This is an automatic English translation of the Dutch original.
Livestreams are very popular. Video footage of things happening right now, as if you’re there. You see them more and more. But a lot of problematic information also circulates in these streams: rumours that cause unrest, or deliberate disinformation.
Video footage you see can be misleading. Even when it’s live-streamed. Then you’ve been taken in. By manipulation. You may even be a victim of disinformation, if incorrect information is deliberately shared in a stream. And then the consequences for our democracy can be serious.
This is what Esther Hammelburg looked at for our quick scan Dynamic disinformation. It shows that there are indications this is a considerable problem, but that it is very hard to research properly.
‘Written text is easy for researchers to label, but dynamic disinformation is not’, she said about this at the start of this project. Together with Post-X director Pieter van Boheemen she is diving further into this subject. Pieter develops tools for automated analysis that are needed for this, for example to convert audio into text.
What did they discover?
Esther and Pieter note that livestreams…
- … create a strong experience of ‘now, here, together’ that connects people;
- … offer an authentic feeling – after all, you see it happen with your own eyes;
- … are deployed to increase impact – including that of terror;
- … are a new form of journalism, aimed at community building.
But livestreams can also be deliberately misleading. Sometimes by manipulating the footage, but also more subtly. Not whole lies or fabrication, but half-truths. See it as gossip, but by someone trying to sabotage our democracy by causing confusion, distrust and unrest.
So we’d do better to talk about manipulation
Esther and Pieter advise shifting the focus of the research from disinformation to manipulation. And then we should not only look at how you can detect deliberately spread incorrect information, ‘but also at the (networked) environments in which problematic content is spread’.
Because Esther and Pieter find it ‘plausible that there are manipulation campaigns aimed at spreading ‘user generated content‘ that can cause or stimulate polarisation and ‘information disorder‘ in a society’.
The researchers also write that content acquires problematic meanings in a symbolic way or within a certain context. ‘For example, a TikTok video can be free of factual inaccuracies and seemingly unproblematic, but acquire a problematic meaning through the use of a particular song.’
We need to know more about it
More research is needed in any case. For now there is the quick scan, which, besides livestreams, also considers disinformation in podcasts, games, TikTok videos and Instagram Stories.
You can read it all in this download.
The research is part of a collaboration with the University of Amsterdam and Wageningen University & Research, and funded by the Ministry of the Interior and Kingdom Relations.
Will you think along with us in our data sprint?
Who is behind this manipulation? And how do you do that, in livestreams? How does the spread work? Which methods can we use or develop to research this? With those questions we get to work during the Digital Methods Initiative Summer School of the University of Amsterdam, from 30 June to 11 July.
There, Esther and Pieter run a ‘data sprint‘: together with experts in live-streaming, influencing, disinformation and data research, experimenting with ways to collect and analyse data. In this way, in an interdisciplinary team, we investigate how livestreams with rumours and/or questionable information around sensitive topics are shared on various platforms.
You want to know that too, don’t you? Then join this research. Sign up with Pieter or ask him your question first.