Rafia Tamanna of 抖阴精品 speaks to Jumanah Bawazir, a researcher of Forensic Architecture about methodology, the investigation of the killing of Abu Sayed and possibilities of building counter-forensic capacity
ABU Sayed, a student activist at Begum Rokeya University in Rangpur, was killed on July 16, 2024. The death became a flashpoint in what would grow into the July uprising. While the Awami League government, toppled in the uprising,听 then claimed that Sayed had been killed with bricks and firearms by other protesters, video footage shared widely on social media had already revealed police presence and the use of live ammunition. Researchers from Forensic Architecture in collaboration with Drik Picture Library undertook a reconstruction of the killing scene using photogrammetry, 3D modelling and video mapping.
听
抖阴精品: Could you briefly introduce Forensic Architecture鈥檚 overall mission and methods, especially how your work bridges investigative research, legal frameworks and public engagement?
Jumanah Bawazir:听We work across contexts, including cases of state violence, corporate crimes and environmental destruction. A core part of our approach is what we call听counter-forensics. This is a method that relies on open-source, public information to reconstruct and understand the events that are often obscured by official control of access and narrative.
Typically, the area where a crime or act of violence occurs is cordoned off. The public is excluded, sometimes even journalists. Our work uses tools such as video analysis, spatial modelling and witness accounts to 鈥榣ook over the cordon,鈥� reconstructing the events that happened in spaces the public couldn鈥檛 physically access, but where evidence often circulates online.
听
抖阴精品: Could you share some of the major cases Forensic Architecture is working on?
Jumanah Bawazir:听Our team has continuously documented crimes committed in Gaza. Much of that work depends on access to open-source information 鈥� videos, tweets, location data, etc. But when communication is deliberately cut off, when journalists and witnesses are targeted, this work becomes far more difficult.
That disruption is not accidental. It鈥檚 a tactic. Once the flow of information is broken, so is the collective capacity to witness. And what often follows is distraction or fatigue. People begin to forget, even if only momentarily. That forgetting is built into the logic of the blackout. It鈥檚 a way of muting accountability. Cutting off information flow disrupts the collective capacity to witness and, with it, the possibility of accountability.
We鈥檝e seen very little responsibility taken by major media organisations for spreading unverified or misleading narratives. And at this point, apologies aren鈥檛 meaningful.
We also have another team working on the German genocide in Namibia.
听
抖阴精品: What was your experience of working on the forensic reconstruction of Abu Sayed鈥檚 killing during the July uprising?
Jumanah Bawazir:听In this case, a large volume of visual material was already available, especially mobile videos and witness clips that had been widely circulated.听
We worked with Drik, which coordinated on-site efforts. One of our key contributions was building a 3D model of the site using听photogrammetry, a technique that involves stitching together photographs taken from drones to create a detailed reconstruction of the space. Drik collected the drone images after a few test runs and we processed them in our office. The resulting model helped anchor all the videos and interviews within a clear spatial framework.
听
抖阴精品: Your investigations mix imagery, modelling and testimony. How do you decide which tools are appropriate for a given case? And, how do you address the legal status of digital evidence?
Jumanah Bawazir:听It鈥檚 always case-specific. A 3D model is most useful in small-, human-scale events like shootings or police violence where you need to understand how people moved, where the shots came from and what sightlines existed. It helps clarify things that raw footage alone may not fully capture.
As for legal admissibility, many of our projects begin with invitation. If we know from the outset that a case is intended for legal proceedings. We handle the evidence accordingly.听
Our methodologies, how we investigate, model, and verify, remain the same whether a case is intended for legal proceedings or not. What changes is the process around documentation. For legal contexts, we may follow specific protocols. We may adapt how we document and present findings to meet courtroom requirements.
听
抖阴精品: In cases with limited video or photo evidence, what techniques do you rely on to construct a reliable picture of what happened?
Jumanah Bawazir:听One of the most effective tools that we use is called听situated testimony. Even when we don鈥檛 have full video or photographic coverage, we can construct a 3D model and use it as an interactive reference during interviews with witnesses. When survivors or witnesses are re-immersed in a spatial reconstruction of the site, carefully and with consent, it often unlocks specific memories tied to location and timing.
It鈥檚 not always about building the full picture at once. Sometimes, we focus on a short sequence, a few seconds, and cross-reference accounts minute by minute. If multiple people describe the same action, position or weapon, we treat that as corroborative evidence. We also use visual references, for example, images of bullets or shells, to help witnesses to identify what they saw. This kind of triangulation is key.
听
抖阴精品: Do you envision local or regional teams developing their own counter-forensic capacities using Forensic Architecture鈥檚 methodology?
Jumanah Bawazir:听Definitely and this is already happening. We helped set up a forensic unit in a Palestinian human rights organisation in Ramallah. We worked with them on the Shireen Abu Akleh investigation. That team now runs its own cases. There are also collectives in places like Paris, Athens and Mexico, now working independently.
The workshop that we did with Drik was part of that broader vision听to decentralise the techniques and make them more available to local actors.
听
抖阴精品: Where do people who want to learn the methods begin?
Jumanah Bawazir:听It鈥檚 more accessible than it looks. The final products we release, the videos and the reconstructions, can seem complex, but every step behind them is learnable. Our web site lists the specific methodologies used in each case and almost all the software and techniques are open-source or available online.
If someone wants to learn how to do the work, they absolutely can. It takes time and practice, but there are tutorials, toolkits and archives to study. You don鈥檛 need special access. What you need is intention and a case to work on. From there, it becomes a process of learning and assembling.
听
Rafia Tamanna is an editorial assistant at 抖阴精品.
