Published on August 7th, 2016 | by EJC0
Geo4Nonpro: Nonproliferation Through The Crowd
Verification takes an important place in issues concerning Weapons of Mass Destruction: it is the bargaining chip in disarmament negotiations and the bane to state secrets and sovereignty – not to mention arsenals. While on-site arms inspectors face uncertainties as they inspect (suspected) nuclear facilities and disarmament researchers speculate on political motives and trade bans, a digital, open-source and crowd-sourced version of verification on the basis of commercial satellite imagery is finally underway.
It’s been a while, but we’ve been hearing this phrase again: ‘Trust, but verify’. Curiously, it was used by two sides of a political argument – the question of whether the Iran nuclear deal was, in effect, a win or a loss for the nuclear disarmament community in general and the Obama administration in particular. The use of the phrase harks all the way back to the eighties: US President Reagan was doing his best to advance relations with the USSR and idealistically envisioned a bright future (NOT lit by the flare of a nuclear detonation), but wanted to portray himself to the Americans and their Russian counterparts as always vigilant.
It’s an appealing slogan, with ring of positivism to it, and it is short and easy to remember. It’s no wonder the phrase stuck in our political communications, all the way back from the age of the Cold War. (We could go on and on the role of linguistics in politics and this phrase; do read this opinion published in The Washington Post for instance.) This is despite its unviability as a working policy: trusting someone or something while second-guessing and doubting them at the same time is impossible. The second you decide to check someone’s trustworthiness, you’ve suspended it to after the verification has had an affirmative closure. This is why journalists, of course, won’t trust enough to publish anything until or unless they’ve verified it – and classic media theory has it that the public uses journalists because they trust that they have done so. (Particularly cynical journalists – or perhaps the stereotypical hard-arsed police detective, the kind who says ‘I’m too old for this shit’ in television series – will be hard-pressed to consider anything verified if it didn’t travel down to the bureau and spit them in the face.) But I digress, the point here being basically to attest: trust is a hard thing to attain, and pretty unchangeable. It either is, or it is not. Either way, it will influence your actions.
This is despite its unviability as a working policy: trusting someone or something while second-guessing and doubting them at the same time is impossible. The second you decide to check someone’s trustworthiness, you’ve suspended it to after the verification has had an affirmative closure. This is why journalists, of course, won’t trust enough to publish anything until or unless they’ve verified it – and classic media theory has it that the public uses journalists because they trust that they have done so. (Particularly cynical journalists – or perhaps the stereotypical hard-arsed police detective, the kind who says ‘I’m too old for this shit’ in television series – will be hard-pressed to consider anything verified if it didn’t travel down to the bureau and spit them in the face.) But I digress, the point here being basically to attest: trust is a hard thing to attain, and pretty unchangeable. It either is, or it is not. Either way, it will influence your actions.
Those actions and the dynamic part of the ambiguous adage stems from the verifying bit. Verification is a process: a trajectory of many different actions and pathways towards establishing a statement’s or a person’s validity. While not necessarily a scientific method, it is an exercise whose carrying out and outcome are held to particular standards and norms. This exercise is where we have seen a shift in the last couple of years. As a method, verification can take many forms, and we’ve seen the verifiable nature of statements diverge over the years; from the political and the personal, in print but more so on social media, in tweets, pictures and video’s. Digital communication has also broadened the scope for the arrays of half-truths, grey areas, lies and propaganda messages, to be produced, communicated and spread.
And so now verification also takes place online. (Do check out the EJC’s Verification Handbook, or First Draft and other toolboxes.) It is completed by journalists but also increasingly by activists, researchers and independent watchdogs. Specialists can weigh in expert knowledge when annotating commercial and open-source satellite imagery. (The obligatory link, of course, being the Bellingcat investigations.) We’re crowd-sourcing knowledge and sentiments through social media, and in a transparent manner.
The original context for the proverb was, of course, the Cold War, in which trust was very strained in geopolitics and within societies, but also in which civil movements sourced the crowds old-school style, to raise their voice against the nuclear bomb. In a context pertaining to nuclear arsenals, some of the terms mentioned above have a specific connotation. Transparency is hard to come by considering ‘all nuclear knowledge is born secret’. And it has been agreed by nation states that it must remain so to a certain extent: as per the Nonproliferation Treaty, the knowledge of and necessary materials for the development of an atomic weapon (or the weaponization) may not travel from nuclear weapons states to have-not-states. Verification, then, very concretely refers to the monitoring and assessment of disarmament, security, and trade agreements.
Upholding verification and pushing the agenda of nonproliferation and disarmament don’t really get the traction they might’ve during the Cold War. With a holey mesh-net of treaties – ratified, signed, contested and everything in between – it’s a field of political studies that’s quite stuffy, too. But, as increasingly information finds its way onto the web and into social networks and digitally connected communities of general public and experts, verification issues on disarmament might well become the subject of crowd-sourced, open-source scrutiny.
This is precisely the thought behind Geo4NonPro, a new platform that was launched at the end of May. It was set up up by researchers of WMD nonproliferation and disarmament at the James Martin Center for Nonproliferation Studies (Monterey, CA). In a blog on the Arms Control Wonk website, Senior Research Associate Melissa Hanham introduced Geo4NonPro as a “’curated’ crowd of experts who will annotate and comment on satellite imagery” to “see if the crowd does better than the in-house experts”.
In an e-mail, Hanham set out the idea for the platform: experts – from fields in WMD disarmament, to architects, political scientists and more – are invited to join the website, where they will be given a new set of tasks each month: to annotate and analyse satellite imagery relating to nonproliferation verification issues. She wrote: “I’ve always been fascinated by satellite imagery, and have been an avid user of Google Earth. Commercial satellite imagery is already used to validate verification claims by both international organizations and governments. Crowdsourcing of satellite imagery has happened in a few other areas like Tomnod‘s varying campaigns, but to my knowledge it hasn’t been used to answer WMD verification questions.” So this is wat the CNS set out to do: they “wanted to see if we could incentivize a crowd to look at satellite images and make meaningful contributions around verification questions. This is still an area that machine learning is rough in, but it’s coming!”
The interaction and the crowd are being divided in two sections for this programme: while the experts are asked – and in some cases the research institute will put out specific calls, looking for experts whose knowledge is particularly useful for a task or whose insight has been lacking in the analysis – to aid in the online verification, the broader public can watch the verification discussion unfold on the public-facing website. Through social media, the team ask experts to answer their specific questions:
When we spoke, Geo4NonPro had launched a week earlier. According to Melissa Hanham (@mhanham), some 100 experts had signed up by then. “We are still looking for more people. For example, we need a specialist in HVAC/ventilation systems right now. Mostly, we want to get people active on the site. We’ve had a lot of signups, but just a few annotations so far. We’re hoping to break the ice a bit to get people energized. here is the possibility that we will receive incorrect information or misinformation, but that is all part of the experiment.” That is also why the James Martin Center researching teams do conduct their own verification research while the imagery tasks are up online.”
A set of how-to videos have since been published on YouTube:
The Geo4NonPro-platform is now a test for the WMD disarmament community, and the first of its kind.
“CNS will keep the website until at least the end of the year. We are working on preparing imagery for Novaya Zemlya in Russia next and then Sinpo in North Korea. If breaking news such as a claim regarding the Iran Deal unfolds, we’ll trying to get imagery up quickly for analysis. Not all verification tasks are conducive to the use of satellite imagery, so we’re going to play it by ear. I’m also interested in using non-optical imagery wherever relevant. I hope to incorporate NIR, SWIR, and TI bands of light.”
About the Author: