the War of Independence
After the Second World War the Portuguese army, as part of the entire colonial
system, began to be forced to change as a result of international pressures and as
the nationalist wars were anticipated and approaching. The Estado Novo was
forced to repeal the Estatuto dos Indígenas which had assured for so long that
the vast majority of the population remained without access to the status of
citizens, and this required the army to deal with a new and unexpected problem:
that of having a growing African contingent in its ranks. Until then the infantry
recognized the categories of commissioned soldiers (white soldiers born in
Portugal or in her overseas provinces), overseas soldiers (African assimilados),
and native soldiers (Africans under the indigenato regime). Forced to change this
system, the regime, through the decree 43.267 of 24 October 1960, introduced
the new categories of 1st, 2nd and 3rd class soldier, in practice corresponding to
the previous ones.15 A little later, this too had to be changed, and although the
colour of the soldiers’ skin ceased to be a criterion, two classes were established
on the basis of formal education and, in particular, of the ability to speak
Portuguese correctly. In practical terms this again meant a perpetuation of the
old distinctions.16 Even so, however, the door was opening ever so slightly for
the Africans.
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 135
17 Subsecretário de Estado da Aeronáutica para Presidente do Conselho de Ministros, 9 de Agosto
de 1960, Arquivo Nacional da Torre do Tombo, AOS/CO/PC-78j, pt3ª, 20ªsd. D.30 (I owe this
information to Amélia Souto).
18 Hélio Felgas, “Necessidade recíproca da ligação metrópole-províncias ultramarinas,” Revista
Militar 13 (7) (1961): 421.
19 See Chefe do Estado-Maior do Exército, nº 332.19, “Incorporação de recrutas indígenas. Passagem
à situação de disponibilidade das praças I que terminaram o período de obrigação normal de serviço,”
Lisboa, 7 de Fevereiro de 1962, Arquivo da SGDN (Forte de S. Julião da Barra), Caixa 6690.5. For the
image of the Africans as terrorists or foreigners, see Moutinho, O indígena, 51.
20 For the development of this argument, see Pezarat Correia, “Participação local,” 145 and passim.
Faced with the threat represented by the Africanisation of its army, a certain
segment of the regime’s establishment had every motive to resist. Firstly,
ideological reasons played a role—paradigmatic in this respect is the strong
position taken by Kaúlza de Arriaga, who in 1960, as Subsecretary of State for
Aeronautics, wrote to Salazar that “a defence concept based on black troops is
impossible, independently of the kind of white control. (…) It is therefore
necessary to reduce the strength and size of our black troops.”17 This view,
undoubtedly common among the upper echelons, is very well conveyed by
Felgas, who commenting on the consequences of a weak metropolitan
contribution to the army wrote that “it is not difficult to foresee that disorder
would prevail, as well as insecurity, as happened in the first months of
independence of the ex-Belgian Congo. Of course, we could adopt the same
solution for Angola, to promote soldiers to colonels and corporals to generals.
But the results would be identical: an army deprived of strength, cohesion,
prestige and discipline.”18
Associated with this view was one that, despite all the integrationist
propaganda of the Nation in Arms, considered the Africans as little less than
potential terrorists, in the international context of the Cold War. This led the
Chief-of-Staff to write, as late as 1962, that engaging the native masses in a
military effort posed great risks, since they were all heavily exposed to the
“propaganda of the enemy.”19
As a result, and despite the strong effort to prepare the army for the African
campaigns, undertaken from the late 1950s onward, the old “philosophy” that
had informed the defence and security system in the colonial territories remained
basically unchanged, with “expeditionary forces” from Portugal coming to fight
the colonial wars and the locally recruited ones playing a limited and secondary
role as second-line troops.20
That this was indeed so is clear from the constant increase in metropolitan
contingents as the wars started, first in Angola in March 1961, and then in
Guinea in January 1963 and in Mozambique in June 1964. Table 1 documents a
100% increase in the total metropolitan contingent during the first half of the
period of conflict (until 1967). Figures in Table 2 show a correspondent modest
increase in the numbers of locally recruited forces during the same period (from
18% in 1961 to 25% in 1967).
136 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
21 See also Quadro A in A. Afonso and C. Matos Gomes, Guerra colonial (Lisbon: Editorial
Table 1. Metropolitan Troops In the African War Theatres
Date Angola Mozambique Guinea Total
1961 28,477 8,209 3,736 40,422
1962 33,760 8,852 4,070 46,682
1963 34,530 9,243 8,336 52,109
1964 37,418 10,132 12,874 60,424
1965 41,625 13,155 14,640 69,420
1966 38,519 19,550 18,868 76,937
1967 43,051 23,164 18,421 84,636
1968 37,547 22,717 19,559 78,823
1969 36,911 23,286 22,866 83,063
1970 36,174 22,633 22,487 81,294
1971 36,127 21,795 23,402 81,324
1972 34,676 22,657 24,036 81,369
1973 37,773 23,891 25,610 87,274
Source: Estado-Maior do Exército, 1988 I: 260.
Table 2
Locally Recruited Troops In the African War Theatres (With % of Total Troops)
Date Angola Mozambique Guinea Total
1961 5,000 (14.9) 3,000 (26.8) 1,000 (21.1) 9,000 (18.2)
1962 11,165 (24.9) 3,000 (25.3) 1,000 (19.7) 15,165 (24.5)
1963 12,870 (27.2) 5,003 (35.1) 1,314 (13.6) 19,187 (26.9)
1964 15,075 (28.7) 7,917 (43.9) 2,321 (15.3) 25,313 (29.5)
1965 15,448 (27.1) 9,701 (42.4) 2,612 (15.1) 27,761 (28.5)
1966 17,297 (31.0) 11,038 (36.1) 1,933 (09.3) 30,268 (28.2)
1967 14,369 (25.0) 11,557 (33.3) 3,229 (14.9) 29,155 (25.6)
1968 20,683 (35.5) 13,898 (38.0) 3,280 (14.4) 37,861 (32.7)
1969 18,663 (33.6) 15,810 (40.4) 3,715 (14.4) 38,188 (31.4)
1970 19,059 (34.5) 16,079 (41.5) 4,268 (16.0) 39,406 (32.6)
1971 25,933 (41.8) 22,710 (51.0) 5,808 (19.9) 54,451 (40.1)
1972 25,461 (42.2) 24,066 (51.5) 5,921 (19.8) 55,448 (40.5)
1973 27,819 (42.4) 27,572 (53.6) 6,425 (20.1) 61,816 (41.4)
Source: Estado-Maior do Exército, 1988 I: 261.
