Version classiqueVersion mobile
OpenEdition Books

The First World War from Tripoli to Addis Ababa (1911-1924)

Shiferaw Bekele
Uoldelul Chelati Dirar
Alessandro Volterra
et al.

Colonial Policies

The First World War Seen from Djibouti: Controlling, Recruiting, Enlisting

Laurent Jolly


For France, the Côte Française des Somalis represented a sea port of call on the way to the more prestigious colonies of the Indian Ocean or the Far East. Thanks to the railway from Djibouti to Addis Ababa, it also became a major route of the Ethiopian foreign trade. When the Great War broke out, it nonetheless found itself threatened by the Turkish presence on the shores of Yemen, the uncertainties regarding the position of Ethiopia in the conflict, and the rebellion in the name of Islam in Somaliland and Ogaden. The French authorities first tried to watch the populations of the small territory by reinforcing the security of the colony. But they also tried to mobilise the population for the war effort. The most visible part of this effort was the enlisting of over 2,000 men in the colonial army to go to war in France.
How can the stability of this colony during the world conflict be explained? How to account for the fact that hundreds of young men enlisted in a foreign army so easily? The article proposes to answer these questions using the study of the individual records of the recruits, which reveal these young men’s individual strategies in a fragile political, social and ecological context.

Texte intégral

  • 1 Service Historique de la Défense-Terre (SHD-T), 26N875-23, Journal de Marche et des Opérations (JMO (...)

1The French Somaliland contribution to the Great War is not often mentioned in studies about the engagement of Africans in the conflict and quite ignored by the populations from this part of the continent. This small colonial possession yet sent over 2,000 soldiers, known as “Tirailleurs Somalis”, to France. They fought in Verdun in 1916, at the Chemin des Dames in 1917 and then during the 1918 battles.1 This army unit known as “Bataillon Somali” was the only one consisting of soldiers from the Horn of Africa to take part in the European war. It was also one of the most rewarded colonial units during the conflict.

2In the Republic of Djibouti, this engagement is hardly remembered. There is only a narrow street in the centre of the capital named rue de Douaumont referring to the battle of Verdun and a War memorial built in 1935, close to the centre of the colonial power. It was taken down seventy years later and can be seen today on the grounds of the French military base. It is thus no longer visible to the public, nor is it a place for commemoration. Instead, it is hidden in the area devoted to the former colonial power, while this part of Africa is at a turn in its history.

3France was the only country involved in the Great War to resort to African combat troops for its conflict in Europe. The Somali soldiers were a small part of the “Force noire”, together with thousands of soldiers from West Africa and from Madagascar. However, the fact that they all enlisted on a voluntary basis made them different. Why was the French army so successful in recruiting in a recently colonised and hardly controlled territory? If recruits were not forced to enlist, contrary to what happened in other parts of the Empire, should their enlistment be considered as a conscious decision and be compared to some sort of “mercenary work”?

  • 2 Centre des Archives du Personnel Militaire (CAPM), Pau.

4This accommodating idea of the “warrior” has been reactivated by decades of political violence in this part of Africa. But is it not relevant to qualify this viewpoint by bringing to light the specific constraints making hundreds of young men enlist in a foreign army to go and fight in an unknown country? To answer these questions, we should first focus on the colonial relationship established in this part of the Empire, where the foreign presence was tolerated rather than enforced. To understand the young men’s motivations to enlist, a corpus had to be gathered. It is composed of Europeans sources, often biased towards the colonial point of view, but also of purely administrative and factual individual records, kept in the French Army archive centre in Pau.2

  • 3 Journal officiel de la République française, Documents parlementaires, séance du 29 juillet 1924, a (...)
  • 4 Jolly 2013.

5This archive centre keeps over 7700 “livret matricule d’homme de troupe” of soldiers recruited between 1899 and 1977, just before the territory became independent. The individual records were held by the soldiers but their military units retained copies which are in Pau today, although some information concerning men recruited before 1916 is missing. The first page of each record mentions personal information about the soldier (identity, parents, residence and personal features), including his clannish identity, an information that clearly became more precise from 1916 on. His military career and successive assignments are recorded on the other pages. For the First World War, 2403 records are available: 1541 for the year 1916 alone (men recruited between December 1915 and June 1916), and 862 “tirailleurs” recruited between January 1917 and October 1918. The total number of records preserved for this period corresponds to the number of recruits given by the 1924 Lyons de Feuchin parliamentary report.3 Consequently, we may suppose that the military institution preserved all the records of the Great War recruits. The article is based on a study4 of 580 individual records (366 for 1916, 108 for 1917 and 106 for 1918), that is to say over 20% of the recruits of the Côte Française des Somalis (CFS) during the conflict. These records were studied to find data about the men who chose to enlist and their personal strategies in a difficult political, social and ecological context.

