1Pakistan armed forces were, in 1947, unified both by their officers’ training at the British Raj’s academies and by their absolute faith in the new state and its cultural basis, Islam1. This sharply contrasts with the highly divided army to which the secessionist state of Bangladesh gave birth at the time of its independence in 1971. Indeed, from the onset, two major specificities distinguished the Bangladeshi army from its counterparts in other South Asian countries.
2Firstly, the Bangladeshi army was not a direct legacy of British colonial military culture but the result of a 'Liberation War'. The group who led this war was highly heterogeneous – rebel officers, privates and civil militiamen recruited by nationalist and Marxist parties – and divided into multiple revolutionary cells and political factions. Moreover, it never formally acknowledged Sheikh Mujibur Rahman (Mujib), the 'father of the nation', as the legitimate leader of the new state. Consequently, when Bangladesh emerged as an independent state it faced an uneasy situation: on the one hand, most of the military elite was not integrated into the power structure, but rather ideologically contested it, and on the other hand, it was considered with suspicion by the first head of state, Mujib (1971-75).
3The second unique quality of the early Bangladeshi army lies with the fact that the Independence war itself had generated strong inequalities and additional divisions within its ranks. Participants in the war (the 'freedom fighters') received honours, privileges and promotions, favours denied to the 'repatriated' – those who willingly or forcefully stayed in West Pakistan during the conflict – upon their return in 1973. After Mujib's assassination in 1975 nevertheless, repatriated Bengali officers (around 60% of the armed forces then), initially polarised against him and his partisans, began to strengthen their position, while the 'Liberation heroes' were, now, discarded from key positions (till 1991). This empowered group enjoyed a key position in the new ruling class (i.e. the army, the top bureaucracy and a class of new businessmen and industrialists). It formed the main support-base to the civilianized governments of Generals Ziaur Rahman (1975-81) and Hussain Muhammad Ershad (1982-1990).
4How do these two decisive historical legacies account for the nature of civil-military relations in Bangladesh? How did democratization, and the subsequent subordination of the army to civilians, impact on the Bangladeshi army itself? To answer these questions, this paper focuses on three issues, which chronologically represent three significant moments in civil-military relations: 1/ the 'army-building' process, its earliest factionalism and the strong influence of party politics, 2/ the exercise of political power and the army’s growing professionalism and, 3/ the democratization process. In this last phase, the civilian power had been able to reclaim control over the military elite but, interestingly, also re-politicized it. Therefore, the trajectory of the Bangladeshi army demonstrates that, contrary to an opinion commonly shared by many analysts of civil-military relations, democratization does not necessarily lead to a higher degree of professionalism, on the contrary2.
5Because Bengalis had been in very small number in the British Indian army, Pakistan's eastern wing’s soldiers amounted to only 1 percent of the entire Pakistani armed forces at the time of the country's independence in 1947 (Islam 2002: 363)3. Thus, one can hardly speak of a 'British heritage', except perhaps with regard to the maintenance, after 1947, of the bio-cultural theory of 'martial races'4 – the Bengalis, seen as puny and 'naturally' rebellious intellectuals, were the black sheep of the Pakistani army. The body count of inductees in 1948 speaks for itself: out of 2,795 fresh privates, only 87 were Bengali (Islam 2002: 366). Just before the 1965 war with India, East Pakistani men respectively composed 6, 15, and 16% of the army, navy, and air force. Eventually, in 1970, the number of Bengali professional fighters reached 34,000 out of some 412,000 in the Pakistani military, including only 300 officers ranking at, or above, Major rank, and just one Major-General (Ahamed 1988: 51-52; Islam: 2002: 369). As a global figure, and over 24 years of cohabitation with West Pakistan, the eastern wing’s strength never amounted to more than 9 percent of the total number of soldiers and officers in the army.
6Whether they would later take part in the Liberation war or not, these Bengali soldiers felt a common resentment against Pakistan’s western wing, and this for two reasons. First, Ayub Khan (1958-1969)’s reforms, guided by a professed will for national integration, consisted, in fact, in creating quotas in the different forces: Bengali officers were generally discarded from promotions in the army, and though some posts were created in the two other corps, most of them belonged to technical and administrative sections5. General Ayub not only judged the Bengalis 'non-martial', but also considered them as an 'oppressed race', whose destiny was to be forever ruled by foreigners and to occupy subaltern positions in the army (Khan 1967: 187).
7Secondly, it was obvious that East Pakistan, with four regiments only, was left totally defenceless in case of an Indian attack. The strategy that was adopted during the 1965 Indo-Pakistan war did little to allay this fear. Indeed, concentrating forces on the western front, launching an attack on Kashmir and occupying Indian territories to create pressure in case of invasion of East Pakistan did not represent guarantees to protection for the Bengalis. Actually, the dereliction of the army’s ideal became blatant in the eyes of the Bengali officers after the 1965 war: the army could not genuinely stand as the guardian of the nation and the symbol of Pakistan’s unity. In 1971, the Bengali military were thus concerned with two distinct, but as potent, threats: East Pakistan’s vulnerability to India and the neglect of the province by the 'centre'.
8Upholding this latter claim, a movement towards greater autonomy had already been gaining momentum in East Pakistan since the mid-1950s, with Mujib and his party, the Awami League (AL), at the forefront. The AL’s landslide at the December 1970 elections, which should have allowed Mujib to become Pakistan’s Prime Minister, never led to the democratization of the regime, as the army and most of the West Pakistani parties were reluctant to transfer power to an assembly dominated by Bengali representatives. The first three months of 1971 therefore witnessed an aggravation of the anti-government campaign in East Pakistan, while the AL’s leftist trend and student front, favouring a complete separation of 'Bangladesh', gained ground in the party’s decision-making bodies.
9Following the crackdown by the Pakistani army in East Pakistan’s cities, and the arrest of Mujib on 26th March, top AL leaders, including some of the newly elected representatives, crossed over to India and formed themselves into a constituent assembly which proclaimed the independence of Bangladesh. They also legalized the creation of a 'Bangladesh government-in-exile' on 10th April, 1971, in a village of West Bengal later renamed Mujibnagar. This cabinet, dominated by pro-India politicians, constituted the 'official' political leadership during the war, and played a critical role in facilitating India’s support for the mukti bahinis6 in terms of logistics and training.
10That being said, the human picture of the battlefield in 1971 was actually much more complex than the clear-cut struggle between the Pakistani and the Liberation armies suggested by the Bangladeshi nationalist historiography. The variety of actors and interests involved in the conflict is worth describing, as it accounts for the post-war civil-military tensions and the factionalism within the Bangladeshi army. On both sides, three different sets of actors took part in the conflict:
11Around 70,000 Pakistani soldiers were deployed in East Pakistan to crush the rebellion. The Bengali soldiers belonging to East Pakistani regiments who survived the crackdown – roughly 6,000 men – joined the Liberation army. The remaining Bengali troops of the Pakistani army were confined in West Pakistan.
12The historic East Bengal Rifles created by the British, composed of 15,000 Bengali men, supported the Liberation army. But the 30,000 razakars, who were non-Bengali volunteers raised after the 1965 war in case of an Indian attack on East Pakistan, stayed loyal to the state and worked with the Pakistani regular troops.
13The main specificity of the war lies in the fact that a significant number of civilians took up arms and joined the respective camps. On the pro-Pakistan side, Islamist parties like the Jamaat-e-Islami raised two militias, namely Al-Badr and Al-Shams, which were supervised by an army officer7; because these militiamen were Bengalis who fought against the independence of Bangladesh, they were – and still are – stigmatized as 'collaborators'8. As far as the pro-Bangladesh camp is concerned, the bulk of the troops were, in fact, scattered civilian militias9. Whereas the regular and paramilitary troops were taking orders from Mujibnagar and the official chain of command, a great number of these militias could be considered as autonomous armed gangs formed of political activists, graduated students, peasants, etc. which were not accountable to anyone. These were partisan militias like Abdul Matin’s East Bengal Communist Party – Marxist Leninist (EBCP-ML), or private gangs like Quader Siddiqui’s, a charismatic godfather in the Tangail district. In some instances, serious tensions even arose between the militias’ chiefs and the Mujibnagar government. That was the case with the Mujib Bahini, a group 5000-men strong, which was led by AL activists close to Mujib and had been created without the knowledge of the government-in-exile or the Liberation army’s command, allegedly to build an alternative leadership if the war was to last longer.
14Out of these 100,000 mukti bahinis’ fighters, no more than 7000, in addition to 100 rebel officers commissioned under Ayub’s or Yahya Khan’s rules, actually joined the Bangladeshi army after the war. The great majority of freedom fighters went back home and let the political leadership take care of building a new nationalist mythology around them, now known as 'Liberation heroes'. However, the war experience left some serious stigmas among the mainly rural population of veterans. Whether we consider the official figure – 3 million dead – or the conventionally accepted death toll – between 1 and 2 million – the planned and massive attacks perpetrated against civilian populations by the Pakistani army and Islamist militias led some researchers to talk of a genocide (Bhattacharya 2000: 102), Moreover, the post-war retribution-based violence that targeted the so-called 'collaborators' turned the conflict and its aftermath into a collective trauma. Direct and indirect civilian participation in this war, as well as ideological divisions within families themselves, strongly politicized large segments of society.
15For quite a few officers, the Independence war was considered as a revolutionary experience, initially because of the way it started. The Chittagong battalions’ uprising of 26th March, 1971, was an act of rebellion against the government, and consequently against the military hierarchy itself (East Pakistan was already under martial law)10. This event represented a dramatic parting from the Ayub era’s: Pakistani officers had already tried to destabilize political order, substitute a head of government with another, and exercise political power themselves as either 'custodians' or 'rulers' (the military was doing it since 1953) (Nordlinger 1977: 21-27). But this was the first time that a group of officers was aiming at breaking up the state11.
16The war of Independence, given its very nature, sowed the seeds of the profound mistrust between military officers and Mujibnagar politicians which would characterize the first years of the new state. Indeed, officers often had to make much more radical choices than those expected by the political leadership. This was particularly the case for the commanders of the 'interior sectors' – in reality militia chiefs operating autonomously in each zone – such as Colonel Abu Taher or Major M.A. Jalil, two men who would oppose Mujib’s government after the independence of Bangladesh and would play an important political role12. Taher and Jalil wanted to preserve the fighters’ logistic independence from India’s interference, even at the cost of a longer war, therefore favouring scattering informal guerrilla units across the land and surgical attacks against the Pakistani army. This strategic option implied a prolonged proximity of the Bengali soldiers toward 'civil society' – here used in the first meaning of 'non-military'.