Table 1 also reveals more modest increases in metropolitan troops during the
second half of the war,21 while Table 2 shows correspondingly a much more
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 137
Notícias., 2000), 518.
22 Thomas Henriksen, “Portugal in Africa: Comparative Notes on Counterinsurgency,” Orbis, 21 (2)
(1977): 404.
23 Estado-Maior do Exército, Resenha histórico-militar das Campanhas de África (1961-1974), 4
vols. (Lisboa: Estado-Maior do Exército, 1988-89) 1: 259 and passim; Cann, Contra-insurreição, 122
and passim.
24 The number of officers graduating from the military academy rose sharply from 68 in 1962 to 146
in 1967, but from then on started to suffer an abrupt decline to only 40 in 1973, as a result of lack of
volunteers. In consequence, the military authorities were forced to mobilise conscripts to fill the huge
gaps in the professional cadres. See David Martelo, “Pessoal e orçamentos. Esforço de guerra,” in
Afonso and Gomes, Guerra colonial, 519-20.
25 For the discussion of Portugal’s financial problems with the war, which fall outside the ambit of
this article, see also, as an example among several, Ministro da Defesa Nacional to Ministro das
pronounced increase in locally recruited troops, whose number reached nearly
half of the total contingent in 1973. Around 1968 there clearly occurred a certain
break in the balance between metropolitan and local forces, with more marked
increases in the latter. This raises the question why, despite all the entrenched
resistance discussed above, did the Estado Novo and its military apparatus
change their attitude so dramatically.
Obviously, the fact that this change occurred when Marcello Caetano replaced
Salazar as President of the Cabinet was not purely coincidental. However, there
are many other factors that also help to explain it.
The main argument for this Africanisation of the Portuguese colonial army
has been based on Portugal’s recruiting problems. In the late 1960s, according to
Henriksen,22 Portugal had, after Israel, the highest percentage of people in arms
in the world, with an annual increase of 11% between the 49,422 documented in
1961, and the 149,090 documented in 1973. In parallel, the percentage of
deserters also doubled, from 11.6 to 20.9, for reasons linked both with avoidance
of military service and with the fact that Portugal was a chronic provider of
migrant labour to Europe and the Americas, through a process that gradually
drained a population of potential recruits already small from the start.23 To these
quantitative difficulties, qualitative ones also have to be added, in the sense that
the expeditionary contingents had serious problems of adaptation to the African
war theatres and that, as the years passed, the Portuguese army faced an acute
shortage of commanding officers, with obvious consequences for its military
efficiency.24
Besides the issue of sheer human numbers, the financial difficulties that
Portugal experienced in coping with the three wars also shed some explanation
on Africanisation. According to this argument, the burden became so unbearable
that the progressive sharing of the war effort with the colonies, through
increasing local recruitment and financial participation, was a way to minimise
the weight. Moreover, this would seem to be quite well in line with the old
Salazar principle of involving each colony in the resolution of its own
problems.25
138 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
Finanças, nº 833/AF/73, Lisboa, 3 de Junho de 1973, in SGDN (Forte de S. Julião da Barra), Caixa
768.2.
26 According to Ferraz de Freitas, Conquista da adesão das populações (Lourenço Marques: SCCIM,
1965), 6, ordering is based exclusively on physical power and provokes the repulsion of culturally
different populations, and for this reason its efficiency tends to decrease in proportion to the decline of
the physical power of the one who exerts it. On the contrary, commanding requires knowledge and ability
to handle the “social forces,” is based on participation, and promotes adhesion of the commanded. As to
accionamento, it was defined as “the set of moves one needs to take to make sure that the population works
with us and becomes prejudiced towards the propaganda of the enemy (...). [Through accionamento] we
attract the populations into our orbit, integrate them in our environment, in our culture, in our civilisation
and nationality (...). This would be one of our purposes. The other is to make them work actively with us
in detecting and combating subversion (...)” (GDT/Serviços Distritais de Administração Civil, 1966:45,
Arquivo Histórico de Moçambique, Secção Especial, nº 237). For the psycho-social work with local
communities in central Mozambique, namely in organising popular operations conducted by traditional
authorities to detect guerrilla movements, see João Paulo Borges Coelho, “A ‘Primeira Frente’ de Tete e o
Notwithstanding these valid explanations, the shortage of metropolitan men
and the high costs of war were not the only reasons behind the Africanisation
process. Despite such difficulties and the fact that they were almost
insurmountable, Portugal was indeed capable of sustaining some level of
increase in the numbers of her metropolitan troops, and the costs involved were
basically covered, even if at the price of going to the brink of economic and
financial exhaustion.
Further factors have then to be brought in to explain the process of
Africanisation. The first, which reveals one of the several internal contradictions
of the Estado Novo, is of a historical and ideological nature. It was based on the
appeal of the integrationist ideology of the Empire and its principle of race
miscegenation, which translated into the revival of white settlement plans and
into the “promotion” of the African populations, particularly under the short but
decisive mandate of Adriano Moreira as overseas minister. In a sense, it bore
elements of continuity with the early days of Salazar’s regime, when important
steps were made to include an African layer at the foundation of the colonial
state and administration, with the involvement of local African authorities in
population censuses, tax collection, and labour recruitment in their areas of
jurisdiction. As mentioned above, this “attitude” also had led to the
establishment of defence forces based on the inclusion of local troops in
secondary roles. The overwhelming majority of these local troops were Africans,
who served mostly as auxiliaries, servants in the barracks, and, given their
knowledge of the terrain, as informers and scouts.
