UN Women dubbed violence against women as the "shadow pandemic" of 2020, warning that the gendered impacts of COVID-19 meant that, globally, women and girls were even more at risk. In Bangladesh, that translated to 1,627 cases of rape in 2020, and 1,321 in 2021 - and this data only reflects the cases reported in news media.

But what should journalists do if a country's legal definition of sexual violence is in direct conflict with reporting that is centred on survivors and acknowledges their trauma?

In October 2020, the newsroom of the Bangladeshi paper that I work for, The Daily Star, was deep in this moral quagmire. Should we stick to language from a 162-year-old penal code that defines rape solely as the penetration of a woman by a man she is not married to, and ignores all other forms of sexual violence? Or should we come up with reporting language that, while not breaking the law, makes readers question the status quo?

The conversation was triggered by the Begumganj case, where violent criminals used objects to rape a rural woman. She was too afraid to go to the police, especially after an elected village official refused her any help. It was only after a video of the violence was shared on social media by the rapists themselves that her story went viral, lighting the fuse for mass protests across the country.

The demonstrations surrounding Begumganj were the largest we'd seen, but they were part of a wider movement for justice in 2020. Throughout the year, despite pandemic-related lockdowns, several incidents of sexual violence led to such demonstrations around the country.

The Khagracchari case occurred a few weeks prior to Begumganj, when a group of men raped an Indigenous woman after breaking into her house. Ten months earlier, protests continued for days when a man raped a university student after abducting her from a bus stop in the capital's Kurmitola.

The anti-rape movement in Bangladesh not only demanded accountability from state authorities who had failed to stem the tide of sexual violence, but questioned a system and society that normalised it.

Trauma, and its impacts on individuals, families and communities, was at the heart of this story. But our reporting during this period was done without guidelines for handling such a trauma-laden issue. While some media organisations have certain rules and practices that have evolved over years of reporting, very few have them written down as organisational policy, and even fewer focused on sexual violence specifically.

Why does it matter how media reports on sexual violence?

Currently, the process of justice in Bangladesh faces numerous obstacles that are deeply connected to an honour culture that shames survivors into silence about their abuse. In the words of Bangladeshi feminist lawyer Sultana Kamal, there is a "misogyny embedded in the legal system [that] has to be addressed to ensure fairness to women seeking justice".

To show the role of the news media in perpetuating this problem, I isolated eight markers of gender-insensitive reporting, then chose four prominent Bangladeshi newspapers - The Daily Star, Prothom Alo, Bd Pratidin and Manab Zamin - and gathered reports related to three different rape cases mentioned above.

I analysed the reporting for a seven-day period before and after each story broke to find out how many of the markers were present. I then discussed my findings with the editors of the four newspapers surveyed, as well as reporters and news editors from these publications. I also consulted media and legal experts, Bangladeshi activists campaigning for survivors of sexual violence, and journalists reporting on similar stories in other parts of the world.

Eight markers of insensitive reporting

There is growing consensus that journalism that deals with traumatic experiences must be done with special sensitivity, keeping in mind the ethical and psychological implications of such work. Much of this reporting has historically been done on a trial-and-error basis, with little formal training on best practice journalism. As British journalist and founder of Trauma Reporting Jo Healey described in a 2019 blog: "Our current culture is to practice on the grieving public until we reckon we get it about right."

Last year, Columbia University's Dart Center for Journalism and Trauma published a style guide for trauma-informed journalism, arguing that "responsible reporting and newsroom decision-making now demand baseline literacy in trauma concepts and emerging reporting standards." This view was reiterated in the 2021 media guide created by the Centre for Women's Global Leadership (CWGL) at Rutgers University, which identified the need to bring together different resources and perspectives from journalists and media experts in one place to ensure quality reporting.

The Bangla language guide on gender ethics in journalism, published this year by media training organisation MRDI, and founded on the principle that gender-sensitive news judgement is essential to good journalism, was also instructive and provided useful context.

The eight markers of insensitive reporting on sexual violence that I chose are all related to language and framing, while interview techniques and agenda-setting are discussed with journalists and editors later in the paper.

Victim-blaming or shifting the responsibility of risk of violence onto the survivor can be done in a number of ways. Examples include what time of the night a victim went out or whether they knew the rapist before. It has the effect of placing the onus of women's safety on women's behaviour rather than on the state and society, ultimately minimising their trauma and perpetuating the idea that sexual violence is a source of shame.