Feeling Threatened by an Unknown Population

  • 5 La Quinzaine coloniale, 10 juillet 1913, p. 483.

6The railway from Djibouti to Dire Dawa in Ethiopia and the urban area of the colonial town were the only parts of the French Somaliland territory (CFS) to be known and controlled, but there was also Obock, an ‘Afar seaside town from where the coasts of the colony towards the Bāb al-Mandab straight could be watched. The shortage of human and material means in the colonial administration, the economic development of a disrupted but strategic possession and the difficulties of living and striving can explain why few Europeans chose to settle there. In 1913, a French colonial review5 published an estimation of the number of inhabitants in the town of Djibouti: out of 16,500, only 600 Europeans were counted. The non-European populations were called Arabs and included Indians, but over 10,000 inhabitants were identified as Somali, ‘Afar or Abyssinian.

  • 6 Archives Nationales d’Outre-Mer (ANOM), Côte française des Somalis (CFS), Affaires politiques carto (...)

7The colonial literature, picturing these people as unscrupulous barbarian warriors without any morals, contributed to their very negative image. Nevertheless, there was a real danger. The murders of Europeans show the pastoral population from the inner areas had no particular regard for these strangers who did not respect their uses or codes of honour. The governor of Djibouti thus described the murder of a French officer by one of his Somali soldiers in Obock in 1918 as a matter of personal vengeance.6

  • 7 Ali A. Hersi 1985.
  • 8 Van Gelder de Pineda 1995: 255.
  • 9 Bardey 1981.

8The European perception of the native populations can be explained by the way France settled on these East African coasts: with financial compensations, they concluded agreements with ‘Afar sultans or with some ‘Issa elders. This practice seems to have started during the short occupation of the Somali coast by the Egyptians7 and was also used by the Europeans. In the CFS, agreements with the ‘Afar sultans allowed French traders to settle on the northern coast of the Tadjourah gulf. In return, the French government paid them pensions in Thalers. As for the Somali ‘Issa, the 1885 and 1917 agreements do not mention any amounts to be paid, but all the dignitaries who signed them were pensioned by France, including the Ugaz, a moral authority among the ‘Issa. Colonial authorities went on paying intermediaries named okals until the independence. The beginning of the work on the Djibouti-Addis railway line (1898) turned out to be an opportunity for negotiations as well, since the company compensated the various ‘Issa clans disadvantaged in their usual activity of transit, as well as the Ugaz, by means of amounts in rupees.8 Negotiated and paid for, the foreign occupation was not imposed, though it did not prevent violence during the development of the railway. Negotiating and then buying one’s safety are omnipresent practices in the accounts of travellers and traders who ventured in the north of the Horn until after the Great War, when they became effectively controlled by the administration, the army and the police.9 Besides, when the British left the interior of Somaliland (1910), they ensured the loyalty of “allied” Somali clans by giving out arms, which was an aggravating factor of political instability in the territory.

  • 10 Cassanelli 2010: 53-66.

9Starting from the beginning of the work on the railway line, special attention was paid to local populations in order to get to know them better. This knowledge was mainly based on genealogies through which the complexity of the Somali segmentary structure could be comprehended and understood with the aim of better using and controlling the population. The understanding of the various Somali lineages, shared by the British, the Italians and the French, was documented via oral tradition and informers whose identity has been forgotten and whose liability may be questioned.10 This way of identifying people according to their lineage, their faction, their clan or their ethnicity, was used in the individual records of soldiers recruited during the First World War and was like an identity card. To which extent can these indications regarding clan be trusted? The knowledge of the Somali clans seems to have been passed on by Catholic missionaries of the Berbera mission, which was evacuated from Somaliland to Ethiopia and Djibouti in 1910. The archives of the mission kept in Djibouti confirm that, as soon as the end of the nineteenth century, such information was indeed taken into account with respect to Somali orphans who were taken in and baptized. The officer in charge of recruiting from 1915 on, Laurent Depui, wrote this information on the records, and it was sometimes completed later on. He may have been helped by a Somali interpreter who had to clearly identify homonyms, which are very frequent among the populations of the Horn. In 1917, he was replaced by a French priest from the Catholic mission, who carried on this task of identification. Lastly, all the “tirailleurs”’ descendants interviewed in Djibouti confirmed these complex clannish indications concerning their fathers or grandfathers.