However, the nationalist wars also introduced a new phenomenon in the sense
that a much broader African involvement became required, well beyond such
targeted recruitment: early on in the process the colonial authorities understood
that the war was also about conquering the population. The “philosophy” of the
Estado Novo, mixed with the first counter-insurgency techniques to “win” the
population, provided the core of a colonial psycho-social doctrine based on the
two fundamental concepts of comandamento (command) and accionamento
(driving, setting in motion).26 These served as a framework for the creation, in
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 139
Malawi,” Arquivo (15) (1994): 72.
27 In 1961, Adriano Moreira, as Overseas Minister, issued legislative diplomas, which created the
Militia Corps as second-line forces in the African colonies. See Guilherme de Sousa Belchior Vieira,
“A auto-defesa das populações,” Revista Militar, nº 2-3 (1962): 215. For an apology of these local
militias, see for instance Jaime de Oliveira Leandro, “As acções contra-revolucionárias e a sua técnica,”
Revista Militar 15 (1) (1963): 65.
28 José Freire Antunes, A Guerra de África (1961-1974), 2 vols. (Lisboa: Círculo de Leitores, 1995)
1: 408-9.
29 João Paulo Borges Coelho, “Protected Villages and Communal Villages in the Mozambican
Province of Tete (1968-1982): A History of State Resettlement Policies, Development and War” (Ph.D.
Dissertation: Department of Social and Economic Studies, University of Bradford, 1993), 165 and
passim.
30 General Carlos Fabião, the “father” of these new militias, would say with respect to their creation:
“I studied, I read ancient documents about the pacification campaigns of the end of last century [19th],
and concluded that past wars in Africa had been fought more or less with the locals, particularly in
Guinea”. See Antunes, Guerra de África, 1: 366.
the early 1960s, of local militias of several kinds in the rural areas of the three
colonies, as second-line troops under the authority of the civil administration and
based on the principle of self-defence against subversive attacks.27 The militia
took a wide range of organisational shapes, from very informal schemes of
village forces acting under the authority of village chiefs, as happened in
Angola,28 to more institutionally militarised groups assuring the defence of the
aldeamentos, the protected villages formed along the lines of what the British
had practiced in their counter-insurgency war in Malaysia, or the North-
Americans in Vietnam.29
As the war situation aggravated, and with the corresponding difficulties
experienced by the armed forces, the militia contingents were brought into more
active roles exceeding the traditional defensive ones. Paradigmatic in this respect
was the case of Guinea, where, besides the normal militias, special ones were
created for offensive operations in their home areas. This new concept, which
led to the emergence of very efficient troops, mixed the old colonial tradition30
with new counter-insurgency theories. These theories, to which the Portuguese
high military commanders were systematically exposed from the second half of
the 1950s onward, had as one of their core concepts the so-called “same element
theory,” according to which the guerrillas could be fought more efficiently by
troops mirroring their organisation, weaponry, knowledge of the terrain, and
even race—i.e., by African combat units. These theories became more
extensively absorbed at a time when the need to adapt and to reinforce the
Portuguese troops became more pressing.
The changes that started to occur at this point brought a new meaning to the
concept of Africanisation. From now on this would not imply merely a growing
percentage of locally recruited or black individuals incorporated in the regular
forces fighting the nationalists, in the same sense as the French jeunissement in
Indochina, for example. More than that, it now meant a process of creating and
fostering combat units of Africans operating more or less irregularly and
140 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
31 See, for instance, Chefe do Estado-Maior do Exército, nº 8392, “Recrutamento de Praças de 3ª
(Indígenas)”, Arquivo da SGDN (Forte de S. Julião da Barra), Caixa 6690.5.
32 See Comando-Chefe das Forças Armadas de Angola, n/ref, n/date, “Plano de formação de novos
GE no ano de 1973”, Arquivo da SGDN (Forte de S. Julião da Barra), Caixa 768, nº 6F. “The cost of
their formation and maintenance is much lower than the one of regular troops. Their incorporation brings
much less socio-economic problems and they are much cheaper to transport.”
33 See this argument in Pezarat Correia, “Participação local,” 147. The period when Marcello
Caetano was in power roughly corresponds to an important decentralization in the conduct of the wars.
If formerly these were conducted in a relatively centralized manner from Lisbon, now three generals
with strong views were appointed to run them from the provinces. Spínola took charge of Guinea as
Governor and Commander-in-Chief in May 1968; Kaúlza de Arriaga went to Mozambique as
Commander-in-Chief in March 1970; and Costa Gomes was appointed as Commander-in-Chief of
Angola in April 1970. In July 1969 the military operations in each theatre were centralised under each
commander-in-chief.
autonomously, and with high levels of operational efficiency. It must be said that
this change in attitude, which implied increased trust in the Africans, even if
forced, was not sudden, nor was it just due to theoretical considerations. In fact,
in the second half of the 1960s the military were facing serious problems related
not only to a shortage of troops but also to the question of what to do with the
Africans demobilised from the regular army. Old suspicions that Africans
demobilised from the regular force, already capable of handling weaponry,
might simply join the nationalist guerrillas, lay behind new efforts to create
auxiliary troops where those men would be kept under the control of military or
civil authorities.31 Moreover, these troops were much cheaper than the regular
ones and their eventual casulaties were much “less repercussion-rich” than those
of metropolitan forces.32
This Africanisation was carried out differently in the three territories of
Angola, Guinea and Mozambique, not just because of different local and
regional contexts, but also owing to the different attitudes and views of the
respective commanders.33 The results likewise varied: while in Guinea General
Spínola planned and fought for the creation of a regular, entirely coherent
African army mirroring the metropolitan one, having perhaps in mind a future
federation of Portuguese-speaking states, in Angola the African irregular units
were much more informal and diverse in nature. Mozambique somehow
combined aspects of the two.