17Yet, the 'interior sector commanders' were, as well, soon at odds with another section of the mukti bahinis’ command-structure itself: the commanders of the three brigades based across Assam and Tripura borders13. The latter’s strategic vision but also, most importantly here, their relationship with the political leadership were of an entirely different nature. They preferred a strategy of 'formal resistance' and conventional fighting across the border, which inevitably meant to cooperate with the Indian military14. This option was eventually chosen by the 'Higher Command of the Liberation Army', as the Bangladeshi historiography terms it, based in Agartala (Tripura) and led by Colonel (retd.) Osmany and Major Khaled Musharraf. Contrary to its military rivals from the interior, this Higher Command immediately submitted to the AL’s influence, i.e. to the government-in-exile headed by AL’s Secretary General, Tajuddin Ahmed15 – as well as to India’s influence.
18At that point, the Mujibnagar government, torn between the interests of the far left and of India, made a decision of enormous significance for the future of the country’s civil-military relations. To impose a political direction on the conflict and discard the communists, the exiled cabinet 'offered' ranks of Major-General and Brigadier-General respectively to the 167 AL Members of Parliament and to the 250 Members of Provincial Assembly, all elected in the 1970 Pakistani elections (Khan 1984: 104). The political calculation seemed symbolic, as the mukti bahinis’ 'big chiefs' were not officially higher than the rank of Colonel. But actually, from now on, the 'general-made politicians', superseded the remaining professional soldiers’ hierarchy16. This also meant that the Mujibnagar government eventually entrusted the army’s command in civilian hands during the whole conflict.
19This decision increased the mutual disdain between politicians and soldiers (the former contemptibly viewed as unit commanders primarily concerned about their future position). It also had two additional, and major, consequences on the military elite: it sharpened its anti-Indian sentiments and convinced the left-leaning officers that the liberation’s core was to be found not in the nationalist party anymore, but rather in the guerrilla units. Yet, its most profound meaning lies elsewhere. To fully appreciate it, the notion of 'politicization', which refers to a growing collective political consciousness (a phenomenon already characterizing the mukti bahinis) is, unfortunately, of little help. The government’s decision has to be understood, rather, as a critical assertion of the supremacy of politics over the military. In essence, this political move represents the mid-war shaping of a revolting and formless mass into a partisan liberation army devoted only to the nationalist cause of the AL. Importantly, it radically differs from the organization of a unified force prior to the conflict, a more common scenario among independence seeking countries.
20From what precedes, it is clear that the Indian Blitzkrieg of December 1971 and Mujib’s triumphant return after his release by Pakistan’s leader Z. A. Bhutto on 12th January, 1972, should not divert our attention from the serious political tensions that existed within the embryonic Bangladeshi army and from the fragile configuration of civil-military relations which prevailed in the days preceding independence. On the one hand, the interim cabinet was supported by a weak military command, increasingly dominated by Indian officers whose objective was to ensure an India-friendly AL government. On the other hand, its leadership was contested by officers, such as Colonel Taher, who stood as civilian-mobilizing agents, and aimed at imposing both their own strategies and post-war communist agendas. And Taher was personally difficult to ignore: the 'Taher bahini' was the first battalion to make its way toward the Pakistani army’s Dhaka Headquarters. Taher himself was given the post of Adjutant-General (the army’s number three) and the command of the Comilla brigade. His young friend, Major Ziauddin, obtained the Dhaka brigade’s command. These two were the most radical and well-organized units of the new army (Lifschultz 1979: 36, 38; Ahamed 1988: 86-87).
21Factionalism, party politics, and tensions between civilian and military sectors formed the unexpected ingredients of an implosive mixture: the nascent state was fundamentally opposed by its own army.
22The new military leadership – liberal, favouring a classic professional army, but still dominated by deeply anti-Indian officers – could have potentially hampered Prime Minister Mujib’s efforts toward a strategic alliance with India. It chose not to, most pressing issues had to be dealt with: factionalism and civil-military tensions born out of the war. It is precisely to solve them that Mujib tried to integrate the various groups into one single institution led, in addition, by a moderate freedom fighter.
23Indeed, Mujib, all the more suspicious of the mukti bahinis as he knew little about their functioning and division lines, appointed the newly promoted General Shafiullah, ex-commander of the S-Force, to the post of Chief of Army Staff (COAS). In consideration of candidates, General Ziaur Rahman (Zia) was looked upon with extreme suspicion because of his ambition and his divergences with the Higher Command during the war. By contrast, Khaled Musharraf was labelled pro-Indian, which left Shafiullah as the least questionable candidate. In addition to this, and so as to maintain a weak and divided army, the government politicized fresh inductees by promoting young civilian freedom fighters (Non-Commissioned Officers – NCOs) and commissioning new AL-leaning officers (Junior Commissioned Officers – JCOs).
24Civil-military relations at independence were also, undoubtedly, informed by the very structure of the new state. Factionalism within an army transformed, seemingly, into a 'spineless body' was not the only factor favourable to Mujib. The vacuum felt within the Bangladeshi bureaucracy after Pakistani officials left also facilitated the establishment of the 'AL rule'. There were simply no other legitimate institutions able to head the newly emancipated nation. The decay of the post-war civil administration was indeed very similar to that of the military bureaucracy. Consequently, in order to hold together his overpopulated country, wherein authority had already began to dissipate, Mujib had no choice but to build a partisan civilian, as well as military, administration officered by AL leaders whom he tried to control, not only as the head of the state, but mainly as the unchallenged chairman of the AL.
25Here lies the meaning of the expression 'Awami League Rule' used by its detractors: top-ranking civil servants, AL-leaning army officers, and state-owned companies’ directors, among others, did not swear allegiance to the state, they were rather using their positions in the party and relying on their freedom fighter backgrounds to acquire power, money and legitimacy. The problem with this model of political domination is that it depends, on the one hand, on a nationalist myth taking its ethos from an idealized Liberation war, and on the other hand, on a nationalist party which defines itself as the only winner of the war. This second tendency echoed the war-time political manipulations of the mukti bahinis by the exiled government, and aggravated the contempt most of the freedom fighter officers had had for corrupt politicians from the very beginning of the war. Moreover, the Mujib government, though glorifying the Liberation heroes, failed to define a clear mission for the post-Independence army in the nation-building process, and even in the defence of the country. According to S. Hussain, ‘this apathy was grown within civil and political society from the fact that the new state of Bangladesh apparently had no physical threat from its neighbour India’, a foreign policy orientation that was not shared by the majority of the freedom fighter officers (Hussain 2003: 7).
26Can it be said, in these circumstances, that Bangladesh was endowed with genuine statehood? Certainly not. The institutional machinery improvised after independence had neither rationality nor legality. The state, like the army, was paradoxically embryonic and degenerative, 'degenerative' in the sense that the state’s resources were siphoned by agents who were not serving the nation (though being civil servants) but were exploiting it through state-owned companies or other agencies. The term 'degenerative' is not to be understood in a 'eurocentric' sense, but rather according to the model of the 'parasite-state', developed by the Marxian trend 'Socialism or Barbary' in the 1950s to criticise Stalin’s USSR17. This paradigm, very relevant to study Bangladesh till the end of the 1970s at least, suggests that socialist states had in fact favoured the development of a 'state capitalism', a society of exploitation led by a new dominant class – the bureaucracy – in which the rulers of the state and economy replaced the erstwhile capitalists in retaining all means of production. Founded on such a 'parasitic' structure, the state was far from fulfilling the goals of the most idealistic among the freedom fighters, namely the social and economic emancipation of the Bengali people and the democratization of the polity.
27Finally, the AL was not – and had never been – a monolithic organization with respect to its hierarchy or its decision-making process. It was a mass party, very well rooted indeed, though divided into several factions led by charismatic and traditional leaders, and thus into a number of ideological trends – pro and anti-Indian, socialist and liberal, anti-imperialist and pro-American – with no real cohesion. Consequently, in order to govern this centrifugal 'League-State', as I term it, whose different components were demanding either their share of the pie or the enforcement of their ideologies, Mujib was compelled to consistently shrink the circle of the ruling elite and to rely on a system of patron-client relations extremely vulnerable to corruption.
28To keep this latent autocratic regime alive, the designation of 'internal enemies'18 was of critical importance, as it authorized regular and massive deployments of police and Bangladesh Rifles (BDR) for 'security operations'. Moreover, the early days’ political instability, marked by uncontrolled retribution violence against the remaining razakars and anti-government activities of underground leftist elements, combined with serious economic pressure, grew to such an extent that it soon led to a real regime’s crisis; the government decided, for the first time, to call the army “in [to the] aid of civil administration”. From 1972 to 1975, the government used armed forces in internal duties on five occasions, such as relief operations during natural disasters, and, already, maintenance of law and order in regions like the Chittagong Hill Tracts where pro-Pakistan armed elements were still active. According to S. Hussain, ‘these employments were counterproductive as frequent clashes with politically motivated forces (…) cemented 'military ego of superiority'’ (Hussain 2003: 7).
29The deep cleavages opposing the military to the AL inherited from the independence war were, thus, bound to reappear, and they did so in a much more damaging way given the peculiar political structure of the new state. The narrative of a former AL student leader, who fought in the 1971 war and then split up with the party, demonstrates it:
After the war, infrastructures were destroyed: For example, it was the case with military equipment and training institutes. In December  already, the mukti bahinis’ officers felt frustrated when [the] Indians took over all sophisticated materials left by the Pakistani army. After [the] independence, Mujib did nothing to give the army new equipment. […] When the Mujib government ordered the military to help the Bangladesh Rifles [BDR] clean the last pockets held by the razakars, they did it, and they succeeded. But when they saw that the BDR were taking arms and equipment for themselves, then fighting took place; soldiers and paramilitary started to shoot at each other.19
30From this period onwards, the regime was contested from within the army. The most radical officers criticized the 'AL Rule’s' very foundation, notably the way properties left by the Pakistanis and Hindus who took refuge in India were distributed among the elite, as well as the post-war lootings. But the most serious dissent came, initially, from men like Taher, Jalil and Ziauddin, who considered that the fundamental problem was the neglect the Liberation army suffered from. It was, in their eyes, the structure most apt to mobilize the nation’s human strengths: it had gathered Bengalis from all social classes and all political sides; it could have even been the catalyzing force to encourage economic and social development. But in abandoning the Liberation army, the government scuttled the expected revolution.
