,
15 tweets,
3 min read
Read on Twitter
Thread. So Twitter is starting a research line of inquiry into whether their site is useful for DE-radicalization. Let's explore. theverge.com/2019/5/29/1864… 
To study this, Twitter will ask how social contamination happens. They'll decide whether they're trying to stop information flow or whether they're trying to change behavior.
Stopping information flow can be modeled as a "simple contagion" requiring just one exposure, whereas changing behavior is called a "complex contagion" requiring multiple exposures.
For example, seeing a meme is a simple contagion. At first you were not infected (didn't see it), then you were infected (you saw the meme). Information transmitted. Same with fake news, disinformation, etc.
The current network theory literature says that a complex contagion (promoting a neo-Nazi rally, making your own Hitler memes, stopping a belief that whites are actually supreme, believing that bleach cures autism, etc) takes multiple exposures.
The work being done by researchers in this field involves - in part - understanding the most effective network structure for spreading a complex contagion. This is a very open question and has many smart people working hard.
The questions researchers ask include: What is the shape of the network that works best to transmit each type of contagion? What are the relationships between the people like? What else happens during exposure?
To answer this, some researchers study known, "real" networks (Al Quaeda, ISIS, white supremacists) to learn how behavior change happens. I'm kind of in this camp. Here's another example by someone much smarter than I am: arxiv.org/abs/1701.08170
Other researchers are building experimental networks, like petrie dishes, just to study behavior change in a controlled environment. Example, again from a top researcher: undark.org/article/spread…
In short, both radicalization and de-radicalization involve swapping one set of behaviors for another.
That's why I was confused about this announcement, because Twitter seems to be claiming their site might be useful to spread one type of complex contagion (de-radicalization) but not for another (radicalization). 🤔
Side note: I hope Twitter doesn't go down the path of experimenting on its own users, like Facebook infamously did/does. makeuseof.com/tag/facebook-s…
Anyhow, I look forward to reading the paper that comes out of this. My guess is that should come out in the mid-2020s. That should be fine, right? Nothing big or important going on for the next half-decade. 🙄
In the meantime, I'd encourage Twitter to ask: Why do trolls and haters come back over and over and over again to this site? Why do they so desperately want to be here?
And most importantly: What is Twitter doing differently, now, today, while we wait for this research to yield fruit?