The first unit entirely consisting of Africans was probably the Tropas
Especiais (Special Troops, commonly known by the acronym TEs), which
emerged in 1966 in Cabinda, when Alexandre Taty, a former UPA/FNLA cadre,
deserted to the Portuguese side with 1,200 men. Organised and controlled by
PIDE, the Portuguese political police, they started to be used in action against
the MPLA, which after having initiated operations in the Dembos, in Northern
Angola, had spread its guerrilla activities to Cabinda in late 1964. The TEs
operated in Cabinda, their home area, as well as in Zaire and Uíge, in Northern
Angola.
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 141
34 For Oscar Cardoso’s testimony on the Flechas, see Antunes, Guerra de África, 1: 401. He
mentioned, in particular, that “they didn’t need logistical support. Gatherers since kids, they could live
out of nothing, with special hability to find food and water. We really had good operational results with
them. We never had a desertion from the Flechas’ ranks.”
35 Antunes, Guerra de África, 2: 705.
36 See in this regard Secretaria Geral de Defesa Nacional, nº 1142/GU, 27 de Dezembro de 1967,
Arquivo da Secretaria Geral de Defesa Nacional (Forte de S. Julião da Barra), Caixa 6114, 5C.
Infiltration of SWAPO and MPLA guerrillas to the South worried very much the South-Africans, who
were ready to meet the costs of Portuguese reinforcements at battalion level in the area. However, the
Portuguese could do little more than reinforce the area with some extra GE groups. See Secretaria Geral
de Defesa Nacional, nº 588/69, 28 de Fevereiro de 1969, and Secretaria Geral de Defesa Nacional, nº
437/RA, 22 de Agosto de 1969, Arquivo da Secretaria Geral de Defesa Nacional (Forte de S. Julião da
Barra), Caixa 6114, nº 5A.
Also in 1966, the war epicentre in Angola moved to the East, with the start of
UNITA activities in that area and with the opening of MPLA’s Eastern front,
which became a threat to the Moxico and Cuando Cubango areas in the
Southeast. Besides dispatching a TE unit to the area, PIDE also began to create
what would become its private ethnic army. The first experiment was conducted
by inspector Oscar Cardoso, who worked in Cuando Cubango with bushmen
groups serving as scouts and information gatherers, and soon as true combat
units, exploiting the cultural distance between these small men of San origin and
the Bantu populations of the remaining areas.34 The experiment was so
successful in operational terms that the concept soon spread to other areas,
particularly Luso (Luena) and Luanda-Caxito, where the Flechas unit was
almost entirely composed of ex-MPLA guerrillas. Towards the end of the war
almost all PIDE sub-delegations in Angola’s war zones had their own private
units of Flechas.
A little later, in 1968, the military also favoured the creation of their own
irregular troops, the Grupos Especiais (or GEs, Special Groups), formed of local
volunteers who were submitted to the same training as the regular military
forces. Organised in combat groups 31 men strong, they were controlled by the
military, usually one Portuguese battalion having one or two of these groups
nearby. Beside the fact that they were cheaper than the regular troops, one of the
advantages gained from the creation of this force, which operated in the North,
East and South of Angola, was their knowledge of local languages, culture, and
terrain.35 The sharp increase in the number of GE groups also had to do,
however, with problems in furnishing replacements for regular troops that had
concluded their operational commission.36
Also in the Eastern area of Angola, groups of Katangese gendarmes formerly
supporting Moses Tchombé crossed the border into Angola during the second
semester of 1967, and were received as political refugees by the Portuguese
authorities. In February 1969 they formed the Front for National Liberation of
Congo, with the objective of overthrowing the Mobutu regime. At this time,
facing an acute shortage of forces in the Eastern part of the territory, the
142 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
37 Comando-Chefe das Forças Armadas de Angola (Quartel-General), Gabinete das Forças
Auxiliares, “Memorando: Acção Fidelidade”, 2 de Agosto de 1974, Arquivo Histórico Militar, 2ª
Divisão, 2ª Seccção, Caixa 139, Nº 7.
38 See, for instance, Zona Militar Leste, Oficial de Ligação do Comando-Chefe das Forças Armadas
de Angola, Luso, 2 de Março de 1971, “Operação Fidelidade: Relatório da visita de inspecção ao campo
de refugiados de Veríssimo Sarmento (Camissomba)”, Arquivo Histórico Militar, 2ª Divisão, 2ª Seccção,
Caixa 139, Nº 16.
39 Both Rhodesians and South-Africans expressed, in 1972, their reservations towards Operation Colt
mounted by the Portuguese, on the grounds that they did not believe the Zambian ANC had enough
credibility to destabilise Zambia at the time. See several documents in Secretaria Geral de Defesa
Nacional (Forte de São Julião da Barra), box 5681, particularly Estado Maior de Angola, 1ª Rep, nº
2490/RA, pº 324.210 AO, 15 de Dezembro de 1972 (I thank Amélia Souto for this source).
40 Comando-Chefe das Forças Armadas de Angola, Gabinete das Forças Auxiliares, n/r, Luanda, 3
de Agosto de 1964, Arquivo Histórico Militar, 2ª Divisão, 2ª Seccção, Caixa 139, nº 8.