  • 11 Ministère des Affaires Etrangères (Nantes), Consulat de France à Aden, Aden 9, Guerre 1914-1918, ét (...)
  • 12 SHD-T 7N2137, Compte rendu de mission en Ethiopie (Somaliland).

10The declaration of war greatly worried the small European population, which was only protected by an auxiliary force of around 200 men recruited without any military status and of uncertain efficiency. The entry of the Ottoman Empire into the war, the proclamation of Holy War in November 1914, and then the military operations around Aden, caused some concern. The colony asked for reinforcements, and Senegalese soldiers were sent from Madagascar to ensure security in the port town. In the peninsula of Djibouti, a defensive line was positioned near the native town and the salt pans. But the main concern in Djibouti was Islam. One dreaded the “fanaticism” of the Muslim population on the coast of the Horn of Africa, since sayyid Muḥammad ibn ‘Abdallāh Ḥasan (“The Mad Mullah”) was leading a rebellion in the name of strict Islam and fight against any foreign presence in British Somaliland, and Lij Iyasu was accused of sympathising with the Muslim religion and the Central Empires in Ethiopia. As soon as hostilities began, some natives from the colony were expelled because they were accused of anti-French activities.11 The qadi of Djibouti was particularly carefully monitored because he ruled over the Muslim population, and Muslim brotherhoods were closely watched.12

11This fear, excessive in many respects, accounts for one of the specificities of French recruiting in East Africa: all the soldiers enlisted in 1916 swore allegiance to France on the Quran.

Populations Requested to Help: Agreements and Enlistments

  • 13 SHD-T 26N76/14, Direction des étapes et des services, JMO.

12The colonial town largely contributed to the war effort, like other African possessions. First, hundreds of stokers from Yemen embarked on ships remaining in the port, being exposed to the dangers of navigating in wartime. Tens of them were counted among the men who disappeared after their ship sank in the Mediterranean Sea. In 1915, over 200 Somali coolies were recruited for the Allied Forces operations on the Dardanelles Strait against the Ottomans.13 They arrived on 1 September and were employed to unload the war ships that supplied the front. On their first day, several of them were injured by Turkish bombings. In a November 1916 report, Laurent Depui, officer in charge of the first recruitment for the war, recalls that the enlisted soldiers did not take the recruitment of unarmed coolies well, as they complained of being unable to defend themselves. He insists on what he sees as disillusionment for these men who expected to take part in the fighting. In early December 1915, almost three months after their arrival, the coolies enlisted for the Dardanelles were taken back to Djibouti. But the most obvious part of the war effort was the recruitment of soldiers to go to France. The shortage of men due to the loss of hundreds of thousands of soldiers at the beginning of the war led European countries to increase their number of recruits wherever they could. Given the fact that they did not know the hinterland well and lacked administrative control, they essentially conducted recruitment on a voluntary basis, which is confirmed by many sources. Each volunteer was given 200 francs for enlisting, which was quite a lot compared to what most native elders of the colony used to receive during the war, that is 60 francs per month. From December 1915 to November 1918, 2,403 men were recruited in French Somaliland, the majority of which was sent to fight in France.

  • 14 ANOM, CFS Affaires politiques carton 123, Brigade indigène, courrier du gouverneur Pascal, 20 juin  (...)

