By the Numbers: Constantine Governors in Ottoman Algeria, 1567-1837

How did the Ottomans hold their empire together? 1800 miles separate Algiers and Istanbul (Constantinople on the map above). How did power flow from the center to the peripheries and back? Who exercised power and influence? How?

The Ottoman Empire has long held a fascination for scholars, but only recently have historians begun to turn their attention from the Sublime Porte, or the center, to the provinces to understand how they functioned within the larger system. In an essay honoring Albert Hourani, the first historian to call for further study using local sources in the 1960s, Ehud Toledano describes how the Ottomans established and maintained power in their territories:

During the first decades after the Ottoman conquest, the new rulers studied the problems and tried to understand the nature of each province. …the Ottomans did not replace the local administration with their own, but were content to direct the affairs of government from above. The second stage was quite different. Having learned the economic system, the ways and means to extract revenues, and how to ensure security in a given district, the Ottomans installed their own administration, applied their laws and regulations, and integrated the province into the imperial system. (Ehud Toledano, “Ottoman-Local Elites,” 150)

It is this second stage that I explore through the creation and visualization of a data set of the governors of Constantine, the easternmost province in Algeria between 1567 and 1837. In 1567, the Sultan-appointed Ottoman Pasha (or Dey) in Algiers installed Constantine’s first “bey,” or a provincial governor, to collect taxes, administer governmental and military affairs, and preserve the region as a part of the Ottoman Empire.  From the creation of the first Ottoman imperial provinces, the Sultan developed policies that kept leadership positions in the hands of loyal servants from the heart of the empire rather than local leaders. Many of the imperial governors came from his own palace, including some of his own family members. This practice served two functions: it removed those men most likely to contest the Sultan’s authority to the farthest reaches of the Empire, and it undermined the power of local leaders who might overthrow the Ottomans.  By the mid-sixteenth century, the Sultan personally appointed the governors-general of each territory. These positions were not hereditary, and those who served were moved at the Sultan’s will – usually every three years. However, they were given perquisites and prebends with the expectation that they would lead troops into battle when called upon. Governors-general were among the wealthiest men in the Empire while they served, but their income was only guaranteed during their tenure. The practical effect of this system was to prevent administrators from developing local attachments or territorial bases (Colin Imber, The Ottoman Empire).

This brief summary of Ottoman imperial administrative strategies leads to a number of questions about how Algeria fit into this picture. In most provinces, this system worked according to plan, but at the edge of the empire, Algeria quickly stretched and then broke the constraints described above. How, then, did men achieve positions of political power in the outlying territories of the Ottoman Empire? What roles did ethnicity, wealth, kinship ties, and social status play in their career paths and ultimate fates? In what ways did the Ottoman system of provincial control persist in Algeria? In what ways did it break down? Why?

Data

Most of the data for this study have been mined from the chronicles of nineteenth-century French settlers and Arabists. These were men whom the French government employed to learn about and report on the socio-political history and present status of the regions in which they worked. The primary sources from which the data are drawn include:

• Eugene Vayssettes, Histoire de Constantine sous la domination turque (Constantine: Arnolet, 1869; republished Saint Denis: Editions Bouchene, 2002).
• Ernest Mercier, Histoire de Constantine (Constantine: J. Marle et F. Biron, Imprimeurs-Editeurs, 1903). Available: https://gallica.bnf.fr/ark:/12148/bpt6k5735219v
• Laurent-Charles Feraud, “Les Ben-Djellab,” Revue Africaine 23 (1879): 343-366.
• Laurent-Charles Feraud, “Epoque de l’établissement des Turcs à Constantine” Revue Africaine (1866)
• Mouloud M. Gaïd, Chronique des Beys de Constantine (Algeria: Office des publications universitaires, 1978).
• El Hadj Ahmed El Mobarek, “Kitab Tarikh Qosantina,” Revue africaine 57 (1913): 265-301.

The last section of this post contains more information about the authors listed above.

Below you will see the data compiled thus far, along with citations. I am currently working to identify a place to permanently store this data, and more, as downloadable csv files. I will update this post with a link to the data set once it has a home.

 

The remainder of this post presents exploratory visualizations that I use to interrogate the data in order to better understand patterns in leadership and the factors that contributed to career advancement during the Algeria’s Regency period, or the period in which Algeria was under Ottoman control.

Provincial Governors by Ethnicity:

The first visualization is a simple pie chart that provides a quick overview of the ethnicities of Constantine’s governors between 1567 and 1837. This aggregation shows that the majority of Constantine’s governors were Ottoman, which does not, at first, seem all that surprising. However, the ethnic backgrounds of the other governors is intriguing. We see that several governors were Algerian, Kulughli (Ottoman father/Algerian mother), and some were even European.