31To these officers, influenced both by Maoism and the experience of guerrilla warfare, the army was not conceived as an institution separated from the government, and its fighters were not cut out to join a classic professional army. Once independence was achieved, soldiers should have participated in industrial production, land reform and irrigation, and eventually in the establishment of socialism at the national level20. They envisioned both a popular and productive army (on the model of the Chinese Popular Army), which would closely work with civil society for the country’s development (Lifschultz 1979: 75-97). Jalil, Taher and Ziauddin were suspended for insubordination in 1972, and joined two clandestine Marxist parties: the Jatiyo Shomajtantrik Dol (Jashod, National Socialist Party) and the Shorbohara Dol (the 'Have-nots' Party), which were composed of AL dissident students and disenfranchised war veterans21. These parties started a struggle aiming at establishing social justice by eliminating corrupted civil servants and local maffiosi, as well as destabilizing the state through terrorist activities.
32To reduce the threat and to, again, counterbalance a doubtful army, Mujib founded his own militia, the Jatiyo Rakkhi Bahini (JRB)22. The JRB was much more than another law-enforcing agency: it represented, truly, the AL’s armed branch. It was designed to purge radical elements both within and outside of the party, and eventually to become the sole 'monopolizer' of legitimate violence in Bangladesh (Ahamed 1988: 51-54, Hakim 1998: 286)23. But after the creation of the JRB, considered by Taher as a concrete indicator of a police state turned against the revolutionaries, the Jashod too founded its own armed branch: Biplobi Gono Bahini (BGB, Revolutionary People’s Army). Taher’s military experience and his authority, drawn from his uncompromising attitude against the corrupted regime, explain the success of his movement. Though officially dismissed from his post, Taher maintained relations with his former brigade and could therefore organize clandestine revolutionary cells within the army. Civilians who had fought in the 11th sector that Taher had led, and upon whose revolutionary fibre he could rely, also joined the BGB’s ranks. An interface was even set up to provide activists and develop networking between the Jashod and the army: this was the Biplobi Shoinik Shangstha (BSS, Revolutionary Military Associations), modelled upon the pre-1917 Russian Soviets24. Yet, its main objective was to gather intelligence and to prepare a coup d’état.
33Under these circumstances, Mujib’s repressive measures were totally ineffective. Though an authoritarian leader, he persisted in neglecting the army, and this was precisely the root of the armed dissent he was facing. In fact, a structural reform of this institution and a redefinition of its relations with civil society were necessary to break the organic link between revolutionary parties and soldiers. This was all the more required after 1973, when the army confronted an additional internal tension, which would have even more decisive consequences on the civil-military delicate balance.
34The political importance of the tension opposing the 'Mukti Juddha'25 to the 'Repatriated' is forcefully recalled by a Major of the Bangladeshi army who quit his post:
35I think Mujib took a wrong decision. He repatriated those [Bengali] officers [from Pakistan] and included them in the army. That was a mistake. They could never be integrated into the army. Division started from this. And I still know officers, a Brigadier with whom I talked recently for example, who says that his batch mates are now Lieutenant-Generals in the Pakistani army. They come from the same batch; they keep strong relations. If Pakistani officers [after 1971] wanted to get information or if they wanted to do a favour to repatriated Bengali officers, it was not so difficult. They have good contact. Especially in 1973, the repatriated had very tight links with Pakistani officers. And I think the assassination of Mujib was initiated because of that decision. [Lower:] I think those people played a major role in the assassination.26
36The 'embryonic' character of the Bangladeshi army relates not only to its nascent nature or its strength but also to the very fact that this army was incomplete. More than half of its actual force was missing. In addition to 20,000 men who were present on the Bangladeshi soil in 1973, including no more than a hundred officers with genuine academic training (Hussain 2003: 6), there were still 28,000 Bengali soldiers and 1,100 officers who had not been serving in the East Bengal Regiments, did not participate in the Independence war and stayed in Pakistan’s western province. In the eyes of the AL government, this group raised even more problems than the freedom fighters did, because in contrast with the latter, the repatriated had never been subject to the nationalist party’s leadership. Nothing was less certain than the 'newcomers’' allegiance to Sheikh Mujib’s government, and there remained the possibility that old hostility toward politicians inherited from the Pakistani academic training could even resurface. The dilemma the nation’s father was facing may explain (it is a hypothesis) the delay between the end of the war and the beginning of the repatriation in September 1973, wherein Mujib had time to recruit partisan soldiers and create the JRB27.
37But the main reason as for why tensions immediately arose between the repatriated and the government was largely material. As mentioned in the introduction, Mujib offered a series of privileges to the freedom fighters after the war. Ex-mukti bahini officers received three to four years seniority as a reward of their services, as well as certain privileges for their families, and their induction into the civil service was also facilitated (Ahamed 1988: 52)28. Mujib did so for two reasons. First, he tried to assuage the tensions between civilians and the military that were born of the war. Second, he emulated the tactics of the war-time exiled government when politicians were made Generals, tampering with ranks in anticipation of the Bengali officers’ return. Nevertheless, unlike Tajuddin’s move during the war, these measures did not primarily reflect the supremacy of politics over the military. It was more of a defensive reflex. Mujib indeed promoted officers who, though not so close to his party, were glorified by their role in the Liberation war, something nobody could question or interpret as being subject to political manipulation. His aim here was to ensure that upon the completion of army staffing, those officers would continue to command the army in exchange for privileges that would win their loyalty.
38When the repatriated were reinserted into the army in 1973, an important cleavage arose which would have long-term effects. Yet, its causes were not only material (ranks, salaries) but also symbolic and political. The freedom fighters claimed that they were the ones who liberated the country, whereas the repatriated were accusing them of having compromised with India to win independence. This second polarization threatened the army’s cohesion on an even larger scale than the first opposition between the communists and the state. Despite the fact that the repatriated constituted the majority of the staff, they quickly understood that they would have little influence: the only way to gain any leverage within the army's ranks was, basically, to organize a coup d’état. Interestingly, the aim of such a coup, which would materialize in 1975, was not to interfere with politics (except for ending friendship with India, these officers had no political agenda), but rather to take over the army’s command. The Mujib government’s power tactics, which consisted of dividing the army to better manage it, could not have backfired in a more cruel way.
39The first coup d’état on 15th August, 1975, in which Mujib, his relatives and associates were killed, has often been described as a “leaderless coup”, as it was perpetrated by mid-ranking officers who acted without the endorsement of their superiors. I would rather define it as a putsch with multiple heads, which would later trigger serious instability both in the army and in the political arena, for the very reason that it gathered different actors with contradicting goals. The crucial variable to understand this coup lies with the opposition against the increasingly isolated 'Mujibist' clique, which temporarily created a convergence between the repatriated and those among the freedom fighters who were feeling threatened by the AL rule. Indeed, the operation was conducted by ex-mukti juddha officers who were made to retire by the government in the pretext of high-handed action during past operations “into the aid of civil power”. But the architects of this putsch were senior repatriated officers who acted in tandem with politicians of the AL’s right wing, such as the prominent Mushtaque. To normalize the situation, the army staff declared martial law and ordered the involved Majors to return to the army's headquarters, an order they did not follow. Rather, the Majors pushed Mushtaque into the President of the Republic’s seat and withdrew from the front stage while influencing decisions, most notably the nomination of a new COAS, Major-General Ziaur Rahman (Zia). Interestingly, Zia was not a repatriated but a former sector commander during the war; but he was the only officer of the staff who could conserve the balance of power between the two groups in the army. The most fascinating aspect of this coup was that neither the constitution, nor the Parliament, nor even the cabinet was dissolved. The objective was to evict the AL leadership, and to favour a politician who would take care of the repatriated.
40The 3rd November, 1975 countercoup logically followed. General Khaled Musharraf, pro-Mujib ex-freedom fighter, took over the main institutions in Dhaka and asked Zia to put the leaders of the 15th August putsch under arrest. As he did not abide by the orders, Zia was also arrested and Khaled proclaimed himself COAS. The scenario identically recurred: Khaled did not undertake the executive power and left the position of Chief Martial Law Administrator (CMLA) to a civilian. But, this was followed four days later by a third 'coup', which appreciably differed from the previous ones. It involved, this time, the well-organized Colonel Taher’s BGB forces. This was less an elitist coup than a vast mutiny which involved soldiers and young BSS officers whose aim was to set Zia free and restore him at the army’s command so that he could activate their revolutionaries’ agenda29. Taher, who had been evicted from the army in 1972, preferred not to appear in the forefront as his authority would be questioned among the repatriated.
41Now, it was Zia’s turn to demonstrate his profound understanding and tactful utilisation of the army's internal divisions. Used by the repatriated officers and the AL’s right wing first, and then by the BSS officers, to facilitate their ascent to power, he now exploited the BSS to evict the main AL leaders, and then turned toward the repatriated to bring the sepoy biplob (soldiers’ revolt) to heel. Zia managed to emerge as the army’s strong man by surfing on two consecutive waves of opponents to the 'AL Rule'. But then again, the executive power under martial law was left to a civilian, and Zia chose not to appear immediately as the head of state, even though he was pulling strings from behind the scene. The recurring characteristic in the 1975 coups and countercoups clearly shows that the position of CMLA was too vulnerable to be sought after, and in any case did not facilitate a close control of the army – the main objective of all the putsch-makers.
42But more importantly for the understanding of this article’s main argument, we see each coup as the strongest signs of the structural instability of the military – a hypothesis valid as well for all the mutinies, internal fights and desertions which will follow from 1975 to 1981. All revolved around two major issues: the influence of revolutionary ideas on the troops and the opposition between mukti juddha and the repatriated. It is precisely because of these two dimensions, and the factionalism they provoked, that none of the putsches aimed at establishing a military dictatorship, that is to say aimed at guaranteeing the interests of a corporation which did not exist in the first place. The different revolts that occurred in Bangladesh until 1981 are better understood as survival reflexes, designed to defend one’s own group against others. It was generally felt that if one of these groups dominates the state, then others would be threatened in their fundamental existence. This existential threat perception resulted from the officers’ diverging political trajectories, which implied very different roles for the army itself, and from Mujib’s effort to manipulate and divide the military establishment which prevented it from defining a clear mission for the institution. In brief, factionalism forbade the development of a corporatist interest, hence the possibility of a stable military dictatorship in Bangladesh. Herein lies the specificity of the Bangladeshi army.
43Prior to Zia’s takeover, the army was quite an incoherent structure with respect to its basis (factionalism of the mukti bahinis), its development (increase of manpower thanks to partisan recruits), and its achievement (divisions between freedom fighters and repatriated). This clear disunity was the primary source of instability within the army from 1975 to 1982. However, after General Ershad’s coup d’état in 1982, the chapter of revolts and putsches in Bangladeshi history appears to have reached its end. To fully understand today’s Bangladeshi army, our analysis now turns to the maturation process that occurred during this period of dictatorship.