13The information on the first page of each individual record, which mentions the soldier’s place of birth and residence, as well as his clan or ethnicity, reveals that most of them came from outside the colony. ‘Afar and Somali ‘Issa herdsmen travelled across the territory of French Somaliland, but these two groups came in small numbers among the recruits of the First World War. This is particularly true of the ‘Afar, according to the available data. This was confirmed by later colonial sources. Several factors may account for the small number of native recruits in the colony. First, it is necessary to understand the decision of the colonial power, as soon as the first security force was created in the colony in 1910, to be protected by recruits from outside the colony.14 These auxiliaries either came from Yemen, or were not Somali ‘Issa. Geographical proximity also made temporary jobs easier to find for Somali ‘Issa herdsmen, who preferred a nomadic lifestyle and for whom integration to the colonial town took some time.

14The Arabs from Yemen, often qualified workers who could easily find employment, were only a minority among the recruits. Many of them were from Tihāma, an area along the Red Sea in Yemen, or from Aden and Laḥij, and less frequently from the coast of the Indian Ocean or from the inner country. The activities of most of the Arab recruits were related to trade, farming, crafts and soldiering (askaris). The presence of professional soldiers among Arab recruits was also related to the political division of Yemen. Moreover, European colonisers had relied on a small group of professionals from Yemen as soon as the first auxiliary units were created in this part of Africa. Thus, in the data available, there are former policemen or askaris from Aden, Djibouti, and Zayla, but these represent a small number when compared to the total number of recruits (about 5% of the sample). However, their abilities were appreciated and used to manage recruits, without any sense of hierarchy. Overall, there was a large number of Arabs to manage the Djibouti battalion.

15The Somalis were by far the most numerous recruits, over 80%. Their birthplace as well as the clan they belonged to clearly show that at least two-thirds of them were from (in descending order): British Somaliland, Ethiopian-controlled Ogaden and French Somaliland or, exceptionally, Italian Somalia. Around 60% of the Somali recruits declared to be nomadic herders and 20% of them presented themselves as coolies, at a time when the activity in the port was greatly reduced. Many of these men were obviously nomads who had recently arrived in town. Without any qualifications, they found employment as day-workers, but they were the first ones to end up out of work when the War broke out. This is also true of the young recruits who had declared themselves as “boys” or waiters in cafés, whose jobs were temporary until they could find better paid and, from their point of view, more honourable employment. The data concerning birthplaces and places of residence are imprecise. They often correspond to an urban unit with which the soldier was identified, like Berbera, Burao, Zayla or Hargeisa, or to a wadi, a mountain or a valley (Nugal valley for example). In the case of Somali ‘Issa or Gadabuursi, the indication might even correspond to some geographical location, such as a stop on the railway line or the kilometre number on the line.

Enlisting: heavy constraints

  • 15 Jolly 2016.

16For a herder also proud of being a warrior, it was degrading to work in exchange of a salary and enlist in the army. But that does not explain why some groups enlisted in larger numbers than others. Their reasons for enlisting should therefore be seen in relation to certain constraints common to all but also specific to some of the larger groups.15

  • 16 Lewis 1975.
  • 17 de Monfreid 1984: 94.
  • 18 Erlich 2014: 135-147.

17The first limiting factor of the pastoral economy was the frequent droughts in this part of Africa. The extreme climate conditions could result in food shortage, but they could also lead to conflicts, with the precarious balance between available pastureland and agreements between the groups of herders being broken. This is just what happened before the War. The terrible drought that struck the Horn of Africa just before 1914 is remembered as Xaaraamacune by the Somali population.16 Henry de Monfreid’s account in 1915 of the women and children rescued in Berbera17 recalls the events of more recent famines in this part of the world. Again in 1918, the colonial authorities in Berbera had to help feed the refugees. As these constraints were not specific to British Somaliland alone, other more specific factors must also be taken into account to explain why so many Somali Issak herders enlisted. This group represented more than half of the number of Somali recruits over the period and almost 80% of the men enlisted in 1917. All Issak clans were represented among the recruits, however a greater number of them were Habar Awal, one of the largest Issak groups in British Somaliland. Other Somali clans, such as the Darod Dolbahanta faction, were settled still further away from Djibouti. These clans were situated around the present towns of Berbera and Burao, and lived on the limit of British Somaliland and Italian territory, or in Ogaden. These groups were often attacked by the murderous raids of Mohamed Abdallah Hasan’s dervishes. This hero of Somali nationalism nevertheless deeply divided the Somali clans: there were those who followed him and those who fought against him. The different Issak clans did not accept his leadership and fought him with British weapons. In the same way, the Dolbahanta were deeply divided during this terrible retaliation from the dervishes. Each actor in the area manipulated the various Somali factions according to their own interests, using rivalries or other conflicts, favouring one to be envied by the others. Lïj Iyassou’s attempts to rally the Somali clans were unsuccessful, although weapons and presents for the elders were handed out.18 The Ottoman campaign for an uprising of all the Somali clans against the Allied forces and the Mullah’s call for Somali union did not influence them. Consequently, the combined consequences of droughts and political conflicts in the North of the Somali area can be traced to the French recruiting in Djibouti. Giving out firearms to the different Somali clans only aggravated the rivalries to get access to resources and violence between the groups.