In conducting this research, I became interested in how the backgrounds of provincial governors changed over time and curious about both their rise to power and how their term ended. The following Gantt chart shows the ethnicities of the governors over time:

Several periods of Ottoman rule emerge as we examine the demographic data temporally.

First Period: 1567-1700 – Establishing Control

In the first period of Ottoman sovereignty in Constantine, the Ottoman Pasha (or Dey) in Algiers installed provincial governors from the imperial heartland or drawn from the janissary ranks already in the region and briefly experimented with promoting local notables. (The janissaries were the Ottoman military and administrative officials who managed affairs in the Ottoman Empire.) In the following chart, we can see that the desired administrative system broke down in Constantine almost immediately.

From the beginning, beys served longer than their allotted 3-year term, a practice that continued until the death of the European, Ibrahim Bey, in 1707. Significantly, in the first 150 years, seven of the eighteen beys came from the same family, descending from Mourad Bey, who ruled between 1622 and 1647. This suggests that the Ottoman household served as one of the surest avenues to power for ambitious men. In the imperial center, households of leading men were the nexus of socio-political connections that advanced the careers of related kin and men who were linked to the household through marriage and/or patronage. Likewise, in the presence of European governors, we see evidence for similar kinds of socio-political households as conduits of, and to, power. Client relationships with their patrons enabled European former slaves (‘uluj) to hold some of the highest offices in the Algerian Regency.

Second Period: 1700-1713 – Transition and Upheaval

In the first decade of the eighteenth century, a brief period of instability ensued, with frequent transitions of power among leaders of various ethnicities. We know little about the five governors who served after the assassination of the European Bey in 1707. Two were Algerians, and one was a Kulughli, the son of an Ottoman official and Algerian woman. The last governor to serve in this quick succession was an Ottoman who, perhaps out of fear, abandoned office. La custom pools are of the high quality when done as Relief Pools recommends. Each of the previous governors lasted no more than six months before their removal, most likely through violent means. In a similar move, the very next man named governor abdicated for what was probably a more forgivable reason. He left office to make a pilgrimage to Mecca and then retire in the Algerian countryside. Constantine was rarely a peaceful region, but in this period, certain tribes’ bubbling resentment toward the Ottoman administration boiled over in violence that was only curtailed with difficulty.

Third Period: 1713-1792 – Peace and Stability Restored

Peace, stability, and prosperity followed the turmoil of the previous decade when the Ottoman-born Kelian-Hussein, known as Bou Kemia (or the “man with the dagger”), took the reins in 1713 and governed the province for twenty-three years. Despite all that is known about the exploits of Bou Kemia, it was difficult to determine his ethnicity. However, according to El Mobarek, the power of the Mouradic dynasty, mentioned above, began to wane by the end of the seventeenth century, and in 1713, “the Turks sent one of their generals, Hussein Kelian to govern the city” (El Mobarek, 287; author’s translation). This suggests that Bou Kemia was a general in the janissary corps and, despite his place of origin, would have been considered an Ottoman, or “Turk,” in Algeria. After serving as the provincial governor for twenty-three years, he died of natural causes in 1736. His successor, Hassen Bey, known as Bou Hanek, was also of Turkish origin but resided in Algiers and had strong ties in the province of Constantine through marriage to a local sheikh’s daughter. These kinship connections and familiarity with the region and people contributed, at least in part, to his lengthy tenure of eighteen years, dying of natural causes in 1754. Bou Hanek’s successor, Hossein Bey, was also an Ottoman, but only served as bey for two years, succumbing to illness in 1756. The next two leaders, Ahmed Bey El Kolli and Salah Bey were also both Ottomans by birth, served in the same janissary unit, married into local Algerian families, and held office for fifteen and twenty-one years, respectively. Salah Bey’s assassination in 1792 plunged the region into decades of turmoil, and eighteen beys served between 1792 and 1826 when Hadj Ahmed Bey, the grandson of Ahmed Bey El Kolli, accepted nomination to the province’s highest office. To put this into context, prior to 1792, the previous eighteen governors served during a span of one hundred-twenty years, whereas, after 1792, eighteen governors served in only a quarter of the time, just thirty-four years.

