Can a Witness Be Convicted? A Clash Between Domestic and ICC Principles

Introduction

The conviction of IGP Abdullah Al-Mamun under the International Crimes (Tribunals) Act, 1973 (ICT Act) presents a profound legal paradox. While the accused was formally granted pardon and testified as an approver status conferring immunity under domestic law, the tribunal proceeded to convict him. This analysis argues that this conviction is void, not merely because it breaches foundational domestic criminal procedure, but because the tribunal’s attempted justification through an imported, and misinterpreted, International Criminal Court (ICC) framework is legally untenable. The tribunal, in seeking to align with international practice, created a hybrid procedure that respects neither the explicit protections of Bangladeshi law nor the precise, limited role of cooperation under the Rome Statute.

 The Domestic Approver Doctrine: A Guarantee of Immunity

The legal status of an approver in South Asian common law is unequivocal. Rooted in the Code of Criminal Procedure and solidified by precedent, the tender and acceptance of a pardon effect an immediate discharge. The individual ceases to be an accused and is transformed into a witness for the prosecution. This is not a matter of judicial discretion but a legal metamorphosis. As established in cases like A.J. Peiris v. State of Madras and State v Ershad Ali Sikder, a witness cannot be convicted for the offenses for which pardon was granted.

The only path to prosecution is the formal forfeiture of the pardon, a rigorous procedure requiring a prosecutor’s certificate of non-compliance, a hearing for the approver, and a fresh trial. In Mamun’s case, this entire domestic legal superstructure was ignored. He was examined under oath, cross-examined, and then convicted a procedural sequence that is a legal impossibility under Bangladeshi criminal jurisprudence.

 The ICC Framework: Cooperation, Not Pardon

In what appears to be an effort to internationalize its rationale, the tribunal implicitly relied on a model of cooperation derived from the ICC’s Rome Statute. However, this reliance is fundamentally misplaced, as the ICC system explicitly rejects the common-law concept of a pardon that extinguishes liability.

The Rome Statute contains no mechanism equivalent to the tender of pardon. Instead, its framework is built on a different premise: cooperation is a mitigating factor in sentencing, not a bar to conviction. As noted by the United Nations Legal Affairs, any discussions or agreements between the prosecutor and the defence “shall not be binding on the Court.” The Court retains ultimate control over questions of guilt and sentencing. This principle was clearly demonstrated in the Lubanga case (ICC-01/04-01/06-3173), during the review of sentence.

The Court explicitly considered whether cooperation during the trial could justify a sentence reduction and underscored that such cooperation does not provide an automatic discount. The Chamber noted that while cooperation already factored into the initial sentencing could be relevant upon review, it is not a guarantee; for it to carry significant weight, the cooperation must be “early and continuing.” This jurisprudence definitively shows that the ICC does not treat cooperation as a complete “get-out-of-prosecution-free” card, but rather as a contingent factor that influences punishment, not liability.

The ICC’s primary concern with cooperation agreements, as seen in the scholarly analysis of the Lubanga trial, often revolves around the tension between the Prosecutor’s confidentiality promises to informants and the Court’s overriding obligation to ensure a fair trial. This complex balancing act is a world away from the domestic approver doctrine, where the act of pardon conclusively resolves the issue of liability.

 The Tribunal’s Jurisdictional Error: Creating a Lawless Hybrid

The ICT Tribunal’s judgment in Mamun’s case suffers from a critical jurisdictional error. It attempted to graft select ICC principles onto a domestic proceeding without the statutory authority to do so and in direct conflict with the un-repealed domestic law on approvers.

By convicting Mamun after treating him as a witness, the tribunal effectively adopted the ICC stance that cooperation agreements are not binding on the court. However, it did so without providing the corresponding ICC-mandated procedural safeguards. Under the ICC model, a cooperating witness like Mamun would have still been tried as an accused, with his cooperation being a powerful mitigating factor argued at the sentencing stage under Article 76 of the Rome Statute.

The tribunal, by contrast, applied the domestic procedure of granting a pardon (which implies immunity) but then imposed the ICC consequence (conviction regardless of cooperation), thereby creating a worst-of-both-worlds scenario for the accused. It granted the form of a domestic pardon but denied its legal substance, all while failing to follow the ICC’s structured process for evaluating cooperation within a sentencing hearing.

This approach renders the proceeding void. A tribunal derives its power from its governing statute and the surrounding procedural law. The ICT Act does not abrogate the domestic law on approvers, nor does it explicitly incorporate the ICC’s specific rules on cooperation and sentencing. The tribunal, therefore, acted without legal authority or ultra vires by convicting a man whom domestic law, which it was bound to apply, legally shielded from conviction. The judgment is not just incorrect; it is a legal nullity.

A Reflection on Legal Certainty and Justice

The conviction of IGP Abdullah Al-Mamun presents us with a challenging legal dilemma that goes beyond this individual case. In seeking to navigate between domestic traditions and international practices, the tribunal’s approach has created more questions than answers. The domestic doctrine of pardon, which offers clear protection to those who cooperate with the justice system, appears to have been set aside in favor of an international framework that operates on fundamentally different principles.

This situation invites us to reflect on some fundamental questions about how our justice system functions. If a witness who has been granted protection can later face conviction without the proper forfeiture procedures, what assurance does any cooperating individual truly have? When courts develop hybrid approaches that draw from different legal systems, how do we ensure consistency and predictability in our laws?

Perhaps most importantly, we might wonder what message this sends to potential witnesses in future cases. The strength of any justice system often depends on the willingness of individuals to come forward and speak truthfully. If the protections offered to such individuals appear uncertain, might others hesitate when their testimony is needed?

These questions touch upon the very foundations of legal certainty and trust in our judicial institutions. While the pursuit of justice may sometimes require adapting to new legal ideas, this case reminds us that such adaptations must be carefully considered, always respecting the fundamental principles that uphold the integrity of our legal system. The resolution of these questions will ultimately shape not only this particular case, but the broader landscape of legal protection and judicial reliability in our jurisdiction. The answers to these questions will determine not only the fate of one individual, but the future of legal certainty in our jurisdiction.

Author Mokarramus Shaklan is an Advocate, Supreme Court of Bangladesh