How do narratives spread, and what does this mean for how we report the news?

"People don't have that much time to get informed," Giovana Fleck points out, noting that for many consumers, particularly amid a global cost of living crisis, keeping up with the news is not a top priority.

There is also disinformation and polarisation to consider. Especially as social media platforms are ill-equipped to react promptly, and news media sometimes struggle to counteract it because of limited resources.

Fleck leads the Civic Media Observatory at Global Voices, an international citizen media publication. The observatory tracks different narratives over time to analyse how and by whom they are spread. That includes through news articles, but also by politicians and on social media or even billboards and posters.

The idea is to provide "an extra piece in the puzzle" of tackling the spread of harmful narratives. By understanding what is happening and how people are talking about it, especially in areas where the international media often has limited coverage, media outlets can cover stories more effectively - and communities can be better informed.

How it works

The research happens in cycles, some of them country-focused and others topic-focused. Teams have followed narratives on specific topics such as covid-19, elections in Taiwan, violence in Ethiopia, and currently there are research projects running in nine countries: India, Turkey, Russia, Afghanistan, Pakistan, Bangladesh, Myanmar, Nicaragua, and Mali.

This work has looked into how anti-Muslim narratives are spread in India, for example, putting into context a recent wave of Islamophobic violence in the country. Researchers identified harmful narrative frames including 'Muslims are outsiders and not real Indians' and 'All Islamic structures in India should be demolished or converted into Hindu temples', identified posts and articles that perpetuated these narratives, and highlighted the role of right-wing groups.

These insights are then made public in the Civic Media Observatory's database and newsletter, and in some cases the team then suggest specific actions to editors, tech platforms or civil society actors.

"Our goal is not to debunk information or say this or that narrative is fake. Our goal is to explain how the narratives form, who's putting it out there, how it's spreading, who are the people profiting, why it is appealing to people, is that harmful or positive, what could be the possible consequences," explains Fleck.

In each country where Global Voices is running a project, there are two or three researchers, either academics or journalists, but all local experts on the country (although for security reasons, some researchers are based outside the country they work on).

Because of the small size of the team, Fleck emphasises that the observatory does not aim to cover all the news in each country, but relies on researchers' local knowledge to select the most important narratives. To avoid burnout - since the work often involves diving into misinformation and harmful narratives - researchers are encouraged to follow their own interests too. Editors also suggest topics to researchers based on the news agenda, "because we all have our blind spots", says Fleck.

This reflects the ethos of the Global Voices newsroom: focusing on local knowledge and a long-term view, covering stories before and after international media 'parachute' in.

The team of five editors reviews all the content submitted to the database, making sure it is clear and accessible so it can actually be used by anyone who would find it useful. This includes avoiding ornate language or technical jargon and being as concise as possible.

To look at the process step-by-step, Fleck shares the example of a narrative registered by the observatory’s researcher in Russia (before the invasion of Ukraine), titled 'feminism is a Western concept that poses a threat to Russian society'. The titles of the narratives are intended to reflect as accurately as possible the way that the messages are spread.

The researcher first registered a news article in Russian state media, which covered - in a positive light - a proposed bill to block feminist content online. The researcher gave concise answers to questions about subtext implied in the article that a general reader might miss, local context the reader would need in order to fully understand the item, and information about reactions to the item.

From there, the researcher tracked how this narrative was being spread more generally, including in news media, in WhatsApp and Telegram messages. She also registered key research terms for this narrative frame, as well as a list of relevant people and groups, which allows other people to replicate the findings and do further research into the topic.

What happens next?

All this research is not intended to sit unused in a database, and action is built in to the observatory's workflow. Each narrative frame is assigned a ‘civic impact score’, ranging from -3 (hateful or disinforming, likely to cause harm) to 3 (accurate and expands understanding). The news article mentioned above was rated as -1 (false, misinforming, inaccurate or biased material).

There is direct collaboration between researchers and the Global Voices newsroom, which produces articles based on the topics and trends uncovered by the observatory.

The team also works with partners at tech platforms, currently including Meta. One of their biggest successes was during the military coup in Myanmar, where Facebook plays a major role in communication - the teams collaborated to develop tools to protect activists online as well as to remove harmful content and profiles that were identified by the Global Voices researchers. Researchers can suggest actions such as flagging to partners, further research, or include in the newsletter or report.

Aside from that, all databases are publicly available for anyone to use in their work, be it journalists and editors, civil society, or even highly engaged news consumers.

A clear advantage for international time-poor newsrooms are the ability to assess whether reporting on a certain line - or even smaller actions such as headline and image choice - can inadvertently advance a harmful or false narrative.

"It may not be the goal of the article or the organisation, but that's how people are going to interpret it. It's a bit heartbreaking, and unfortunately it's still a part of our practice [of journalism]. It's rushed, it's low pay, people take decisions without considering the possible impact," she says.

Newsrooms can also use the data to understand what local or historic context is important to include in reporting.

"It's a good and comprehensive tool to have a sense on not only what are the narratives but who are the key people transmitting them, and what we should be watching out for in a year. And especially in terms of subtextual messages - things our researchers evaluated that can pass on the wrong message - [the database can show editors] how to avoid that, how to be aware from the start of how your public will interpret what you write, which could be about a certain word choice for example," she adds.

Fleck cites an example from her home country of Brazil, where President Jair Bolsonaro is currently campaigning ahead of the October 2022 election, and is re-using many of the narratives from previous campaigns, including a strong pro-gun stance.

But Bolsonaro's loosening of restrictions on gun ownership have led to an increase not only in ownership but also in violent deaths and robberies, she says.

"It's a very clear illustration in an election period - but people often forget that we have seen these narratives before. One of the things about the observatory is that we can go back and retrieve information from the past in a very concise and clear form, and use it for covering stories in the present, or even trying to anticipate how harmful it could be in the future, because now we know how those narratives played a role in society and how people reacted."

For Fleck, the main value of the observatory's work therefore comes from its long-term nature: following how narratives shift, rather than a typical news article which might act more as a snapshot in time.

"We follow narratives usually for a year at least, to see how those narratives are evolving, and they change a lot. It’s really diving inside the rabbit hole, many times, and seeing how things are changing in these countries."

Free daily newsletter

If you like our news and feature articles, you can sign up to receive our free daily (Mon-Fri) email newsletter (mobile friendly).