Sensationalising reporting fixates on violent details and creates a spectacle of the crime. Highlighting the cruelty and unusualness of it all, it tries to reel in readers with shock value. All reporting guidelines agree that this sort of journalism is triggering for survivors and fails to uphold their dignity and agency.

Extraneous details and jigsaw identification Including unnecessary personal details of the survivor increases the risk of jigsaw identification. This means that even if the survivor is not named directly, enough details allow readers to piece together their identity. In Bangladesh, the law protects the survivor's right to anonymity, and forbids the media from publishing information that could lead to jigsaw identification as well.

Victim vs monster stereotypes A common frame used in reporting is that of the ideal victim and perpetrator, where survivors are portrayed through the lens of "helpless victim" and rapists as "monsters". In the Bangla language, victimising of the survivor is done with words like "dhorshita" and "nirjatita". While difficult to translate literally, these terms make the violence a part of the woman's identity - in the way you would call someone a student or a lawyer, these describe survivors as "the raped/abused one". In MRDI's Bangla guideline, it is suggested that newspapers should scrap these terms for "bhuktobhogi", the legal term used for victims of all crimes, not just rape. MRDI also suggests using the term "dolboddho dhorshon" instead of "gono dhorshon" to describe gang-rape, since the word "gono" carries with it a positive connotation.

Survivor and perpetrator: erased and over-represented The other side of the victim vs. monster narrative is the unbalanced representation of the survivor and perpetrator, where one or the other is erased from the narrative. Examples of this are headlines like "woman raped", and not "man rapes woman", where a passive tone of reporting focuses solely on the survivor.

Episodic vs thematic framing When reporting on sexual violence, journalists have a responsibility to highlight patterns, trends and contexts at an institutional level. When rape cases are reported on as isolated events, with a focus on details of crimes and individual behaviours and motivations in the private sphere, the journalism takes an isolationist approach known as episodic framing. Experts suggest employing a thematic frame instead, where the reporting focuses on a contextual, big-picture narrative to address the whole cycle of violence, from prevention to prosecution, and identifies gender-based violence as a human rights issue that requires policy intervention.

Lack of female sources To date, reports on GBV tend to be dominated by actors from male-dominated institutions: police, politicians and lawyers. Witnesses, relatives and community leaders interviewed also tend to be male. Globally, there is a glaring underrepresentation of female voices in media reports. According to the GMMP, in 2020, only 19% of subjects and sources in Bangladeshi news media were female, as were only 11% of reporters.

Poor image choice Finally, many images in reports continue to refer to trauma and the act of rape in indirect ways, such as by showing a hand being held forcefully, or a woman in tears, running/shrinking from a male figure. Such images can be extremely triggering for survivors. It is important for media organisations to stop using images that portray women as passive, helpless victims.

How did Bangladeshi papers fare?

Of the 258 reports that I looked at in my analysis, 63 were from the Kurmitola case, where a man abducted a university student near a bus stop in the capital and raped her before fleeing the scene on January 5, 2020. The Khagracchari case - where Bengali robbers gang-raped an Indigenous woman in a south-eastern town on September 24, 2020 - received far less attention: only 21 reports were printed, despite the fact that demonstrations were organised to protest settler violence against minority communities. Soon after, Begumganj became the straw that broke the camel's back, culminating in the biggest nationwide protests yet when the video of a gang-rape was released online by the rapists themselves on October 4, 2020. Over the one-week period analysed, I found a total of 174 print reports related to Begumganj.

More than 68% of these reports used isolationist frames and 49% completely erased the victims from the narrative. Almost 66% of reports had zero female sources. Less than a fourth had the voices of rights activists or GBV experts.

These are astonishing findings when you consider that all the reports concerned female survivors of violence. This huge gap could also explain the treatment of survivors in Bangladeshi media.

On top of the sensationalising, around 36% of reports portrayed them as helpless victims, with no allusion to their trauma or resilience. In fact, almost half of the reports did not refer to them at all - the survivors were completely erased from the narrative.

In the 23% of articles that did include voices from the Bangladeshi women's rights movement, 80% referred to gender-based violence as a human rights issue and discussed the structural problems that lead to it.

Almost all quotes related to protests were from male demonstrators, mostly leaders of student bodies affiliated with political groups. Many of them referred to women as their "mothers" and "sisters", and used other patriarchal and often victimising language to talk of survivors. Journalists reproduced these comments without interrogating this representation of women as passive actors needing protection, rather than citizens actively demanding their rights. On the other hand, when female protesters' voices were included, many spoke of wider issues such as rape culture, gender inequality and hate speech against women.