41 Interviewed in 2001, Costa Gomes said that “the reason why I got along with PIDE was the
following: When I arrived in Angola already as Commander-in-Chief, I heard from the Department of
Operations of a conflict which had occurred involving PIDE’s Flechas and a Commando company of
Portuguese authorities launched the project “Fidelidade” (Fidelity), based on the
promise to support the “liberation” of Zaire in exchange for the participation of
this force in counter-insurgency operations in Angola, particularly against the
MPLA. Code-named Fiéis, they received military supplies and training from the
Portuguese armed forces, as well as political supervision from the PIDE.37 This
became one of the most effective forces in counter-insurgency operations,
despite chronic problems with discipline that translated into frequent riots and
desertions and that stemmed from complaints about low pay and from the fact
that these men did not feel they were receiving enough support from the
Portuguese government in their struggle against Mobutu’s Zaire.38
At the same time and a little further to the South, a similar project, codenamed
“Operation Colt,” was implemented to receive a smaller group of
Zambian ANC dissidents who arrived in Angola in 1967.39 In 1968, PIDE
organised them into a combat group of 45 elements code-named Leais, under the
same kind of agreement as the one established with the Fiéis, namely to fight
against the Angolan liberation movements in exchange for support in their
struggle to overthrow Kenneth Kaunda’s Zambian regime.40
The wide, flexible, and diverse utilisation of African irregular troops that was
practiced in Angola did not have any parallel in the other war theatres. It was the
result of several specific factors. The first one was, of course, the vast area to be
covered in counter-insurgency activities, together with the great difficulties the
Portuguese had in replacing, let alone increasing, the number of their troops, and
together with the inevitable financial aspects. The second factor was the
competition among the three nationalist movements, which gave rise to
desertions by trained guerrillas who went to join counter-insurgency operations.
Thirdly and very importantly, it is necessary to take into account the attitude of
General Costa Gomes, who managed to establish a good relationship between
the military and PIDE,41 one that undoubtedly allowed the spread of this kind of
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 143
the army and resulting in some casualties. I asked myself how could PIDE be waging a war without the
Commander-in-Chief knowing about it? I asked the Governor to convene a meeting with the people that
mattered and told them: PIDE can do whatever wars but not on behalf of the Governor or PIDE itself.
It has to be on my behalf”. See Público (Lisbon), 12 (4153): 14 (2 de Agosto de 2001).
42 Still according to Costa Gomes, interviewed by the Washington Post in March 1971, “the African
troops combating against subversion are increasing in number. Their training and experience rendered
them outstanding professionals.” And in another interview with Época magazine, on 21 December 1971:
“TEs are a very cohesive group (…) GEs are growing from year to year. (…) As do the Flechas. All the
irregular troops have proven their merit. That is why we will make efforts to increase their numbers.”
See Arquivo Histórico Militar, 2º Divisão, 2ª Secção, Caixa 194, nº 10.
43 Estado-Maior do Exército, nº 2592, 7 de Setembro de 1966, and Secretaria Geral de Defesa
Nacional, nº 27/66, 18 de Janeiro de 1966, Arquivo da Secretaria Geral de Defesa Nacional (Forte de
S. Julião da Barra), Caixa 6879, nº 3D and 3F respectively.
irregular war, so that unlike many of his fellow generals he was able to shift a
signficant operational burden onto the autonomous activity of the African
units.42 Table 3 illustrates the high level of autonomous activity undertaken by
the irregular forces.
Finally, besides the already discussed advantages inherent in the employment
of foreign irregular troops, the colonial authorities in Angola also had in mind
interfering at a broader regional level, keeping Zaire and Zambia in particular
under pressure.
Table 3
Irregular Forces Operations in Angola in 1972. Number (With % of Total)
Force Jointly with the
Army
Autonomously Total
Flechas 128 (29) 316 (071) 444 (100)
GE 922 (31) 2,010 (069) 2,932 (100)
TE — 342 (100) 342 (100)
Fiéis 369 (24) 1,170 (076) 1,539 (100)
Leais — 46 (100) 46 (100)
Source: Comando-Chefe das Forças Armadas de Angola, Gabinete das Forças Auxiliares,
Resumo da Actividade Operacional (1º e 2º Semestres de 1972), in Arquivo Histórico
Militar, 2ª Divisão, 2ª Secção, Caixa 139, nº 9.
In Guinea developments were quite different in this regard. When the war
started, locally recruited militia groups were created to assure the “self-defence”
of the populations, freeing the expeditionary army for offensive operations. By
1966, Guinea already had 18 militia companies, and the authorities were
requesting funds to create more, although still acknowledging the risks involved
in having to deal with a “considerable volume of people armed, equipped and
trained”.43 The substitution of General Schulz, a quite conventional and
144 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
44 Antunes, Guerra de África, 1: 363 and passim; Cann, Contra-insurreição, 137,
45 See the exchange of correspondence between Spínola and the Secretaria Geral de Defesa
Nacional, particularly Secretaria Geral de Defesa Nacional, nº 1929/RA, 13 de Abril de 1969; Comando-
Chefe das Forças Armadas da Guiné, nº 4626/69, 10 de Outubro de 1969; Comando-Chefe das Forças
Armadas da Guiné, nº 15746/AP/01, 22 de Dezembro de 1969; Secretaria Geral de Defesa Nacional, nº
1291/RA, 3 de Abril de 1970; Comando-Chefe das Forças Armadas da Guiné, nº 2268/AP/01, 11 de
Julho de 1970, Comando-Chefe das Forças Armadas da Guiné, nº 2673/AP/01, 10 de Agosto de 1970;
and Comando-Chefe das Forças Armadas da Guiné, n/ref., Abril de 1971, in Arquivo da Secretaria Geral
de Defesa Nacional (Forte de S. Julião da Barra), Caixa 718, nº 1C.