44In the mid 1970s, the Bangladeshi political arena was in complete turmoil. We have shown that the situation was in part attributable to the army’s intervention. However, it has to be acknowledged that the demise of the multi-party system began well before. The Mujib government had already demonstrated its bent toward civil dictatorship by, among others policies, banning politico-religious parties, evicting the communists who rejected the Delhi-Moscow axis and, eventually, creating a one-party system. This system was based on a political organization which was no longer the AL, but a much more authoritarian core composed of Mujib’s relatives and close associates, the Bangladesh Krishak Sramik Awami League (People’s League of Bangladesh’s Peasants and Workers – BAKSAL)30. Most of the civil society’s organizations, which had made the 1971 mobilization possible, were considerably weakened31. This authoritarian drift yielded one serious consequence: the only genuine opposition to civil dictatorship came from the army.
45Under these circumstances, Zia’s pledge to legitimate his rule through the restoration of law and order was welcomed with apathy by a public opinion which in any case was no longer politically mobilized. The general’s first move was to extend this de-politicization process to the army itself: it became the prime target of purges. The BSS and Jashod leaders who had carried Zia to power were systematically arrested whereas the repatriated, less threatening because of their apparent disorganization, became the regime’s fresh allies. But for the soldiers who had experienced the Liberation war and 7th November 'revolution', Colonel Taher’s arrest, followed by his death sentence, was an act of treason. This triggered a series of unprecedented rebellions. Taher indeed became potentially more threatening, post-mortem, to Zia. The resentment brought forth from the execution resulted in a marked military instability, which inspired chain-reaction mutinies in several regiments. From November 1975 to October 1977, roughly ten serious revolts questioned the military hierarchy. Although none succeeded, all resulted in widespread bloodshed (Haque 1981: 191; Hakim 1998: 299)32.
46The aborted coup attempt of 2nd October, 1977, was undoubtedly the most serious of them. First, it occurred in the capital, Dhaka, whereas the other mutinies generally took place in peripheral units. Second, this coordinated operation of soldiers, NCOs and JCOs belonging to the army and the air force, clearly aimed at assassinating Zia and creating a soldiers’ revolutionary committee on the same pattern as the November 1975 sepoy biplob. But the most immediate incentive for this revolt was the avenging of Colonel Taher’s death, now raised by his men to the status of a martyred 'Liberation hero'. The coup, whose effects would last until January 1978, failed because of the loyalty of several top-ranking officers, like the Chief of General Staff of the Armed Forces Division, Major-General Abul Manzoor, an ex-freedom fighter who had fought with Zia in 1971, and the commander of the Dhaka brigade, Lieutenant-Colonel Ameen.
47The main source of instability was, unquestionably, to be found in the army: the issue was not about civil-military relations, but rather about army's internal relations. This is clearly attested by the fact that, within a few months of his takeover, Zia felt the necessity to found a special police force of 12,500 men, which strongly resembled Mujib’s JRB33. Because he could not rely on the army to help him hunt down the Jashod’s clandestine groups, he was, in a way, forced to create his own militarized police force. Zia’s uncomfortable position among his peers was further demonstrated by his decision to keep his COAS position until the June 1978 presidential election. And even after being elected president, he chose not to settle in the presidential palace but rather to stay at his COAS residence inside the army’s headquarters (Ahamed 1988: 125).
48Although the COAS constitutionally held the army’s command, power in the military was actually contested by several chiefs – another strong specificity of the Bangladeshi army at that time. In fact, each General Officer Commanding (GOC) was the only functional leader of his division. The Dhaka headquarters, and thus the COAS, was theoretically the supreme decision-making body with respect to nominations and promotions. But the different GOCs had comfortable latitude in managing their own personnel. In fact, during the Zia years, every time the higher command intervened in the divisional officers’ nomination process, rebellion burst out, sometimes triggered by the GOC himself. The coup d’état of 30th May, 1981, which led to Zia’s assassination, is a case in point34. It was triggered by the COAS’ decision to transfer Chittagong’s GOC who was 'ruling' his division as his own private 'game preserve' (Liberation war veterans were concentrated there and benefited from the highest posts). Once the menacing general had been asked to live, his men simply rejected their (theoretical) allegiance to the COAS and launched a coup. This was all the more ironical because strengthening the 'local chiefs’' power was a result of Zia’s will to ‘make Dhaka sure for him’, by placing most of the senior officers, especially mukti juddha, outside the capital (Franda 1982: 241-243).
49Though they eventually led to his own fall, the different measures taken by Zia to 'tame the tigers' are worth noting because they progressively professionalized and unified the army. They revitalized a certain type of authority in the military. An interviewed retired officer puts it this way: ‘Zia raised the army to a standard. Ershad did the rest’35. First, Zia significantly increased the army’s strength. After the October 1977 coup, five new divisions were raised and were equally distributed all over the territory (Ahamed 1988: 110). The result was impressive: in 1981, mukti juddha and repatriated soldiers amounted to respectively 15 and 25% of the total strength; the remaining 60% were the product of Zia’s expansionist policy (Ahamed 1988: 117). His aim was to superpose a fresh apolitical recruitment to the ideological factions that had existed since 1971, so that the paternalistic relationship between commanders and their men would be hampered. This 'de-ideologization' of the army was also visible in his effort to emulate K. Menon’s plan for the Indian army and diffuse power among armed forces through a system of frequent transfers of highly ranked officers (Ahamed 1988: 112).
50Secondly, Zia launched several initiatives so as to modernize the army. The fresh inductees received a new type of academic training based on the principles of soldiers’ responsibility vis-à-vis their institution and constitutionalism; the training also promoted a modified version of history stressing a continuity with Pakistan while India’s role in Bangladesh’s Liberation was sharply abridged. The 'batman system', that allowed officers to use privates as their servants, was also abolished in 1977. Both officers and soldiers’ wages were increased at a faster pace than inflation, which attracted new volunteers. Re-evaluation of rents in the cantonment was beneficial to the privates. Eventually, Zia also offered facilities to retired officers so as to join lucrative jobs in the private sector or to start their own businesses after retirement (Ahamed 1988: 129). No need to say that these modernizing efforts were made possible thanks to a substantial rise in the military budget: from 13% of public expenditures in 1974-1975, it rose to32 % the next year, and continued to regularly increase every year. From 1972 to 1981, military expenditures grew by 186% (Huque and Akhter 1989: 183).
51Zia’s behaviour towards officers solidified his reputation as the man who re-established discipline within the army, a vision still present in contemporary public opinion:
There was a time when discipline was much more serious in the regiments. […] What used to happen under Zia? The officer gives you an order, you do it, you ask question later. If you do not do it, the President will kill you.36.
52Zia may be considered as a moderate dictator in the civilian aspects of his rule – he is often said to have restored the multi-party system that Mujib suspended – but he left a lasting impression in the military by reintroducing a practice which had been done away with since the 1857 Sepoy mutiny: the systematic execution of rebel officers. For instance, after the October 1977 aborted coup, he declared that out of the 460 officers and privates who had been brought to court, only 63 were acquitted37. The air force officers, who had the largest number of mukti juddha, were the first targets of Zia’s disciplinary violence and never truly recovered. After the trials, only 11 officers were left in this corps (Franda 1982: 243). Subsequent to these purges, but also in alignment with the national strategic vision, the army still towers over the other two corps to this day.
53Yet, despite these measures, Zia did not manage to make the military a safe constituency to back his rule. It is through a civilianization of the regime that he drew, out of the electorate, the legitimacy he was lacking in the army. He established, for instance, a system of local representative bodies very similar to Ayub Khan’s 'Basic Democracy' and also managed to develop a successful populist and religiously coloured nationalist discourse. He allowed several polls to be organized (referendum, presidential and parliamentary elections), lifted the ban on politico-religious parties, most notably on the Jamaat-e-Islami Bangladesh (JIB), and eventually founded his own organization, the Bangladesh Nationalist Party (BNP). These moves drastically restructured the political landscape as the BNP attracted anti-Mujib AL activists, ex-'collaborators' as well as radical leftists (who gave up both ideology and their life underground). They would also have a profound impact on the country’s political future: from now on, and up to the present, political parties no longer consist in ideology-based organizations but rather act as extremely personalized power instruments, which work on patriarchal and patron-client-relationship patterns; the only notable exception being the JIB.
54General Ershad’s coup d’état on 24th March, 1982, was the last of a long list. Therefore, our main objective now is to understand the reasons why the army eventually stabilized in the 1980s. The general concept of stabilization encompasses three related processes we will try to explain in this part: 1/ the army’s institutionalization, i.e. its managing a defined space in the institutions of the state (in which we will see that Ershad’s attempts at consistently militarizing the administration turned out to be a failure, as they were stopped by the opposition parties, and eventually renounced by top-ranking officers themselves); 2/ growing professionalism in its ranks, which means that the military hierarchy no longer feels it has a political role to play in the nation’s destiny, and sticks to its role of managing a defence-oriented force (here ambition and greed still remain, but are limited to a quest for prestige, power and money within the army); and 3/ the sociological changes the rank and file went through in the 1980s. The hypothesis of the army’s stabilization will allow us, in turn, to explain the army’s behaviour at the height of the 1990 movement for democracy and its acceptance of the pre-eminence of civil power.
55Ershad’s putsch differed from previous coups in many respects. It occurred after several months of political disorder during which Justice Abdus Sattar was acting President though the actual power lay in the hands of the COAS, General Ershad, who had been appointed by Zia in 1978. Considered then as unambitious, he had kept a low profile and won Zia’s trust. However, Ershad stood as the repatriated officers’ leader and, after Zia’s death, became very vocal about political issues. He regularly intervened, for instance, to emphasize the need for a constitutionally defined role for the military in state institutions. A few lines of an interview he gave to the New York Times are worth reporting here:
Insha’Allah there will be no more coups, but what you have to do, is to find constitutional solutions so that no coup and no killing occur after one year, ten years, or ever again… Our army is an efficient body, well disciplined and most honest, it is made of organized forces which are devoted to the nation. The potential of such an exceptional force (…) can be genuinely used for productive and nation-building purposes, in addition to its role in national defence.38
56The above statement is all the more interesting in that it points out the Bangladeshi army’s maturation during Zia’s rule. The military is described as a unified corps, confident in itself and its potential. Ershad goes to the extent of using Taher’s concept of productive army, though with a very different meaning. According to Ershad, the rationale for unending revolts in the army lies with the fact that the latter was not given the place it deserved in the state apparatus.