Another statistic points to a possible reason for a lack of female sources: less than 10% of reports were produced by identifiable female journalists.

Out of the 24 reports identifiable as being written by female journalists, only one report did not include a female source. It was, however, difficult to ascertain the reporter behind a large proportion of reports that were credited to the newsdesk.

For Aasha Mehreen Amin, joint editor at The Daily Star, this is simply not good enough. "There will always be not enough reporters, but the question is: are newspapers willing to give gender-based violence the same treatment they give to politics, economy and even sports?"

"We can be more innovative, especially in our use of editorials and opinion columns, in keeping the topic of gender-based violence in the public eye," she added.

Journalists need to be sensitive to biases in class, race, religion and location

Of the three cases of sexual violence selected for analysis, the Kurmitola survivor - a middle class, university student from the capital - was the most present in narratives. At the same time, reporters failed to follow-up police information on how the perpetrator had previously raped other homeless and disabled women, showing a considerable apathy towards survivors from marginalised groups.

This apathy was extended to the other two survivors, who come from working class, rural backgrounds. The Khagracchari survivor, who is also an Indigenous woman, initially received so little attention that Prothom Alo and Bangladesh Pratidin reported on the protests surrounding it before they actually reported on the case.

Surprisingly, even though the Begumganj survivor created a storm of debate surrounding sexual violence in Bangladesh, her trauma was the most underrepresented in news reports. She was erased from over 53% of reports, and 38% presented sensationalist details about her abuse.

Information from police and medical reports is only muddying the waters

All the reporters I interviewed spoke of consulting the police's first information reports that are filed when a case is initially accepted at the station, and of how these reports tend to be filled with insensitive comments, graphic descriptions, and sometimes even incorrect information.

Another common tendency was to repeat information from police reports or police quotes without interrogating their statements in any way. For example, the Kurmitola rapist turned out to be a homeless man with a history of drug addiction and, during that period, police routinely made comments linking drug use with sexual deviancy. This narrative was reproduced without question in a number of reports, ultimately erasing the simple fact that sexual violence occurs across all sections of society.

Confusion over reporting practices is negatively affecting survivors

Journalists also spoke of the issues they faced while speaking to survivors. One of them shared how reluctant she once felt in approaching a rape survivor who was still traumatised and had already had too many journalists interview her. What is concerning is that despite her qualms, she still felt compelled to conduct the interview.

In the CWGL 2021 media guide, meaningful consent is defined as consent that comes from the survivor, is given for specific use at an appropriate time, is trauma-informed, and where requests for consent are repeated. Ideally, the reporter should have a line of communication open with the survivor to ensure that she understands the risks of sharing her experiences and that her vulnerability is not taken advantage of in any way.

In the absence of training on how to deal with vulnerable and traumatised persons, journalists also struggled in managing emotions. Many felt intense guilt for not staying in touch with survivors after gaining their trust. One reporter told me about continuing to communicate with a survivor because she was suicidal. Another mentioned how she tries to step out of the role of journalist and be their friend. While she spoke of how she consoled them, it became clear that in the process, she had ended up trivialising their pain.

What can Bangladeshi journalists do right now to make sure they treat survivors with empathy and dignity? Australian journalist Isabella Higgins, who covered Indigenous Affairs for ABC for five years and dealt almost exclusively with vulnerable communities, and she suggested the following:

Make sure survivors know they don't have to answer your questions. People who have experienced trauma from marginalised communities can forget that it is their right to say no to us, and we need to empower them to do that.

Start your interview by asking them: "What do you want to get across? What is your message?" Use their own words as much as you can, and let them set the tone and mood of the story.

Don't sit face-to-face and stare them in the eyes like you are about to interrogate them. Conduct interviews in the way that people talk to each other, sitting side by side, while walking or while they are doing an activity. How much eye contact they want to make should be up to them.

Never promise that your interview will make them feel better, or that actions will be taken as a result of their story being published.

Ask questions that reflect their resilience. I often interviewed people when they were on their healing journey and not right after the traumatic event, so that they can talk about what they have overcome and how they are so much more than the trauma they experienced.

Be conscious of the power you have as a journalist, but also remember to maintain professional boundaries.

Ultimately, bringing in reporting practices that are trauma-informed and centred on survivors will not only be for the sake of gender justice, but for the sake of good journalism as well.

From Reuters Institute for the Study of Journalism.