46 Secretaria Geral de Defesa Nacional, Info nº 409/AF/73, 20 de Julho de 1973, Arquivo da
Secretaria Geral de Defesa Nacional (Forte de S. Julião da Barra), Caixa 435, nº 1B. See also Secretaria
Geral de Defesa Nacional, nº 550/RA, 13 de Outubro de 1970, Arquivo da Secretaria Geral de Defesa
Nacional (Forte de S. Julião da Barra), Caixa , nº 6812, nº 1B, Secretaria Geral de Defesa Nacional, nº
551/RA, 13 de Outubro de 1970, Arquivo da Secretaria Geral de Defesa Nacional (Forte de S. Julião da
Barra), Caixa nº 6812, nº 1A, and Secretaria Geral de Defesa Nacional, nº 279/RA, 5 de Maio de 1970,
Arquivo da Secretaria Geral de Defesa Nacional (Forte de S. Julião da Barra), Caixa nº 907, nº 6C.
47 SAF, nº 372/AF/72, “Encargos com forces especiais ou irregulars”, Arquivo da Secretaria Geral
de Defesa Nacional (Forte de S. Julião da Barra), Caixa nº 760, nº 4.
48 The Centres for Commando Instruction (CICs) were created in Guinea in 1964, Angola in 1965,
and Mozambique in 1969, for training these elite troops. Many of the Guinea African commando
combatants were trained in Angola.
conservative commander, by General António Spínola brought profound
changes. Through selective recruitment among normal militias, he fostered the
creation of Special Militias organised in combat groups and operating fairly
autonomously.44 Spínola structured these militias along the structural lines of the
Portuguese army, in companies subdivided into platoons. He faced a tough battle
with the upper echelons to end the distinction between metropolitan and locally
recruited soldiers, arguing that discrimination against the latter involved serious
risks for the Portuguese African campaign, and threatening that its perpetuation
“would force us to redefine our counter-insurgency policy based on African
forces.”45
This conflict continued when Spinola pressed to increase the number of
special militias. Of the five requested in 1968 only two had been authorized by
1970. Lisbon’s resistance clearly had to do with financial constraints, and also
with fears that “the informality brought on by the Africanisation of the war is
spreading to an informality of procedures.”46 It was perhaps because of this, and
in an attempt to counteract the tendency of provincial commanders to keep
creating new African forces that, at the end of 1971, the Minister of Defence
ordered the centralisation in Lisbon of all expenses relating to the irregular
troops.47 In 1970, Spínola again engaged in a struggle with the upper echelons in
order to create, using elite African combatants from the militias and the
organizational pattern of the Portuguese army, companies of African
Commandos, structured very much like normal commando companies created in
the other theatres,48 but manned entirely by Africans and carrying out very
special combat operations both within Guinea and in Guinea-Conakry and
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 145
49 Afonso and Gomes, Guerra colonial, 204-4; Antunes, Guerra de África, 2: 698, 719; SAF, nº
391/AF/71, 25 de Fevereiro de 1971, Arquivo da Secretaria Geral da Defesa Nacional (Forte de S. Julião
da Barra), Caixa nº 907, nº 6B; Comando Territorial Independente da Guiné, CO 45-246 A/G, 21 de
Fevereiro de 1973, Arquivo da Secretaria Geral de Defesa Nacional (Forte de S. Julião da Barra), Caixa
nº 733, nº 3A.
50 He translated himself the seminal book of Sir Robert Thompson, Defeating Communist
Insurgency.
51 See in this respect several reports in Arquivo da Secretaria Geral de Defesa Nacional (Forte de S.
Julião da Barra), particularly boxes nº 772.1, 5294, 5298, 6102.1, and 3036.2.
52 In Mozambique, the first four commando companies were created in Montepuez in 1969, and
experienced limited growth, despite their relative operational success, to five in 1970, and eight by the
end of the war in 1974.
53 Including, among other, the creation of paramilitary groups linked to PIDE, acting in the North
as hunters of wild game and, in fact, gathering intelligence on the nationalist guerrillas’ movements and
contacts.
54 Estado Maior do Exército 1989 (4), 1988: 193. For an important testimony by Captain Van Uden,
one of the GEs’ trainers, see Rui Rodrigues, ed., Os últimos guerreiros do Império (Lisbon: Erasmus,
1995), 242 and passim.
Senegal. The discussion of the project took two years and only in 1972 did they
start to operate.49
In Mozambique, this process occurred a little later and in a way that combined
features of the other two cases, as well as some that were specific to the area.
The first stages of the war unfolded under the command of General Augusto dos
Santos, an admirer of the new counter-insurgency theories.50 In the mid-1960s,
with Costa Gomes as second-in-command, Dos Santos sponsored important
“experiments” involving local militias. One of these led to successful
collaboration with the Rhodesian authorities to form units of African scouts.51
However, in 1969 Kaúlza de Arriaga replaced General Dos Santos. The new
Commander-in-Chief had an entirely different way of conducting the war. For a
long time the operational involvement of Africans was limited to local
recruitment in the regular army or in commando companies,52 irregular African
units being entirely out of the picture, with the exception of limited paramilitary
experiments.53 Only in 1973 were the first solutions involving African units
implemented on the ground, with the creation of Special Groups of Parachutists
and other Special Groups, formed in the central regions of the country on an
ethnic basis, recruiting volunteers, and operating in particular in their home
areas, in coordination with and under the control of the military.54
In contrast with the Angolan case, Kaúlza’s reservations with regard to
African forces operating outside the military’s sphere of control, as well as his
conflicts with PIDE, probably help to explain why the implementation of a
Flecha-type project took so long in Mozambique. The introduction of Flechas
was being discussed between PIDE and the Rhodesian authorities since 1972.
The latter were very much interested in the unfolding events in Mozambique and
favoured “lighter” and more local alternatives to the way the war was being
conducted by Kaúlza de Arriaga. It is therefore probable that discrete and low
146 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
55 See, in this respect, the testimony of Lieutenant-Colonel João Santos Fernandes in Rui de Azevedo
Teixeira, ed., A Guerra Colonial: Realidade e ficção (Lisbon: Ed. Notícias, 2001), 63 and passim.
