Steve Stottlemyre from the Office of Intelligence & Threat Analysis, U.S. Department of State and Sonia Stottlemyre from Georgetown Public Policy Institute, have recently published an article titled “Crisis Mapping Intelligence Information during the Libyan Civil War: An Exploratory Case Study”. The article touches on one of the most interesting topic related to crisis mapping applied to civil unrests or conflict settings, and has some very good points in it, as well as some very big mistakes and misrepresentations. The article is definitely worth a reading though!
Let start with the good points:
1. The way Twitter users fused crowd-sourced data during the Libya War resulted in the creation of tactical military intelligence. This is indeed a super interesting matter, that in a way leads us to a broader discussion about the fact that tasks performed by hierarchical centralized systems may now be taken on by networked decentralized systems, in a way that may (or may not) lead to the same outcomes. What this means is also that crowdsourcing is creating a new way to process information that before was only possible by organizations that had the means and the money to do it. These new processes are so decentralized and embedded inside such a huge network (the Internet) that may be able to reach the velocity and accuracy of the centralized hierarchical systems and gain value because completely free and much larger in volume.
2. During the Libya Crisis War it was members of the “crowd” who planned and directed collection efforts, and established operating procedures for intelligence operations. This point is in a way very similar to what this article is formulating when talking about the “self-regulation” of Twitter. The social media space seems to be more and more described as a self-regulating and self organizing environment more than an “anarchy”. Balances and collective planning seem to happen in this space even with the complete lack of a unique authority to direct it.
3. Twitter provides both a platform for reporting information, and much of the infrastructure required to convert information into intelligence. The power of Twitter in this sense is incredible and undeniable: this platform I think has by far exceeded the expectations of the same people that created it. More than Facebook has done, or if you want in a different way, Twitter is being used for a range of tasks that all together make it one of the best real-time coordination tool. The division of tasks and the consequent combination of its component can be done, again, in a decentralized, real- time dimension, while the vetting of the information combined is distributed to the all network.
4. Twitter acted as a platform for collaboration on and compilation of intelligence products. Again, the inherent structure that Twitter has and the use of hastags makes is a very efficient curating system. What is happening is that this curation process is being done collectively and intelligently in real-time, making it possible not only to access already curated information, but also to have a sort of continuous verification/vetting system that constantly reiterate itself.
So, let’s make it clear here: people creating crisis maps and people using social media were and are creating intelligence. This was true for the Libya war, as it is for ANY crisis mapping deployment or social media coverage of an event. The first time I had a conversation about this, it was with Heather Blanchard, co-founder of Crisis Commons, in 2010, discussing about PakReport, a crisis mapping deployment in Pakistan. A big kudos to the authors of this paper to have reached the same conclusion 3 years later!!! 🙂
Now, let’s go to some of the major mistakes in this paper and to some of the weird points that the authors make in an attempt to prove their thesis.
Vocabulary and Glossary mistakes
1. Crisis Mapping is not equal to Twitting. Unfortunately it looks like the authors of this article are a bit confused about the vocabulary they use, when they attribute to people twitting the definition of “crisis mappers” and the contrary. For example, the Libya Crisis Map was clearly a crisis mapping effort, but did never engaged in active twitting, while people twitting were not necessarily the same one creating crisis maps out of twitter messages, even if involved in the curation of the data that was subsequently mapped. I suggest the authors to read this blog post to learn more about this.
2. Hashtags do not equal @. This is the most disconcerting mistake in this paper, namely because you would think that people writing a paper about social media would have done their due diligence work in understanding how social media work. In the paper the authors infer that people using the hashtag #NATO wanted to address the information in that tweet to NATO. At the contrary though, people use hashtags to underline a topic, or a specific actor involved in the action reported. For this reason if I am tweeting that the US have just passed a law on cybercrime, I will add an hashtag to US and one to cybercrime, but the reason why I do it, it’ s because I want people interested in the cybercrime topic to find that information, as well as people looking for information about the US. The very interesting part of this mistake in understanding how Twitter works is linked directly to the intentionality that the authors want to attach to everyone that used the hashtag NATO. In fact the thesis they are trying to support is that everyone that used the hashtag NATO wanted to actively pass information to NATO.
The missing point here is that as much as we can assume that the NATO was following its hashtag, we can infer that the rebel groups would have done that too, as well as the Gaddafi forces, as well as the media, as well as everyone on the Internet that wanted to see what the NATO was doing during that period of time. This of course is a very different issue than the Twitter messages that had @NATO or @NATOPress, since this was indeed a way to make sure that the specific accounts were getting that information. Without going into details about the actual intentions behind the willingness of people adding the @NATO to their tweets, those two groups cannot be merged together, nor can their motivations.
As one of the person managing the Libya Crisis Map project I have to say that I am definitely pissed off by one main factual mistake done in this paper.