  • 19 Jardine 1923: 198.

18A last indication in the records confirms this relationship between political violence and the hardships of surviving with military enlistment: it is the high number of orphans among the recruits. Almost two-thirds of the records mention that they were fatherless. Only 24% of them still had both a father and a mother. There were fatherless recruits among both Arab and Somali soldiers, but they were more numerous among the latter. The proportion of fatherless recruits was particularly high in the clans most persecuted by the dervishes. British or French archives, as partial as they may be, mention the deadly raids by the Mullah’s men on the groups refusing to join him: stolen herds, ransacked camps, as well as killings of women, children, and above all adult men. In his book published shortly after the end of the dervish revolt, Douglas Jardine, colonial officer in British Somaliland, mentions a real “holocaust” in which a third of the men in the protectorate died.19 Because these men represented the strength of their clans, they were the most important targets in order to weaken a group that could then no longer defend or perpetuate itself. For a young herder, being fatherless was a source of anxiety regarding the future. His lineage and his uncles were in charge of him but he was no longer first in line to enter adult life through marriage. Leaving the camp to try and manage all by themselves became an option for these fatherless “social cadets”. Finding employment, even precarious, in the new town of Djibouti, enabled them to get the amount of money equivalent to the herd they did not have. Avoiding such risks may also have led them to a recruitment office as an alternative. Moreover, enlisting as a “temporary worker” in the army got them an immediate reward, and being a soldier was well suited to their ideal of a warrior. This readiness to go to war is likely to have been one of the factors behind the achievements of the Djibouti unit during the conflicts in the North of France.

19The “tirailleurs” recruited in Djibouti arrived in France in June 1916 and in July, they left for Verdun, where they were first employed as workers. When they arrived in France, the recruits from the Horn were considered undisciplined, frail and hardly able to fight. But in August several colonial troops, among which the recruits from Djibouti, were “tested” on the front lines. Their resistance to the bombings won them the trust of the military staff. In October 1916 two out of their four companies were chosen to take part in the attack at Fort Douaumont, where they reinforced a distinguished European colonial unit (the “Régiment d’Infanterie Coloniale du Maroc”, RICM) with which they remained associated until the end of the war. The capture of fort Douaumont near Verdun on 24 October 1916 was a glorious success for the French army, widely related by the propaganda of the period, but it also proved the efficiency of the “tirailleurs Somalis”. From then on, the “tirailleurs” were employed as “shock troops” along European units. During the “Chemin des dames” attack (in Cerny-en-Laonnois in May 1917 and then in the fort Malmaison battle in October 1917) many of them were employed as “trench cleaners”: they had to kill with a knife the enemy soldiers hidden in the trenches that had been taken. Then, in June 1918, they successfully resisted at Mont de Choisy during the last German attack. After the Great War, the colonial consideration for the populations of the Horn changed: the dangerous and barbarian pastors had become a “troupe d’élite” well-appreciated by the military.


20The analysis of the reasons making hundreds of young men without any patriotic ideas enlist in a foreign army shows how the relative prosperity of the port of Djibouti made it the main place to find employment along the coast of the Horn of Africa. For these volunteers, who were often nomadic herders, enlisting meant taking some temporary work which they definitely chose for want of a better situation, but which was attractive enough to make them renounce their independence and their idea of freedom. The port and the railway line at once changed the commercial cash flows to the detriment of the old trading places on the northern Somali coast. The war allowed and increased the creation of migration routes and professional opportunities favouring Djibouti. It is also noticeable that the different calls to join the Holy War or to fight Europeans were fruitless. On the contrary, the large number of the men willing to enlist reveals individual strategies prevailed on uncertain political motives. However, all sorts of constraints determined these choices. Hasty conclusions about the belligerent attitudes of the various populations in the Horn should thus be avoided, and one should rather insist on their great capacity to adapt to life’s hazards.


