HEFAZAT-e-Islam presented itself as an ostensibly cohesive and homogeneous entity through its May 5 rally at Shapla Chattar. Underlying this apparent unity, however, is the demonstrated potential of Bangladesh's Deobandis to disagree and compete, and often to undermine one another.

Hefazat's lifespan will depend on the willingness of these men to temper their personal ambitions and on their ability to collaborate durably. The past behaviour of Deobandi leaders in the educational and political arenas suggests this won't be easy. Hefazat possesses the elementary ingredients to challenge the status quo through its large and self-propagating network of dedicated men keen on transforming the political culture of Bangladesh. Yet power struggles within the movement will likely keep it from reaching adolescence.

While the total number of Deobandi (qawmi) madrasas in Bangladesh remains a mystery, one can broadly place their strength in the range of 5,500–10,000 (this estimate excludes maktabs). Each of these institutions administers the Dars-e-Nizami syllabus, each is devoted to propagating the Deobandi worldview, and each is committed to resisting state interference in the qawmi madrasa system.

Yet, Bangadesh's qawmi madrasas are characterised by latent but very real divisions. The competing desires of rival maulanas to establish and wield social control have given rise to five qawmi madrasa boards in addition to several less prominent ones. The largest of these boards is Befaqul Madaris al Arabia Bangladesh. Headquartered in Dhaka, it has 3,600 member madrasas across the country and is the only qawmi board that claims and aspires to represent all of Bangladesh's qawmi madrasas.

The other boards exhibit only regional aspirations. Ittehad-ul-Madaris, based in Chittagong's Al-Jamiah Al-Islamiah Patiya, oversees approximately 600 madrasas in Chittagong Division along with a handful in Feni and Noakhali. Azad Deeni Iddara-e-Talim Bangladesh supervises around 500 madrasas in Sylhet Division. Tanzeem-ul-Madaris is headquartered in Bogura's Jameel Madrasa and represents madrasas in North Bengal.

Similarly, a board in Gohardanga supervises the affairs of qawmi madrasas in South Bengal. Representatives of these boards often trivialise the cause for their division (e.g. "We give more importance to Farsi and logic" or "We use different books to teach Arabic literature"), but the truth is that each board represents a power centre—an area of social control—enjoyed and dearly guarded by a clique of powerful maulanas.

Though mostly latent, these divisions do find expression whenever one sphere of social control impinges on another. In 2006, Mufti Abdur Rahman of Dhaka's Islamic Research Centre sought to unite all the qawmi boards under the banner of Somelito Qawmi Madrasa Shikhya Board. He succeeded in bringing the regional conglomerations on board, but failed to convince Befaqul Madaris.

Since Befaqul Madaris already claimed to present a platform for qawmi madrasas across Bangladesh, it considered Rahman's exercise redundant at best and, at worst, a direct threat to its own national aspirations. These intra-qawmi divisions also feature prominently when the state attempts to modernise and/or recognise the qawmi sector's curriculum.

In 2006, the BNP-led four party alliance toyed with the idea of recognising the Dawra-e-Hadith degrees of qawmi madrasas. One impediment to the plan's materialisation was the government's requirement that any qawmi madrasa board issuing these recognised certificates operate as a government entity, similar (and in parallel) to the alia system's Bangladesh Madrasa Education Board.

All qawmi madrasas oppose this stipulation, insisting that their own men control qawmi affairs, independently of the government. But even if the government concedes to this demand, it will struggle to please all qawmi quarters. Befaqul Madaris hopes to leverage its countrywide presence, thus favouring establishment of a single national qawmi board. Its regional rivals expect the government to recognise region-specific entities, respecting the currently decentralised nature of the qawmi landscape.

In this particular case, the government proposed establishment of a single qawmi board after Islami Oikya Jote MPs loyal to Befaqul Madaris lobbied for this arrangement (these MPs included Mufti Muhammad Waqqas, Mufti Shahidul Islam, and the late Mufti Fazlul Haque Amini). But the representatives of regional boards in Sylhet, Chittagong, North Bengal, and South Bengal turned down the offer.

This account shows that while all Deobandi elements favour government recognition of their degrees, this shared desire is subordinated to the more selfish objective of establishing social control at the expense of fellow Deobandis.

The unwillingness of one maulana to subordinate his ambitions to those of another is equally evident in the political arena. At its peak, Islami Oikya Jote (IOJ) existed as a broad coalition of Islamic parties, its constituents including Bangladesh Khilafat Majlish, Khilafat Andolan, Bangladesh Islami Andolan, Jamiat Ulema-e-Islam, Khilafat-e-Islam, Nizam-e-Islam, Ulema Committee, and Firaizi Andolan.

But, the inability of leaders to accommodate each other's ambitions caused constituent parties to progressively exit the coalition. Today, only four of the original constituent parties remain part of IOJ. Bangladesh Khilafat Majlish (BKM) left IOJ after a leadership dispute arose in 2001. BKM expected its amir—the late Sheikh-ul Hadees Allama Azizul Haque—to assume chairmanship of IOJ's central committee.

Other parties within IOJ preferred his former student Mufti Fazlul Haque Amini for the role. Similarly, Jamiat Ulema-e-Islam (JUI) left IOJ in 2008 when Mufti Amini sought to register IOJ as a political party. JUI took the view that IOJ was a coalition of willing members, and that registration of IOJ as a party would make the constituent parties (including JUI) irrelevant. It insisted that Mufti Amini should only be allowed to register his own constituent party—Khilafat-e-Islam; not the coalition as a whole.

This record of Deobandi power struggles in the educational and political domains should inform our expectations of Hefazat-e-Islam's trajectory. There is no denying that Hefazat is an expression of widespread Deobandi sentiment.

While one might disagree with the tenor of its thirteen demands, the latter represent an ideological common denominator shared by the large majority of Bangladesh's Deobandis. That said, Hefazat's fate will not depend on ideological homogeneity alone. Within this large body of likeminded men, the human qualities of sincerity and selflessness are distributed unequally. The occasional show of synchronised selflessness can create street noise at Motijheel. But disturbing the status quo requires far more.

The writer is a DPhil candidate at the University of Oxford.