Thread by @sophiecdyer

🧵It's #InternationalWomensDay and given the uptick in media coverage of #digitalsleuths: Let’s talk about representation in open source research!

When we use open sources to report on war, whose voices, experiences are we amplifying?

#Russia #Ukraine #OSINT #OSINTUkraine

Like most groups, @amnesty is publishing analysis partly based on open sources from #twitter as well as Telegram, Facebook, VKontakte, etc. For us the focus has is on possible war crimes.

E.g. this thread by the Crisis Evidence Lab ↓

https://twitter.com/amnesty/status/1499357512193220610?s=20&t=pflLjVZMYKxGaE9nGIAkMA

Social media is a critical source (great!) but it privileges certain types of violence. Violence that is less spectacular e.g. dehydration, starvation, lack of access to medical supplies and electricity, or racism.

HT @bobtrafford 🙏🏻
hrw.org/news/2022/03/0…

We know, too, from past conflicts that more intimate forms of violence, e.g. sexual and gender-based violence (SGBV), can be over looked by #OSINT researchers and investigators.

See this paper by @KAlexaKoenig and peers ↓
academic.oup.com/jicj/article/1…

There is also a digital divide: older people + people w/ disabilities are less likely to be online. As a result, the experiences of these groups are not well represented by #OSINT

See @RawyaRageh easy-to-read on living w/ disabilities in armed conflict ↓amnesty.org/en/latest/news…

Same for young children, with good reason. On the flip side, photos of children in war zones have a special currency in war reporting.

How can we do better to include them and their experiences in research and reporting? 🚸

The identity of the researcher matters, too. Their language skills, gender + location can all be multipliers of bias.

Still shocking few womxn are visible in #OSINT …or not so shocking considering the harassment womxn experience online.

👀 @manisha_bot is researching this!

Because of these factors, while crucial to war reporting, #OSINT generates narratives in which certain types of violence + demographics are over represented – while others are conspicuously absent.

For a how-to on mitigating some bias see ↓ by @m_paquett
citizenevidence.org/2021/03/12/gui…

Ultimately, #opensource investigations are about power relations and how to dismantle them …so too is #feminism!

Gabi Ivens of @hrw (who helped me write this thread 💓) and I wrote a paper on this → files.cargocollective.com/c949312/What-W…

So did @libbyjmcavoy →
openglobalrights.org/centering-the-…

As a team we've spoken lots about maps… Maps are compelling but despite the best caveats they exude completeness. The commanding "view from above" is seductive! Even when what’s being mapped is partial.

See #DataFeminism by @kanarinka + @laurenfklein ↓
data-feminism.mitpress.mit.edu/pub/ei7cogfn

Plus, maps feel neutral. Alert! 🚨

This is an interesting thread about how many conflict maps that show the advance of the Russian military may actually be projecting the view Putin wants us to see.

https://twitter.com/Calthalas/status/1498998318755680260?s=20&t=zniaWNbqHOAQvGkjgr9cVQ

To sum up, #opensource and #OSINT offers racial possibilities but let's keep talking about #representation and #power in our work, let's shine a light on the limits of our research methods, and finding inclusive and imaginative of ways to communicating our findings.

#Ukraine

This unravelling thread is dedicated all the womxn in Crisis Response @amnesty – esp. the Evidence Lab crew: @milena_iul, @AI_Micah, and @rayadamsrowfarr – and all of our allies.

Thanks again to @bobtrafford + Gabi Ivens for feeding into its writing.

*radical