Ali A. Hersi, ‘Egyptian administration of Somaliland (1874-1884). A prelude to disaster’, in International symposium on the African Horn, 5-10 January 1985, University of Cairo, Institute of African Research and Studies.

Bardey, A. 1981. Barr-Adjam: souvenirs d’Afrique orientale, 1880-1887 (Paris : CNRS, 1981).

Cassanelli, L. 2010. ‘Speculations on the historical origins of the "total Somali genealogy"’, in M. V. Hoene and V. Luling, eds, Peace and Milk, Drought and War. Somali Culture, Society and Politics. Essays in Honour of I. M. Lewis (New York: Columbia University Press, 2004), 53-66.

Erlich, H. 2014. ‘From Wello to Harer: Lïj Iyasu, the Ottomans and the Somali Sayyid’, in E. Ficquet and W. G. C. Smidt, eds, The Life and Times of Lïj Iyasu of Ethiopia (Berlin: Lit Verlag, 2014), 135-147.

Jardine, D. J. 1923. The Mad Mullah of Somaliland (London: Herbert Jenkins, 1923).

Jolly, L. 2013. Le tirailleur « somali » : le métier des armes instrumentalisé (début xxᵉ siècle, fin des années 60), thèse de doctorat, Université de Pau et des Pays de l’Adour (2013).

Jolly, L. 2016. ‘Les tirailleurs de la Corne de l’Afrique, déconstruire le mythe du guerrier’, Le mouvement social, 256 (2016/3), 101-116.

Lewis, I. M. 1975, ed., Abaar, the Somali drought (London: International African Institute, 1975).

Monfreid (de), H. 1984. Journal de bord (Paris: Arthaud, 1984).

Van Gelder de Pineda, R. 1995. Le chemin de fer de Djibouti à Addis-Abeba (Paris: L’Harmattan, 1995).


A Somali soldier’s individual record kept in Pau (CAPM). His name is partially undisclosed.

The first page of another individual record with social information.


1 Service Historique de la Défense-Terre (SHD-T), 26N875-23, Journal de Marche et des Opérations (JMO) du bataillon somali (1916-1918).

2 Centre des Archives du Personnel Militaire (CAPM), Pau.

3 Journal officiel de la République française, Documents parlementaires, séance du 29 juillet 1924, annexe n°335.

4 Jolly 2013.

5 La Quinzaine coloniale, 10 juillet 1913, p. 483.

6 Archives Nationales d’Outre-Mer (ANOM), Côte française des Somalis (CFS), Affaires politiques carton 123, dépêche télégraphique du gouverneur, 6 janvier 1918.

7 Ali A. Hersi 1985.

8 Van Gelder de Pineda 1995: 255.

9 Bardey 1981.

10 Cassanelli 2010: 53-66.

11 Ministère des Affaires Etrangères (Nantes), Consulat de France à Aden, Aden 9, Guerre 1914-1918, état de siège en CFS, arrêté d’expulsion du gouverneur et rapports du lieutenant Depui (6 octobre 1914) et du brigadier Thomas (2 novembre 1914).

12 SHD-T 7N2137, Compte rendu de mission en Ethiopie (Somaliland).

13 SHD-T 26N76/14, Direction des étapes et des services, JMO.

14 ANOM, CFS Affaires politiques carton 123, Brigade indigène, courrier du gouverneur Pascal, 20 juin 1910.

15 Jolly 2016.

16 Lewis 1975.

17 de Monfreid 1984: 94.

18 Erlich 2014: 135-147.

19 Jardine 1923: 198.

Table des illustrations

Légende A Somali soldier’s individual record kept in Pau (CAPM). His name is partially undisclosed.
Fichier image/jpeg, 352k
Légende The first page of another individual record with social information.
Fichier image/jpeg, 168k


Chercheur associé LAM/UMR 5115-CNRS-Sciences Po Bordeaux, Laboratoire « Les Afriques dans le monde », Université de Pau et des Pays de l’Adour, France

© Centre français des études éthiopiennes, 2018

Conditions d’utilisation :



Offert par