57Ershad’s coup is distinct from any previous one because it eventually legitimates the establishment of a military dictatorship as such. First, Ershad immediately occupied the position of CMLA, i.e. the head of state in this case, in order to act as fast as possible, even if it meant alienating the political parties from the opposition. Second, he extended the militarization of civil institutions, which had started under Zia, to the Secretary level39. The obvious aim of calling officers and soldiers out of the cantonments and posting them in the administration was to get rid of the military’s threat. Its major consequence was that the civil administration became as compartmentalized as any part of the army’s structure, with an extreme concentration of power in the Presidency (Hasanuzzaman 1991: 41). Thirdly, instead of stigmatizing the army’s indiscipline, Ershad entrusted it with establishing a new moral and political order. Neither JRB nor special police force was needed: the military was directly in charge of internal security. ‘The district-level police administration was so largely militarized that, for a particular period, Superintendents of Police in 53 districts out of 64 were army officers’, as Hakim points out (Hakim 1998: 289)40.
58But these concessions given to the military cannot as such explain the stabilization of the army during the 1980s. The critical sociological changes it had gone through since the mid-1970s contributed, in a more decisive way, to pacify its ranks. The systematic eviction of freedom fighters from the army staff brought forth by the 15th August, 1975, coup and achieved after Zia’s assassination (30th May, 1981) had profoundly transformed the leadership of the institution. At the time of Zia’s death, 3 out of 50 Brigadiers and Major Generals had actually participated in the Independence war. The army staff and the divisional command were fully controlled by the repatriated officers. Their political and military culture, which profoundly differed from the veterans’ as it was much closer to the Pakistani military ethos, is one of the reasons why order was restored in the army. Politically conservative, ideologically anti-communist, they did not have the ambition to change society in a significant manner. And in terms of military culture, they envisioned a pyramidal and very hierarchical army structure which drastically differed from the 'egalitarian' conception of many ex-freedom fighters.
59However, the most significant change the army had undergone was the introduction of a new generation of officers who were commissioned after 1975 into mid-ranking posts. This young generation never experienced the mukti bahinis’ type of organization or the guerrilla warfare. By contrast, it had been 'de-ideologized': its nationalism was no longer directed towards a social revolution, nor was it based on the cooperation between an idealized Liberation army and civil society’s actors. The 'Zia generation’s' nationalism aimed at preserving Bangladeshi sovereignty, both militarily (defending its borders against India) and culturally by underlining the Islamic character of the nation. Under Zia and Ershad, 'Bengali' nationalism was thus transformed into 'Bangladeshi' nationalism, in order to distinguish the Muslim-majority Bangladeshi citizens from their Indian neighbours in West Bengal. Yet, this shift did not reflect an ideological change in favour of political Islam (the Islamization of the state was in fact very limited), but rather expressed an increased pragmatism due to a perceived Indian threat.
60There is another important variable which accounts for the stabilization of the army in the 1980s. In parallel with the growing role of its officers in the state apparatus, the 1980s army as an institution continuously turned its 'raison d’être' outward, i.e. toward the 'Indian threat'. This actually reduced the incentive for an intervention in domestic politics (Shamim 1994: 351-366).41 The changes in the national defence strategy, implemented by Zia, were confirmed under Ershad’s rule: the Bangladeshi army assumed a mainly defensive role while directing its strength towards a possible Indian attack. This 'realist' behaviour was, in fact, the military response to growing tensions with India regarding foreign policy and water sharing, among other issues. This is partly due to the fact that, according to a retired civil servant and diplomat, military instruction textbooks had not changed substantially from the time Bangladesh was still a part of Pakistan, and India was still considered as the prime enemy.42 Another ex-officer explains hostility towards India by the cultural factor, yet links it to more mundane causes:
It is traditional. It is the philosophy we have been taught. And here, religion has an important place. For example, our war cry is: “Allahu akhbar!” And it is not only the case in the army, anti-Indian feelings are well-spread among civilians, and it has diffused more and more in the officers’ ranks. […] And there is the insurrection in the Chittagong Hill Tracts [CHT]. 60 % of all armed forces are constantly deployed in the CHT. In a way, 100 % of the soldiers have served at least once in the CHT. And everybody knows this conflict started because of India. So if you are posted there, one way or another you will have to hate India. Another reason may be the regular crossfire with the [Indian] Border Security Force.43
61Finally, a critical factor in the army’s growing professionalism lies with the participation in United Nations’ peace-keeping operations, which started under Ershad’s rule. The first of these missions was the UNIIMOG (United Nations Iraq-Iran Military Observer Group) in 1988. The performance of Bangladeshi troops during that operation was recognized, and allowed a Brigadier to be appointed chief military observer. From 1988 to 1997, they participated in eight missions, including UNIKOM in Kuwait and UNMIH in Haiti. In 1998, Bangladesh was the third largest provider of UN troops. Given the significance and the frequency of its participation, the UN even suggested the creation of a special reserve battalion formed of Bangladeshi troops which could be called at any time (Ali 1998: 12). Apart from the prestige those operations brought to the army as an institution, the financial benefits individuals could draw from it were an important factor in weakening political ambitions among officers and social demands from the soldiers. The following comment by a retired officer highlights this:
Under his rule, Ershad decided to send troops in United Nations operations. It was quite a good thing, but not really a military decision. Officers began to earn money. And when they saw what dollars can do in Bangladesh, all of a sudden the whole army changed. People started to look for money, in a way we became more selfish. I think this is the major change in the last years.44
62The new configuration of the army and the de-politicization of its troops were also the result of the legitimizing strategy that Zia and Ershad implemented to strengthen their regime. In order to rule, they had to make alliances with one or more social groups who would allow them to get rid of the previous 'Mujibist' elites. The two General-Presidents therefore tried to build a coalition of interests comprised of the army, their first support base (though a more fragile one in the case of Zia), the bureaucracy, and finally the upper middle class businessmen and industrialists (Nuruzzaman 2004: 43). These networks progressively formed what we might term a 'ruling class' and developed a common interests in terms of government policies: doing away with socialism, while paradoxically encouraging protection of domestic industries, refusing Indian hegemony and strengthening ties with the United States, Pakistan and other Muslim countries, etc.. In this configuration, the army held a place and a prestige it never had before. Officers also got new opportunities and access to power thanks to their relations with the business community and the political elite.
63Yet, the army’s relationship with civil society was, in the late 1980s, very different from the quasi-fusion that had prevailed during the 1971 War. After the joint struggle of the army and civil society for the achievement of a revolutionary ideal and in the nation-building process, the military cut ties with civil society. It fortified its relations with the ruling elite and satisfied its corporative interests, yet made little effort in making inroads in society at large, neither at an ideological level nor at a policy level. Except on two issues – the history of the Liberation war and the insurrection in the Chittagong Hill Tracts – the military in power did not make any significant effort to limit the freedom of press, reshape the schools’ textbooks, or curb political activities in universities, for instance. Zia and Ershad were cunning politicians, more than sheer military dictators, who managed, through relatively non-repressive means, to evict the groups supporting Mujib’s socialist policy, while making alliances with powerful Bangladeshi families and creating ties between the army and the business community.
64The opposition parties were left, therefore, with a clear field. As a consequence, and contrary to Zia’s time, the instability which affected Ershad’s government, and would eventually lead to his demise and to civilian rule in 1991, came from the political parties, and not from the army. The lift of martial law in 1986 triggered the first protests and allowed opposition parties to mobilize outside their activists’ circles while giving them a better visibility. The old tactic of hartal, which nurtured the nationalist movement in the 1950s and 1960s, was reintroduced in public space. Mixing two types of collective action – demonstrations and general strike – it aimed at economically paralyzing the economy and bringing the government to heel. Yet, the protest movement was divided. Four poles were leading it: the Left, led by the pro-Moscow Communist Party, which had been fractioned and weakened by the creation of BNP; an alliance led by the AL under the leadership of Sheikh Hasina, Mujib’s daughter; a coalition of liberal and conservative parties led by the BNP and its new chairperson, Khaleda Zia, Zia’s widow; and finally, the JIB.45 These four blocks eventually forged a grand alliance in November 1990, only one month before Ershad resigned.
65To remove the opposition’s threat, Ershad used repressive measures – but, interestingly, police and paramilitary troops, not the army, were handling law and order management. According to several sources, the reason why the army was not called in to deal with internal security was both the diminished support Ershad enjoyed within the military and the pressure exerted by foreign donors.46 The General-President also tried to bring dissent to an end through political manoeuvring. He founded his own organization, the Jatiyo Party (JP), and sought, like Zia, to absorb opposition party leaders (this worked as far as AL leaders were concerned). Successive polls (referendum and presidential elections in 1985, parliamentary elections in 1986 and 1988) were also held. Massively rigged – the JP controlled the Parliament – they nevertheless helped in civilianizing the regime.
66Finally, when the situation became too critical, Ershad also emulated the opposition parties’ clashing tactics by arming and criminalizing his party’s student branch and syndicates. This led to the 1987 and 1990 pitched battles on the campus of Dhaka University, the epicentre of protest. This tactic eventually proved how out of steam Ershad’s government had run. The Bangladeshi state had slipped so far from Weber’s ideal type of legitimate violence monopoly that the government was compelled to activate groups which would have normally threatened its survival simply in order to maintain itself. These armed gangs, some of whose leaders were given positions in Ershad’s party, were used to confront opposition parties in the street, and were 'rewarded' by controlling commercial areas and exercising protection racket on them. As it had already happened during and after the war of independence, though on a wider scale, this privatization of legitimate violence started a process of criminalization of politics, which continues up to the present day.
67More decisively for the short-term, Ershad’s strategy proved totally counterproductive, for it obviously implied more public disorder and insecurity. As the government was playing the opposition’s confrontational rules, law and order deteriorated and hartals critically destabilized the economy, which was the opposition parties’ ultimate goal. Therefore, at the apex of the street protests, two of the regime’s three pillars – the bureaucracy and the upper middle class – withdrew their support, as Ershad was no longer able to guarantee a stable economic environment or the aid influx.
68Ershad’s resignation in December 1990 did not put a final end to the triadic configuration of the state – the bureaucracy, the upper middle class, and the army –, functioning on patrimonial and patron-client based relations. As we shall see, the democratization process rather encouraged its perpetuation.
69In the last days of Ershad’s government, political instability could have easily justified, in the military’s eyes, a direct return to power through a coup d’état and the imposition of martial law, as was done during the 1975 and 1982 crises. For the first time, however, the opposite happened. Here we examine the factors that prompted the army to abandon Ershad and to let the democratization process begin.