The mistake is about why and when the volunteers working in the project where reporting about military operations happening in Libya. Interesting enough the assumption that the authors make is that we started reporting about military operations because we wanted to actively support the NATO Operations in Libya. What looks strange is that the Stottlemyres did not connected the fact that military related reports were increasing in the platform to the fact that military actions overall were increasing in number – since there was a new actor in the battle field, and namely NATO. The second factor that they seem to ignore is that NATO military operations were much more visible and reported than military operations done by the rebel groups or the Gaddafi soldiers.
In addition to this, the authors seem again to ignore a very important point here: if we are assuming that the Libya Crisis Map was reporting more military related information to support the NATO , why not to support the rebels? or the Gaddafi soldiers? as the internet is accesible to everyone in the same way, no causality can be drawn by the simple fact that more information was reported.
In addition to this, what really strikes me in this strong tentative to accuse 300 volunteers of wanting to support a military operation that caused thousand of victims, in a very complex emergency, is the subtle idea that the Standby Task Force is a unique body composed of people that are all politically aligned, or, the even more annoying idea, that we, as the Core team, could have been instructed people to search and publish information specifically for the purpose of supporting NATO operations in Libya. Both those scenarios are not only unrealistic but also offensive.
To be added to this is the simple fact that the SBTF was acting under the activation of UNOCHA, one of the most independent and imparcial body of the UN. In fact, if the authors of this paper would have taken some more time to actually support their accusations, they would have seen that once UNOCHA took full control of the deployment, the reporting of military related operations stopped entirely. One of the main reasons why this happened was because, while in the beginning of the deployment the SBTF was mandated to give UNOCHA an overall idea of what was happening, and what was the humanitarian situation, in the second part of the deployment, when UNOCHA had more information coming from the ground, the focus switched to more attention to the provision of humanitarian relief.
Let me also add something else here: in order for the SBTF to give an overview of what was happening on the ground in Libya, the location and intensity of the combat was indeed a very useful information. For example knowing that Benghazi was under attack for days, and that the port was blocked, was indeed valuable to infer that civilians would have been in need of water and food, and that they most likely would have been trying to run away, causing an influx of IDPs and refugees in other areas. All in all, saying that combat information do not have a relationship with humanitarian needs is like to say that hunger has no relationship with availability of food.
Specifically in this instance, I would like to think that the authors of this paper are rather ignorant than to think that they are intentionally trying to accuse the SBTF to be an ally of NATO, which would not only be malicious but also dangerous for some of our volunteers, that live in Libya and could be subject to repercussions due to those accusations.
A piece of advice for the authors, is also to remove the twitter account names from the article, especially when they are publicly accusing those people of being NATO supporters.
Totally not supported assumptions
As I said before, this paper is a very good piece of research when it comes to the relationship in between the military intelligence process as done by the army, and the same process as done by the collectivity in the social media or in crisis mapping projects. I find this topic extremely fascinating and definitely in need of more research and possibly in depth research. What is very curious about this paper is the decision of the authors to infer intentionality for everyone using social media or doing mapping to support the NATO. I have been reading this paper over and over and I cannot find a good reason for the authors to add element this to their paper.
In addition to this, what is extremely curious is that, while in the paper the authors repeatedly use several reasons to support the argument of intentionality to support the NATO (i.e. the use of the # NATO or the number of tweets or maps monitoring the military operations), they also specify in the conclusions that “we cannot precisely extrapolate the motivation of crisis mappers who created finished intelligence products, nor can we determine how responsive crisis mappers would be to official PIR and RFI issued by military commanders.” It definitely looks like some confusion is going on there, but we can also notice that there is a huge stretch in the tentative to infer intentionality by using an argument that could be used as well as to infer intentionality to support the Gaddafi soldiers or the rebel groups, or whoever was in the field at the time.
The main questions were basically not asked
This paper is just scratching the surface of the real issues. What is a shame is that the authors of the paper did not asked the right questions, as if they did not want to actually find out the two main issues:
1) If the military was actually really using the data produced by crisis mappers or social media;
2) and related to that, if the data produced was of any additional value to what the military already had.
In this matter, clearly the authors of the paper did not noticed that the Libya Crisis Map, for example, had a delay of 24 hours – meaning that data posted one day on social media would have only been visible and usable the day after in the Libya Crisis Map. Would that data be of any value for military purposes?
The authors themselves say that “Public information is unavailable about the extent to which military commanders used information from crisis maps during the Libyan Civil War. Nevertheless, commanders had access to such information, and likely used intelligence products derived, at least in part, from information pulled from social networking websites.” Forgetting for a moment that this is stating the obvious, since it would make no sense to even think that military do not look at social media data, the actual question for me is to which extend, and how, this data was used by the military, if it was ever.
Hoping that someone competent in the issue will take this topic on, I would love to understand this and know more about how the collectivity is being (or not) more reliable, fast and articulated than the military is in creating intelligence.