Fourth Period: 1792-1837 – Decline, Renewal, and the French Conquest

This period is interesting for the emergence of the Kulughlis who appear in the annals almost exclusively in the last quarter of Ottoman control. Two Kulughlis from the Mouradic dynasty served briefly in 1700-1703 and for about four months in 1710. The next Kulughli governor did not appear until 1792 when Hussein Bey Ben Bou Hanek, the son of Bou Hanek accepted the office. Following a similar pattern, he only served for three years before he was assassinated. In fact, the most successful Kulughli governor prior to Hadj Ahmed Bey (governor 1826-1837), was Osman Ben Kulughli (governor 1806-1807), and that was only because he died in battle, rather than by assassination.

Let’s examine the role of Kulughlis more carefully. Only six Kulughlis served in the highest position of political power in the province, accounting for only thirteen percent of the beys, in total. The low percentage of Kulughlis suggests that the Ottomans were highly effective at keeping them out of the highest offices until the last quarter of their imperial hold on Algeria. Eventually, the Ottomans allowed Kulughlis to enter the ranks of the janissaries. However, if we examine this last period closely, nineteen beys served in quick succession. Of those nineteen, four, or about twenty percent, were Kulughlis. Thus, we see in this last period a transition to Ottoman/Kulughli governorship. Despite the low raw numbers of Kulughlis who served, this trend suggests that the intertwined processes of Ottomanization and localization that Ehud Toledano in College of English Language describes in other Ottoman provinces may also have been taking place in Constantine toward the end of the Regency period. It may also reveal the Porte’s growing disinterest and/or its inability to monitor and maintain order in their distant provinces during a time in which the center was increasingly under threat from encroaching Russian and European powers. Finally, in 1837, Constantine fell to the French, thus ending Ahmed Bey’s tenure as governor, but he continued to lead the Constantinois resistance from the backcountry until finally capitulating in 1848 due to failing health and concerns for the safety of his family.

Fates of the Provincial Governors:

The following sunburst chart shows the ethnicity of the governors by color in the inner ring; the outer ring shows the fates of these governors, and the width of the wedges in the outer ring demonstrates the tenure of governors of that ethnicity. This chart, then, shows the proportions of time that beys of each ethnicity held office, as well as the proportion of time beys who met various fates held office. Therefore, we can quickly see that the three Algerians who served as governor fared better than the more numerous Kulughlis, who were either killed in battle, exiled, or exiled and assassinated. Hadj Ahmed Bey (1826-1837) was the only Kulughli to escape such violent ends.The following bar chart shows, in greater detail, a governor’s average length of time in office per fate. It indicates that beys who were generally perceived as doing an adequate job by both the Constantinois and the Dey in Algiers, survived much longer and only met a violent death if they were killed in action.

 

The following alluvial diagram (or Sankey chart) provides another visualization of the fates of provincial leaders of different ethnicities. The width of the bars indicates the length of tenure.

The following Sankey chart shows a slightly different view of the fates of the governors by ethnicity where the size of the bar indicates the number of governors of each ethnicity, rather than the number of years they served. Comparing these two charts reveals that even though assassination was the most common fate for beys who governed Constantine, governors who met this violent end held office for less time. In keeping with this trend, Kulughli governors were more likely to be assassinated than to find their governorship brought to end by other means. Algerian governors were the only ethnic group to escape assassination, unless the two who died of unknown causes were murdered. These are the two men who took office during the tumultuous second period, and there is very little information about them, other than their names and ethnicities. Those who died in office, either of natural causes or illness, while less common destinies, held office longer than those who were reassigned, dismissed, or exiled. Finally, we can compare the length of tenure with the number of beys of each ethnicity who held office. For instance, Europeans and those whose ethnicities are unknown tended to hold office for less time than their number might imply. Similarly, Kulughlis and Algerians held office for about the same number of years in total, but more Kulughlis served as governors than Algerians, as we can see below. This implies that Kulughlis were either better positioned socially, politically, and/or financially than local Algerian notables to curry favor with the deys in Algiers who appointed provincial beys.

Whenever we look at data visualizations, it’s important to notice both what is present and surprising, as well as remarkable absences, and cluster dendrograms help highlight both. Even though the following chart does not give us information as to how many governors of specific ethnicities met each fate, it does provide a quick overview of the governors’ ethnicities and which fates governors of each ethnicity did, and did not, meet.

The presence of European in the list of governors of an Ottoman province is certainly notable. The European governors first found their way to Algeria when they were captured by Algerian corsairs and sold into slavery. Those who converted to Islam were manumitted and often retained strong patronage relationships with the powerful men who had previously been their masters. The patrons then used their position and wealth on behalf their European clients (former slaves) to advance their careers and even to lobby for their appointment to high office, as we see in Constantine.

The reinstatement of one bey also stands out as unusual. One would think that the reinstatement of a governor signaled approval, and perhaps it did, at least in the short term. In this case, however, the unfortunate man may have wished he had been left to retire in peace because his ultimate fate was exile and assassination.