70A first explanation of the 1991 transition from military to democratic rule, derived from the rational choice theory, lies with the cost-benefit ratio that ranking officers take into account while considering a possible takeover. This means that an inner putsch would have looked like a step backward, which was unthinkable in such a politicized civil society as was the Bangladeshi one in the late 1980s, notably because opposition parties, which eventually united, had recovered their mobilizing power and were controlling the street. This scenario was all the more costly because the army had already ceased to be the main agent of repression against opposition parties; the armed cadres of the JP were assuming this role. Moreover, this would have also meant, in any case, to increase the level of repression and, given the extreme polarization of the political environment, different intelligence agencies, in particular the Directorate General of Forces’ Intelligence (DGFI), had already expressed great reluctance to do so, aware of the serious civil confrontations it could lead to. Moreover, Western embassies in Dhaka made it clear that they would not support another military junta (Hakim 1998: 292).
71But another, and more important, reason accounts for the military’s withdrawal: the nature of its corporative interests. The continuation of a regime stained by corruption was also dangerous to the army as an institution; it threatened its prestige and unity47. The opinion of an officer, who was pushed to retire in 1998, illustrates this hypothesis:
Ershad understood that mid-ranking officers must be controlled, because they are the men who directly command troops. For instance, the August 15, 1975, coup had been organized by such officers. So, he was making some of them millionaires, whereas others only received their usual salaries. Or, if officers were becoming dangerous, he used to punish them by appointing them to degrading positions. Every man who was not corrupt began to think that Ershad was destroying the army’s prestige.48
72This quotation shows that corruption among the military was an individual corruption, wherein each officer aimed to personally get rich to the detriment of the institution. A comparison with the situation in the Pakistani army might be helpful here. Through the distribution of lands and military foundations, which provided numerous retired officers and their families with social services, veterans collectively benefited from the rent-seeking system, during years of military rule and even after the return of civilians to power in the late 1980s (Blom 2005: 47-49). As S. Wilkinson also points out, ‘as collective enterprises, [officers’] growing role in the economy does not have the same damageable effects to the military discipline as when [they] individually seek to get rich, as they did in countries like Indonesia and Bangladesh’ (Wilkinson 2000: 216). In 1997, for example, it was the army as an institution, which asked the President to dismiss the Chief of the navy, Admiral Mansurul Haque, with charges of corruption, because he had personally grown richer. In the case of Bangladesh, the system based on the coalition of interests we have previously described, did not allow the army to collectively benefit from economic growth, and did not even put officers on par with one another. It was therefore expected that, at one point, the 'losers' would eventually refuse to let that system last. They just used the pretext of a loss of prestige to abandon Ershad. As a matter of fact, the dramatic drop of foreign aid at the end of the 1980s triggered the withdrawal of both these officers as well as the privileged ones.
73A third cause of the army’s withdrawal from power, structural as well, can be found in the political instability that was tearing the country apart. In the officers’ minds (particularly the eldest), the disorder which followed eight years of peace revived the memories of the army’s bloodsheds of 1975, 1977, and 1981. For instance, during the 7th November, 1975 sepoy biplob, the COAS, 67 officers from the army and the navy as well as about 2,000 listed men were slaughtered by low-ranking privates. Even though the army had been ideologically reunified by Ershad, these bloody memories, in addition to internal tensions coming from an inequitable system of corruption, and to a political environment of opposition parties prepared for civil revolt49 were acting as a repellent.
74These memories, however, did not cast any doubt on the tragic state of that situation. Consequently, Lieutenant-General Nuruddin Khan, appointed as COAS in September 1990, contacted leaders of both the AL and the BNP, and informed President Ershad that he was no longer able to support him (Ved 1991: 1272). Zillur R. Khan gives details about Ershad’s surrender: ‘While withdrawing their support from Ershad and making possible his subsequent arrest, military rulers tried to keep their distances from the President, thereby using him as a scapegoat for the political and economic corruption of the nine years of military rule’ (Khan 1997: 587). This decision paved the way for power transfer to a caretaker government led by Chief Justice Shahabuddin Ahmed. The army had officially withdrawn from political stage. Nevertheless, because of this neat and unconditional withdrawal, the democratization process did not lead to an institutionalization of relations between civil power and the army. The regime reforms focused on relations among political parties in decision-making institutions, most notably the Parliament, on the modalities of access to power, and on the way to guarantee a peaceful transition in every future election.
75Here again, the comparison with Pakistan offers valuable hints. When the military 'left' politics in 1985, the transfer that took place through the establishment of a civilian government looked rather like power sharing between politicians and the military elite. Indeed, General Zia-ul-Haq, who clearly represented the army’s interests, remained President. Even though the situation in Bangladesh apparently retained similarities with respect to the civilianization process, the main and critical difference is that Ershad was both the army’s representative and the chairman of the party he founded (the JP), whereas in Pakistan, Zia-ul-Haq entrusted the Muslim League’s leadership to a civilian, Muhammad Khan Junejo. Moreover, before lifting martial law, he had the 8th constitutional amendment voted, which allowed the President to dismiss the central and provincial governments and dissolve the Parliament. Given Pakistan’s civil-military imbalance and the pillar role that the COAS played, it was implicitly understood that the President, be he a civilian or a general, would not do anything without the COAS’ consent.
76Therefore, the 'balance of power' Zia-ul-Haq had put in place could carry on even after his death; it was this set of guarantees that made the military elite accept the appointment of a civilian President and the organization of elections in 1988. Indeed, even if the government was supposed to represent the democratically elected majority’s interests, the President remained the spokesman of extra-parliamentary forces, the army in particular, and could use the 8th amendment as control means towards the civilian government. As M. Waseem explains, ‘it provided [the military] with a strategic space within political power, even after they formally withdrew from the game’ (Waseem 2000: 154). The dependency of the Prime Minister on the President restored the old colonial system of diarchy - executive powers were disputed between the head of the provincial government and the Governor. This meant that when issues arose between the two heads of the executive (head of the state and head of the government), it was the COAS who solved the crisis, thus becoming the referee in the political arena, as it happened when President Ishaq dismissed Nawaz Sharif in 1993. When the latter resorted to the Supreme Court which eventually gave reason to his plight, the COAS himself imposed a compromise by asking the two men to resign. Similarly in the 1997 judiciary-executive tussle, Jehangir Karamat, the then COAS, intervened to ask President Leghari to resign to ensure stability.
77The Bangladeshi scenario is completely different. The non-institutionalized nature of civil-military relations did not lead to an implicit army’s rule, as in Pakistan, after the restoration of democracy. The 12th constitutional amendment, which was voted soon after the first democratic elections in February 1991, restored the parliamentary system and the President lost the prerogatives he had had since 1975. The diarchy found in Pakistan was thus rendered impossible, which allowed Bangladesh to avoid the numerous political crises and the ministerial instability which dramatically weakened democratization in the former. Moreover, the Bangladeshi Prime Minister (PM) was entrusted with the defence portfolio, giving him or her actual power on the armed forces. The representative of both parliamentary and extra-parliamentary forces was embedded in the same person, which implied that the military’s influence on politics, when it was found, was directly exercised through bargains with the PM, and not through the COAS’ arbitration like in Pakistan50. In Pakistan, though the defence portfolio is vested in the Prime Minister, the appointment of the chiefs of Armed forces is vested with the President under the 8th amendment.
78But the key variable, and major difference with Pakistan, lies with the status, and stature, of the COAS, and with the command structure of the army. In Pakistan indeed, the COAS and the 9 Corps Commanders, among others, play a key role in that structure, which allows the establishment of a 'junta' that somehow controls the civilian rulers. By contrast, in post-1991 Bangladesh, when conflicts arose between the civilians and the military, it turned at the advantage of the government, which used the COAS as a fuse wire51. For example, the dismissal of Lieutenant-General Nasim, ordered by President Abdur Rahman Biswas in May 1996, created an indisputable precedent. A retired top-ranking officer analyzes its crucial consequence on the civil-military balance as follows:
At that moment, the President wanted to sack two officers. The COAS refused. But we [the Director General of military intelligence and I] did not back the COAS. So, the civilian government won confidence about the fact that the government can dismiss a COAS without losing the support of the rest of the army. (…) It has become common now. (…) The chiefs of the navy, of the air force, in fact the three chiefs were sacked by Khaleda Zia [in 2001], who appointed other chiefs. Many generals who had got promotions under Sheikh Hasina [from 1996 to 2001] were retired without any reason. Now, the civilian government is more confident, because in 1996, we sided with the government to protect its interests. So [the civilians] think they can retire any commanding officer at any time. And this is what Khaleda Zia has done [from 2001], very confidently. And nobody resisted. On the contrary, now it has become so critical that officers are happy: whenever a post falls vacant they think they will be promoted to that position. It is selfish interests that work. Patriotism has gone. It is real facts I am telling you here. We are not in Pakistan, where Musharraf takes over power and several units surround the official buildings, where it is claimed on the radio that the COAS took control of the situation… People are very selfish and egocentric here.52
79And truly, the only time a PM actually dismissed a COAS in Pakistan’s history, and though it was his constitutional right, it was him who was immediately made to resign from power by the COAS – as Nawaz Sharif experienced in October 1999. This was possible because the COAS could count, contrary to the PM, on the military hierarchy’s support against any attempt by a civilian to interfere in army’s posting, especially at such a high level. It needs to be recollected that the Pakistan Army had not taken lightly an earlier incident when Army Chief Jehangir Karamat was forced to resign, as he proposed the constitution of NSC in the Naval War College. But in Bangladesh, as the interviewed officer underlined, the PM could very confidently direct change in the military hierarchy. Yet, as proved by the contrasting example in Pakistan, this confidence may be found only if a relation of allegiance exists between the military elite and the government. So, what are the determining factors in forging this allegiance? In other words, how is the staff in general, and the COAS in particular, chosen to make this link possible?
80Let us take the example of the access to the Prime Minister’s Office (PMO). Given the absolute centralization of the Bangladeshi political system, this constitutes an essential attribute of power for both military and civil affairs. The following excerpt from an interview with a retired officer will help to understand how it precisely works:
- Presently, who are the main generals who have access to the PMO?
- The Director General of DGFI, the Director General of Special Security Forces, the Commanding Officer of the Presidential Guard Regiment, the COAS, the Brigadier Commanding the 46th Brigade, based in Dhaka, the General Officer Commanding [GOC] the 9th Infantry Division in Savar, the GOC of Comilla… I believe all of them have access to the PMO.
- Is it institutionalized?