Fernandes claims that the project had the support of the Mozambique provincial government and it was
the central government that prevented its implementation in 1972.
56 As late as July 1973, the army refused to supply automatic weapons to the Flechas, on grounds
that all the lots were already consigned. PIDE approached the South-Africans who expressed a
willingness to finance the supply. See Estado-Maior do Exército, nº 2522/LM, “Prioridade no
fornecimento de material à delegação da DGS em Lourenço Marques com destino aos ‘Flechas’,”
Arquivo da Secretaria Geral de Defesa Nacional (Forte de S. Julião da Barra), Caixa 435, nº 1C, and
Estado-Maior General das Forças Armadas, nº 2333/RC, “Fornecimento de espingardas G-3 para os
‘Flechas’ de Moçambique contra pagamento da RAS,” Arquivo da Secretaria Geral de Defesa Nacional
(Forte de S. Julião da Barra), Caixa 795, nº 8A.
57 The numbers were gathered from several sources, including direct reports in Arquivo Histórico
Militar, 2ª Divisão, 2ª Secção, Caixa 139, nº 5; and Arquivo da Secretaria Geral de Defesa Nacional
(Forte de S. Julião da Barra), Caixa 763, nº 1C. They vary considerably, due to several factors including
different estimation criteria.
profile “experiments” with Flechas under the aegis of PIDE were in fact taking
place since 1972,55 despite the difficulties.56 Only in 1974, when Kaúlza de
Arriaga had been already dismissed by Marcello Caetano, and on the eve of the
military coup that ended the war, did they start to operate on the ground.
Table 457
African Special Forces
Year of
Creation Theatre Designation Origin Organisation
Nº of
Men in
1974
1966 Angola TEs Ex-guerrillas Groups (17) 800
1967 Angola Flechas
Ethnic
troops/
Ex-guerrillas
2,270
1968 Angola GEs Locally
Recruited Groups (99) 3,250
1968 Angola Leais Foreign
troops 127
1969 Angola Fiéis Foreign
troops Companies (16) 2,400
1970 Guinea Special
Marines Local troops Detachments (2) 160
1970 Mozambique GEs Local troops Groups (12) 860
1972 Guinea African
Commandos Local troops Battalions (3) 700
1973 Mozambique GEs Ethnic
troops Groups (83) 2,720
1974 Mozambique Flechas Ethnic
troops ?
The comparative analysis of how Africans were used as soldiers by the
Portuguese at the three fronts of the colonial wars shows that, beyond very broad
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 147
58 Chefe do Estado-Maior General das Forças Armadas, “Memorando: Africanização das Forças
Armadas nos TO Ultramarinos”, Secretaria Geral da Defesa Nacional, nº 12131/GC, 1 de Maio de 1973,
Arquivo da Secretaria Geral de Defesa Nacional (Forte de S. Julião da Barra), Caixa 1634.1.
59 The debate about the Portuguese decolonisation is very much on, with some arguing that
everything was done that was possible under the circumstances, while others claim that it was conducted
on the basis of abandonment and betrayal.
strategic guidelines, the factors that mattered were local context and local
commanders. Costa Gomes, perhaps the most successful general, sought good
relations with the civilians and employed African units within the framework of
a counter-insurgency technique. Spínola, by contrast, appealed for a more
political and psycho-social use of African soldiers. Kaúlza, the most
conservative of the three, feared African forces outside his strict control and
seems not to have progressed beyond his initial racist perception of the Africans
as inferior beings and terrorists.
Whatever the approaches and their degree of success in terms of furthering
colonial interests, the fact is that on the eve of the war’s end Africanisation had
been accepted as the only way of maintaining the colonial project, in circles as
high as that of the Portuguese Chief-of-Staff. According to him, African troops
were more efficient, more cost-effective, cheaper and susceptible of delivering
better results not only in military terms, but also politically. Moreover. if
properly organised in militarised villages they could fight forever.58 In 1974 he
therefore proposed a substantial reduction in the number of metropolitan troops
and a decentralisation of the financial resources thus spared, so that the local
commands could create further African units. Clearly, the plan was to promote
preconditions for civil war in the African territories, if not politically, at least
militarily.
A Note on the Colonial African Soldiers After Independence
Owing to several factors, the decolonisation process that followed the military
coup of 25 April 1974 was far from smooth and planned. In Guinea and
Mozambique, where the nationalists appeared strong and unified, Portugal’s
attempts to manage the process through referendums or political negotiations
were cut short by threats that war would resume unless independence and the
transfer of power were prioritized as immediate objectives. In Angola, by
contrast, the liberation movement was divided, not only amongst themselves but
also internally, particularly the MPLA. Moreover, profound contradictions were
manifest at the level of the Portuguese decision-making circles, with the
President of the Republic António Spinola trying to revive attempts at federation
with the colonies, while senior government officials and the military had very
different plans. As a result, decisions were very much informed by political
events unfolding in Portugal, and were dominated by improvisation and the need
for quick response.59 This context became particularly dramatic in the
demobilisation of the African forces formerly serving on the Portuguese side in
148 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
60 Josep Sánchez Cervelló, “La revolución portuguesa y la questión colonial. Que descolonisatión?”,
in Instituto de História Contemporânea da Universidade Nova de Lisboa , ed., Portugal e a transição
para a democracia (1974-1976), I Curso Livre de História Contemporânea, Lisboa, 23-28 de Novembro
de 1998 (Lisbon: Edições Colibri, 1999), 89.
61 Presidência da República Portuguesa, “Relatório” (in Centro de Planeamento e Coordenação da
Direcção do Serviço Histórico Militar do Exército Português, nº 1438/88/C, 22 de Junho de 1988),
Arquivo Histórico Militar, 2ª Divisão, 2ª Secção, Caixa 169, nº 17.