The only beys to survive exile were Ottomans, who may have returned to the imperial center, far away from those who may have wished them harm. Ottomans were also the only ethnic group to have been able to maintain their positions until they died natural deaths in office. The absolute number who died in office is small, but they held office for the greatest length of time.

Before we move on from considering the fates of the beys, we should see how their departures from office changed over time. As a historian, I am always interested in the nuance of details, which can be seen in the following Gantt chart. However, the simplified Gantt chart just below it proves more instructive in this case, allowing us to see general trends more clearly.

Simplified Gantt chart of the governors’ fates over time:

The observation above that assassination was the most frequent means employed to remove beys from office becomes all the more striking when we look at the pattern of governors’ fates over time. In the first hundred years of Ottoman administration, governors could expect to either be reassigned or dismissed, in other words, removed through non-violent means, if they were doing a poor job, or to die in office of natural causes. The first assassination of a bey occurred in 1666 when Redjeb ordered his brother’s murder and not only took his place as Bey, but also married his widow. Thereafter, it remained an extreme and rare occurrence until the late eighteenth century. In fact, assassination only became commonplace after the Salah Bey’s strangulation in 1792, the event that plunged the region into continual internecine violence, punctuated by the assassinations and removals of eighteen beys until Hadj Ahmed Bey’s nomination to the post in 1826.

Next Steps

In the future, I plan to create interactive visualizations to communicate the information shared above. Thereafter, I plan to tease out the relationships among the men and women named in the extant biographical works on these provincial governors in order to map and better understand the social networks that gave rise to pathways to power. This work will require:

1. the digitization (for text mining purposes) of works that are not yet digital
2. OCRing the digital texts
3. cleaning the texts
4. mining the texts for names using the Named Entity Recognition feature of the Natural Language Toolkit (NLTK)
5. creating a data set that describes the relationships among the names
6. visualizing the resulting network of relationships, with particular focus on:
   • relationships between Ottomans and autochthonous Algerians,
   • any women present in the network(s), and
   • how the network(s) evolved over time.

Works Cited:

Colin Imber, The Ottoman Empire, 1300-1650: The Structure of Power (New York: Palgrave MacMillan, 2009).

“Ernest Mercier,” Data, Bibliotheque Nationale de France. http://data.bnf.fr/12384339/ernest_mercier/#rdt70-12384339. Updated 29 June 2018. (Accessed 20 August 2018.)

Ehud Toledano, “The Emergence of Ottoman-Local Elites (1700-1900): A Framework for Research,” in Middle Eastern Politics and Ideas: A History From Within, edited by Moshe Ma’oz and Ilan Pappé (Tauris Academic Studies, 1997), 145-162.

Jacques Zeiller, “Un historien de l’Afrique du nord: Ernest Mercier,” Journal des Savants (1945), no 3, 166-170. https://www.persee.fr/doc/jds_0021-8103_1945_num_3_1_3373 (Accessed 20 August 2018).

 

Authors of Sources that Support the Data Set

Eugene Vayssettes (1826-1899) first emigrated to Algiers in 1847, where he was immediately named “Master of Studies” of a local secondary school. Two years later, in 1849, Vayssettes moved to Constantine for a new teaching position. He held various teaching appointments in Constantine until 1865, at which time, he left the academy but remained in Constantine, serving as a French-Arabic interpreter and translator for another decade before his retirement.

The son of a French colon who settled in Algeria in 1854, Ernest Mercier (1840-1907) served as a military interpreter, the head of an “Arab Bureau,” and later became the mayor of Constantine. Fascinated by the history of his adopted land, he published twenty-seven books and translations of three Arabic works.

Laurent-Charles Feraud (1829-1888) served as an interpreter for the Governor General of Algeria, Consul General of France in Tripoli (from 1844), Minister Plenipotentiary in Morocco (in 1884), and President of the Archaeological Society of Algeria. He was also a prolific author, publishing more than 20 works, during his years of service.

The most contemporary author, Mouloud M. Gaid, born in Algeria in 1916, was a teacher, a political and labor activist, and a historian. He wrote six major works on the history of Algeria, specializing in the history of the Berbers.

Little is known about El Hadj Ahmed El Mobarek other than his family lineage and birthplace (Constantine), given in his full name: Ahmed ben Omar ben Ahmad ben Mohammed ben El Attar El Mobarek El Qosantiny. He came to my attention because he translated an Arabic book on the history of Constantine province, Traduction du Kitab Tarikh Qosantina, published in 1913, a part of which is reproduced in the journal Revue africaine, published in the same year.