- Yes and no. It is institutional, because those are vital positions for the government. At those posts officers may do something, so they have to be kept under control. It is also informal, because those who are there, in a way or the other, they have good relations with the government. Nobody can sit in those positions unless he has good relations. For instance, it may be one of the PM’s cousins, or maybe the PM knows them personally, or they stand loyal to the PM… There must be a relation between them.53
81Here again, nothing is institutionalized, and the Bangladeshi political culture’s 'repertoires of action' – highly patrimonial – usually take precedence over the rules defined by hierarchy, seniority or professional experience. The military elite’s loyalty is not truly directed toward the state, and even less toward civilian power, but toward the Prime Minister as an individual who heads a kinship network. The most striking example lies with the nomination of a new COAS by Prime Minister Sheikh Hasina in December 1997, who was none other than her own uncle, Lieutenant-General Mustafizur Rahman54. It is personal links of allegiance between two central figures of the political system which determine the relations between civil administration and the army as institutions, rather than the contrary. Given this patrimonial structure, it is not very surprising that the army was so easily re-politicized during democratic rule.
82Finally, in contrast with the case in Bangladesh, the Pakistani Army has defined a sphere of interest – foreign policy and security. Pakistan’s India policy, Afghanistan policy and its nuclear policy are completely in the army’s control. This is one of the reasons why, when the transfer to a troika system took place in 1988, Benazir Bhutto was not given oath of office for fifteen days till she conceded on these issues. The Bangladeshi army does not have much role in its foreign policy. Its perception of Indian threat is not as deep and essential as the Pakistani army’s. The circumstances of state formation in the cases of Pakistan and Bangladesh have defined the armed forces’ role vis-à-vis India. Pakistan’s army’s role has been linked to the survival of the Pakistani state, which has not been the case with the Bangladeshi army.
83Different interviewed officers evoke the 'sense of insecurity' felt by political parties with respect to the army from 1991 onwards, which may be considered as paradoxical in light of the fact that the military had withdrawn from politics. The following excerpt helps us to understand the causes of such a sentiment:
- Today, civilians have this sense of insecurity which makes them want to control the defence sector.
- Do they think they could be overthrown?
- They believe the army is the most influential actor when the question of overthrowing the government arises.55
84This perception that the army could still pull the strings behind the scene, though it no longer exercised political power, is an obvious legacy of its interventionist past. The constant links between the army and a particular leader (Taher, Zia or Ershad) or a political or ideological faction (freedom fighters, repatriated…), which lay at the roots of factionalism, explained also the 1991 politicians’ defiance vis-à-vis the army.
85Now, political leaders paradoxically answered to their own fears by awakening factional divisions. The sole objective was to 're-appropriate' the army. Khaleda Zia, who headed the government after the BNP won the 1991 election, played a key role in this process. She strived at all costs to appear as the new leader of the military institution. She first sought to weaken 'Ershad’s army' by retiring, or sending as ambassadors, the oldest and/or most involved officers of the previous regime. Then, in order to reduce the influence of Ershad and his partisans she could not evict, she tried to discredit them by leaking out information about organized corruption under Ershad, as well as criminal activities he had allegedly sponsored. This initial attitude of civilian power would bear heavy consequences, as exposed by a former officer:
After the fall of Ershad, we received official white papers about his corruption. A doctor was killed during the 1990 uprising, his name was Dr. Milon. The killer, his name is Povi, was a student leader. He got ammunitions from Rowshona Ershad [his wife]. She gave it to him in a canteen of the Cantonment Military Hospital. All those things were released as official papers and were sent to all units. Officers read it before soldiers. The planning of the 1982-1983 operations [to make the coup d’état successful] was disclosed. These are secret things; they should have never gotten out. When those things were known, there were obviously some officers who were more enthusiastic than others, who wanted to be in the new government’s good books, and who were thinking: “Okay, if this information is disseminated in the army, Ershad will be weakened”. I am just giving you an example. Whoever did it, and I don’t know who, obviously he made some enemies. He did it against the rules. So the rules were violated and the government accepted it. It is how it slowly started.56
86In parallel with this move, Khaleda Zia played the score of her defunct husband’s prestige; though terribly harsh with the military, his memory still enjoyed the aura of the righteous and honest freedom fighter; some had even raised him to the status of a martyr after his assassination. To help her in this strategy, a critical figure was Major (retd.) Iskandar, Khaleda’s own brother, who gave up his military carrier in the mid-1970s and became one of Zia’s closest collaborators. Iskandar personally knew every officer who had been commissioned and promoted by his brother-in-law, and served as intermediary between the PM and the staff in order to give these officers the most important positions (in today’s army, though Iskandar has retired long ago, an 'Iskandar’s Brigade' still exists). Though exercised outside the institution, his influence was widespread:
From 1991, Major Iskandar had been controlling everything in terms of promotions in the army. It was what we used to discuss in our ranks, but we were junior officers (…). He got retired in 1976, or even earlier. His batch mates are highly placed today, they are Generals or Brigadiers, and he knows them personally. He can tell them: “I post you there, and you take care of me”. And I know he is my contact with the political party. If I take care of him, he will take care of me.57
87‘Taking care of one another’, as this retired officer puts it, was and still is at the bottom of the relation between not only the military elite and the political parties, but also between these two and the business community. The army is the main buyer of national industries and a significant real estate investor, through the Defence Officers Building Houses Societies. A general never directly supports a political party, but facilitates the settling of contracts between the Armed Forces Division and a company, the latter pouring a commission back to the party. While priding himself on building a new warehouse or a new mosque in his unit, which is commonly understood as him helping the party in power, an officer might be favoured at the time when promotions are decided. But the patron-client relationship goes further. When this officer gets retired, if he does not possess enough wealth and influence to directly obtain the investiture of the party, the latter will still find for him a leading position in a structure profitable enough to 'subsidize' the party and win its ticket in the next election.
It is a chain: you helped me when you were in the army; after your retirement, I will help you. (…) For example, the chairman of Red Crescent, honestly… It is very honourable in a sense, but it is all window dressing to make money. The Red Crescent does nothing here, it just steals money. So, you stay five years chairman, you give enough money to the party, and you ask your nomination for the election. It is like that: put them where they can make money, a part of this money will drop in the party’s funds. There are calculations: for example someone believes the Awami League will win the next election. So he stays low-key, and when the AL wins, he serves it; after 5 years, he can expect a good position. It is long-term investment!58
88By the time the AL came back to power in 1996, another personal and 'clientelist-based' dividing line was added. This new faction was composed of former officers from the Jatiya Rakkhi Bahini (JRB) who had been reintegrated in the army after this force had been dissolved. Ostracized under Ershad and neglected during Khaleda Zia’s government (1991-1996), they obtained promotions and commanding positions after 1996. But in contrast with the 'Iskandar’s Brigade’s' officers, deeply-rooted and in good terms with the last elements of the old generation from the Pakistan Academy, the JRB officers were subject to strong segregation by the rest of the military elite. Negatively considered because of their non-military past, they were, and continue to be, blamed for being the main beneficiaries of the changes in the military hierarchy after 1996.
89The mechanisms of the post-91 factionalism are very similar to the 'tradition' inherited from the Independence war and strengthened afterwards. Because the three consecutive cabinets felt insecure with the partisan features of the army, every one of them did whatever it could to divert its rules and to promote mid-ranking officers in order to forge a circle of friendly generals, either united by their political past (like the ex-JRBs or ex-mukti juddha) or by their common batch (in the case of Iskandar’s Brigade). The central question to ask, then, is whether the army’s politicization became a factor of political instability or not. To answer it, a close examination at the politico-military crisis of May 1996 is relevant.
90Soon after Khaleda Zia’s government handed over power to a caretaker government headed by a retired Chief Justice, free and fair election according to the 13th constitutional amendment were announced. This was also made necessary by the paralysis of democratic institutions since 1994, when opposition parties (AL, JP and JIB) started to boycott the Parliament while claiming that the coming election would never be fair unless organized by a 'non-party caretaker government'. They had also been asking that this change be enshrined in the constitution in order to make this practice systematic at the end of every government’s tenure. The BNP did not initially answer these demands and contested parliamentary election alone in February 1996. Under pressure from the opposition’s hartals, the new Parliament, monopolized by the BNP, eventually had to give in and have the amendment voted, but added a 'detail', barely noticed: during a period of ninety days, while the caretaker government is supposed to organize the elections, the higher command of the armed forces is entirely vested in the President (Islam 1996: 22). Yet the President at that time, Abdur Rahman Biswas, had been elected by the outgoing BNP majority as per the rules of the parliamentary system.
91Therefore, and in contrast with the non-partisan caretaker government’s prerogatives, the army’s control was not in neutral hands. The crisis started when President Biswas decided to dismiss two officers considered close to the AL. The COAS, General Nasim, refused that order and openly confronted civilian power. He ordered units to march against Dhaka in order to pressure the President. This crisis, hastily described as a coup attempt, was finally solved thanks to the loyalty of the 'two-star generals', including the head of military intelligence (DGFI), who managed to arrest General Nasim. Eventually, democratic elections were organized in June 1996, which gave the AL the majority in Parliament. As a majority of officers refused to support the so-called coup, several scholars saw in this event a proof of the army’s growing professionalism.59
92That professionalism was the main reason behind the aborted 1996 'coup' remains to be demonstrated. A fairly different interpretation can be proposed. After nine years of stability in civil-military relations during Ershad’s government, this politico-military crisis clearly exemplified the renewed politicization of the armed forces and the re-emergence of factionalism, yet of a different sort that the one prevailing in the 1970s. The two main political parties, using their influence on the army to give a bias to the coming election, nurtured this new factionalism. One should not forget, indeed, that the military was to be deployed to assure law and order on the polling day. Several interviews with retired officers, who played a role in these events, indicate that the loyalists were all supporters of BNP, whereas the officers who were initially dismissed had had contacts with the AL with the next election in view. The COAS himself was seen as close to the AL. Moreover, the three laid-off officers, including Nasim, were all ex-freedom fighters, while President Biswas had commanded razakar units in the 1971 war.