62 Comando-Chefe das Forças Armadas de Angola, Gabinete das Forças Auxiliares, “Memorando:
Acção Fidelidade”, 7 de Agosto de 1974, Arquivo Histórico Militar, 2ª Divisão, 2ª Secção, Caixa 139,
nº 7.
the colonies. Enveloped in substantially different scenarios in the past, as we
have argued, Angola, Guinea-Bissau and Mozambique thus experienced
developments that were also quite divergent in this particular regard.
Research still needs to be done regarding the destiny of the irregular African
forces in Angola. Several indications, however, point to the probability of a high
level of integration. Firstly, the fact that there were three competing nationalist
movements and that war resumed after the 1974 coup certainly favoured the
integration of these highly skilled soldiers among contingents already fighting a
new war, the civil one, which lasted almost until the present day. Secondly, the
1975 South-African invasion of Angola pressed the newly independent regime,
which had a weak army, to recruit in great haste people capable of fighting in
this new context.
FLEC re-emerged in Cabinda, integrating in its ranks most of the former TEs
of Alexandre Taty,60 while significant numbers of former GEs may have joined
the FNLA, which rapidly became the most influential military force in the North,
threatening Luanda directly. The foreign troops, particularly the Fiéis, quickly
became a sensitive issue. At the time they numbered around 2,400 in total, based
in three camps (Chimbila, Camissombo and Gafaria), and were militarily
organised in 13 commando companies (Tigres), plus around 300 in auxiliary
roles. They remained disciplined, with a strong military command, and still
obeying the orders of the Portuguese armed forces.61 The authorities considered
several alternatives, namely their local reintegration as civilians (as most of them
had left their families back in Congo, they had established new families in
Angola), negotiating at diplomatic level the return to Zaire of the ones who
wanted to do so, or discussing with the respective governments their passage to
Rhodesia and South Africa. The two first options were discarded due to the
strong animosity of the local population in the areas where the troops were
concentrated. This prevented a simple demobilisation of the force, for fear of
retaliations. Moreover, General Costa Gomes vetoed any contacts with South
Africa and Rhodesia in this respect, fearing the aggravation of an already tense
and intricate regional situation.62
J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50 149
63 See Chefe do Estado Maior das Forças Armadas de Angola, Despacho: Acção Fidelidade, 7 de
Agosto de 1974, Arquivo Histórico Militar, 2ª Divisão, 2ª Secção, Caixa 139, nº 7.
64 See conclusions in Chefe do Estado Maior das Forças Armadas de Angola, Despacho: Acção
Fidelidade, 7 de Agosto de 1974, Arquivo Histórico Militar, 2ª Divisão, 2ª Secção, Caixa 139, nº 7.
65 See Cervelló, “La revolución portuguesa,” 93 and passim. According to Emile Kalambo Ilunga,
interviewed in Brussels in May 1997 by Guido van Hecken, “en 1975, c’est le governeur portugais, qui
était un governeur rouge comme on l’appelait, qui a favorisé la jonction entre les katangais et Agostinho
Neto. (…) C’est grâce à leur participation à la guerre que Neto a pu conserver Luanda depuis 1975
jusq’à ce jour. Alors, il y avait donc là une sorte de devoir moral du MPLA vis-à-vis les katangais.” (I
thank David Hedges for this source). A little later, in 1994, the Lusaka Accord prescribed that all foreign
forces should leave Angola. However, the Katanguese had been by then already formally integrated in
the Angolan 24th special forces regiment, and therefore could have remained within the country.
66 Again, it is possible that some elements crossed the border to join the South-Africans in the South.
67 See, among other, the testimonies of General Almeida Bruno, and Mamadu Mané, in Rodrigues,
Os últimos guerreiros, 76-77, 166-168. There are widespread allegations that lists of former African
special troops were sent from Lisbon to Bissau, by someone well placed, and that these lists were used
In face of the uncertain future, the commander of the Fiéis threatened a
suicidal attack against Zaire (“since we are going to die anyway, we prefer to die
there rather than here in Angola”). This caused serious diplomatic concerns for
the new Portuguese authorities, who wanted to keep these men in Angola as long
as possible in order to avoid problems that might affect decolonisation in Angola
and Mozambique.63 It is nonetheless possible that substantial numbers of the
Fiéis in fact did manage to join the South-African forces, partly because a
conservatively inclined section of the new Portuguese authorities pressed in this
direction.64
The main solution that emerged, however, was to contact the Angolan
nationalist movements, particularly the MPLA, probing the possibility that the
Fiéis might be accepted as political refugees in Angola. This was convenient for
the movement, particularly the presidential faction of Agostinho Neto,
threatened by the FNLA, which had rapidly become the most influential force in
the North and must have absorbed many former GEs into its ranks. There are
indications that despite Zaire’s offers of amnesty, most of the Fiéis were
incorporated in the MPLA. The latter, also able to rely on important contingents
of Flechas, managed to retain Luanda and to regain military control in the
Northeast and in Cabinda.65 The case of the Zambian Leais was not much
different, and it is probable that they too were merged into the UNITA and
MPLA forces.66
In Guinea, on the contrary, matters took quite a different turn in this regard.
This was the case in territories where the nationalist war effort was the most
advanced, and where the new Portuguese authorities thus had less room to
negotiate. Guinea-Bissau reached an independence agreement rapidly, and the
African forces that had fought on the colonial side became more than irrelevant
as a threat to the new order. Consequently, many of them, particularly from the
African Commandoes, were arrested and it has been alleged that hundreds, if not
thousands, were simply shot after summary trials.67
150 J. P. BORGES COELHO, PORTUGUESE STUDIES REVIEW 10 (1) (2002): 129-50
to locate and arrest them.
68 See the testimony of Captain van Uden, one of the GEP commanders, in Rodrigues, Os últimos
guerreiros, 248.
69 Later, President Machel met with many of these Mozambicans. Some of them were arrested on the
spot, while others were allowed to resume normal civilian lives. There are records of the various sessions
of this meeting