93So, the answer to our question is: yes, the army’s factionalism was a factor of political instability or, to put it in other words, the renewed factionalism implied by the alternation of 'clientelist' and dynastic parties in power has been threatening, in return, the democratization process itself. And the crisis occurred, expectably, during the interim period before the planed 2007 general election. At this juncture and intrinsically, the acting government does not have any popular legitimacy and the army’s higher command is left to the presidency, a necessarily weak institution in a parliamentary system. Neither this time did the head of the government have any legality: President Iajuddin Ahmed indeed took over as Chief Adviser of the caretaker government without following the process of the 13th Amendment. Moreover, political turmoil and violence in street battles caused by the AL and the BNP had reached intolerable levels, and foreign actors were regarding the impartiality of the promised election with more and more suspicion. By participating in these polls as the last guardians of the cherished though doomed 'free and fair' election – the electoral commission was biased towards BNP, and several monitoring foreign missions had already been recalled – the military would have severely endangered its credit60. Exactly for the same corporative motives as in 1991, the army followed the opposite path. It confiscated power from political parties by pushing Iajuddin to stall the election, declare the state of emergency, and resign as head of the government, while Fakhruddin Ahmed, a high-ranking bureaucrat backed by the military, took over as Chief Adviser61. The given reason: graft. Like at the end of the 8-year-long Ershad’s rule, corruption of leaders of the main parties and corruption/politicization of civil servants during the 16-year-long civilian rule had not only crippled the economy, it had also threatened the prestige of the army as an uncorrupted institution. As an interviewed officer explained (in 2003), one of the developments that account for this resentment lies in the military hierarchy’s perception of being manipulated by the government, notably because of the latter’s populist use of the army to combat crime and corruption in 'joint operations', and the renewed contacts with inefficient and corrupt civil agencies implied by these operations, to which we shall now turn.
94In contrast with political parties and law-enforcing agencies, whose level of corruption and links with criminals are for both notorious, the public poorly understands the relation between the army and political parties. As a result, the military still enjoys an unparalleled prestige. Normally, the isolation of soldiers and officers in cantonments shields them from the inherent vices of the Bangladeshi politics, which have existed since the restoration of a competitive regime. This is obviously a 'myth'. The perception of the military as the last of the incorruptible emerged when Khaleda Zia’s government launched a massive crackdown on crime in October 2002, 'Operation Clean Heart', under the double pressure of foreign donors who were seriously concerned with internal (in)security, and of a public opinion which was expecting the BNP to keep its electoral promises of 2001.
95For the government, the task was delicate: How to appear sincere in such an anti-crime operation while most of the party leaders, including Members of Parliament, were patronizing local godfathers who, in return, were giving the party a share of their benefits and preventing opposition activists to work in BNP’s constituencies? The success of the operation lay with the extreme confidentiality with which it was conducted. According to interviewed officers and press sources, the group of people involved in the decision-making process was limited to the Prime Minister’s Office, with Khaleda Zia at its head, Lieutenant-General Mashud Chowdhury (COAS) at the executive end, and Brigadier Rumi (Director General of DGFI) as the main intelligence source62. The exceptional nature of 'Clean Heart' was rooted in the fact that the Home Ministry had been kept uninformed for the first two days, during which the army could arrest the most untouchable criminals thanks to the list provided by the DGFI: 1500 people were taken into custody on the first night alone, including leaders of both the AL and the BNP63.
96Three aspects of this operation are noteworthy with respect to civil-military relations. Firstly, it was made possible only by the relationship of trust, which then prevailed between the PM and the COAS. This was obviously crucial, as during the 3-month crackdown, a total of 24,000 soldiers were to be deployed64. It had therefore been prepared several months ahead of time, with a series of drastic actions aimed at ensuring that the PM would get the military’s support in due course. As an interviewed officer says:
The army could never have been deployed if the PM did not have good relations with the COAS. Because when the army is spread all over the territory, it takes a snap to take control of the country. But here, the COAS is theirs, the main generals are theirs [the BNP leaders’], they are loyal to them. So under these conditions it is easier to ask the army to conduct such an operation.65
97The BNP’s political manoeuvres which forced the previous President, Badrudozza Chowdhury, to resign in June 2002, and led to the retirement of the COAS, Lieutenant-General Haroon-ar Rashid, a few days later, were actually the preliminary measures to the deployment of the army; these two men are said to have opposed involving the army for internal security matters66. The nomination of Iajuddin Ahmed as the new President, as well as the replacement of General Haroon by General Mashud, who entertained close links with Khaleda Zia (he belonged to the loyalist side during the May 1996 crisis), eventually facilitated the crackdown.
98Secondly, the military was not in a monopolistic position: it had to share parts of the anti-crime drive with civilian law-enforcing agencies. The Home Ministry as well as the police, the Ansar67 and the Bangladesh Rifles got involved in the third phase of the crackdown thus renamed 'Joint Drive Operation'. Then, the number of arrests dramatically dropped. As an interviewed officer who participated in the operation explains it, the military was only in charge of the first and second phases – arresting and interrogating in the cantonments68. Afterwards they were compelled to transfer the detainees to the police so that the legal procedure could be followed. This is when the corrupt practices normally observed in the civil administration came back in full force: many arrested people who had links with the BNP, for instance, were eventually released.
99Thirdly, the operation stopped on 9th January, 2003, only after the military hierarchy negotiated the vote of an amnesty law with the government, which would protect the army from any legal case filed under civil courts against officers with respect to any of their deeds during those 87 days. A retired officer’s remarks about this law precisely express the sentiment of indignation it provoked in the concerned segments of society:
It was an artificial way to protect the army, but it was not necessary, I think, because by giving amnesty, you indirectly admit that the army illegally killed people. It had an adverse impact. It apparently protected officers. […] Justice should have followed its own process if people had been killed by some military […]. The truth would have come out, and the army would not have been blamed, not entirely. Maybe out of 10 cases, 3 of them would have shown abuses during interrogation in custody. But 7 of them would have been proved innocent. By giving amnesty, you suggest that the military can do whatever it wants. You admit, without any proof and any trial that yes, the army killed people. It is a bad precedent for the future. If there are more abuses, they will put pressure on the government to get amnesty.69
100As far as the institutional links between the civilians and the military are concerned, but also regarding democracy, 'Clean Heart' operation represented, in essence, a double-failure. On the one hand, in order to get the expected results, this drive gave to the army the primary role in serving arrests and in gathering intelligence. This implied a lack of transparency in the decision-making process because roles were not institutionalized and everything depended on the personal relationship between the PM and a handful of generals. Moreover, the haziness of the operation led the army to proceed in a 'military fashion' and demand legal protection thereafter. This situation fairly resembled a state of emergency, and as such was the first hit against the democratization process started in 1991. On the other hand, 'Clean Heart' demonstrated the inefficiency of the collaboration between the army and the civil law-enforcing agencies; the latter’s central flaws – corruption and links with criminals – have never been dealt with by the government.
101Given the country’s bustling past, this operation, which revitalized the military’s role in the domestic sphere, is, as shown, an important development. It has already been followed by other punctual operations at the local level, such as 'Spider Web' in July 2003 which intended to arrest members of outlaw leftist organizations in the very unstable southwest. Moreover, the state’s police has undergone an unprecedented process of militarization with the creation of the Rapid Action Battalion in 2004. Half of this elite force is composed of military, and its civil members are trained by officers. Beside, 'Clean Heart' left a bitter after-taste to the military itself. Military officers are surely conscious of the gravity of the law and order situation, yet many of them also felt manipulated or betrayed by the political power70.
102This paper highlighted two of the most original characteristics of the Bangladeshi army, namely factionalism and politicization. It demonstrated how they seriously hampered the institutionalization of civil-military relations in this country. The analysis also suggested that the political power’s manoeuvres, either civilian or military, aggravated the army’s structural instability born of the Liberation war heritage.
103Our study shows that the military’s professionalism and the institutionalization of the armed forces’ role were, in fact, encouraged during the 1980s military rule. Though functioning on a corruption-based model which was harmful to the institution, Ershad’s regime still allowed the military to re-invent its own tradition and behave like a classic force defending the country towards an external enemy and intervening in the 'new world order' through UN peacekeeping missions. These policies deepened a maturation process that could have transformed it into a conventional army, isolated from civil society like most of the South Asian armed forces. However, we have demonstrated that the return of civilians to power in 1991, triggering permanent confrontational politics between the two main parties, has reintroduced partisan cleavages into the army and threatened the stability of civil-military relations.
104Our findings, especially in the third part of this paper, have serious policy implications. Because it was generally understood that Ershad’s resignation had paved the way for a professional, incorrupt and non-partisan army, the Bangladeshi troops have been increasingly implicated in the democratization process as allegedly neutral agents. Indeed, since 1996, they have regularly been used to maintain law and order during electoral periods. But the way they fulfil this task, under a partisan President’s command and with a list of 'criminals' made up by the outgoing government in order to evict its political adversaries, cannot guarantee the neutrality of the army’s role. Involving the military in the electoral process and in the maintenance of law and order on polling-day is, in our view, a dangerous process; this should be left to agencies controlled by the neutral caretaker government. Every interviewed officer ultimately asserted that the soldiers’ place was either in the barracks or on the battlefield and that repeated contacts with civil society would only increase the risks of corruption in the army.
105On 11th January, 2007, while a consensus between the AL and the BNP could not be reached to hold credible polls, President Iajuddin cancelled the election, imposed a state of emergency and stepped down as Chief Adviser of the caretaker government in favour of Fakhruddin Ahmed, a retired bureaucrat backed by the army. Since that event up to this paper’s writing, Bangladesh has been governed by a 'bicephalous' structure. The caretaker government has been vested with the powers of an elected government in addition to its legal role of making the political and institutional environment suitable for free and fair elections. But to be able to perform this task, Fakhruddin’s cabinet has been seeking help from the military, which created a 'National Coordination Committee against Crime and Corruption' headed by the 9th Infantry Division’s General Officer Commanding (GOC), Major-General Masud Uddin Chowdhury, and which drafts in all seven GOCs as members. This 'coup that dare not speak its name'71 aimed in particular at launching a 'clean-up' operation against both the AL and the BNP, so as to break up the 'political dynasties' that had been ruling the country since 1991, and to make a new leadership emerge.
106Whether they lead to a lasting period of army interventionism or not, the events of 2007 have already reshuffled the fragile civil-military equation of Bangladesh. In absence of an elected Prime Minister, President Iajuddin’s theoretical command over the army has been overstepped by the COAS and a handful of active and retired generals, who also supervise all the law-enforcing agencies through the National Coordination Committee. The COAS, Lieutenant-General Moeen U Ahmad, did not hesitate to take a public stance on national issues and policies, while assessing that the ‘armed forces [were] only assisting the civil administration’72. And, as he said, ‘no one [has been] spared’ in this crusade against corruption73. After 16 years of democratization de façade, during which the alternate political parties have been manipulating the army’s hierarchy and distorting its prestige while re-awakening its political factions, the military actually showed the strength of its corporative interests in a way that backfired against politicians, as well as its intention to still play a role in Bangladesh’s political destiny.