Liberia: Building Trust in Tools

Some information in this post has been updated. Occasionally, in partnership with others, Ushahidi takes on projects that help test certain applications of the platform in order to improve the tool. Ushahidi will also take on projects on a consultancy basis from time to time; in the unique instance of Liberia, I will be acting as a consultant for Ushahidi, working with partner organizations on the ground to implement Ushahidi as each partner sees fit.  Over the next six months, I'll be checking in regularly on this blog to provide updates on how Ushahidi is being used in the Liberian context by local organizations and networks of early warning actors. ________________________________________________________________________________

In my first week back on Liberian soil, I've been (re)learning the complexity of this simple fact: Ushahidi is a tool.  It's a tool defined by versatility - a Swiss Army knife of sorts that can serve multiple purposes and offers many useful approaches to a range of challenges.  But here's the funny thing about tools - they only work in collaboration with others tools and in cooperation with their surroundings; a hammer, for example, is only effective if it has a nail, and that nail will only hold if it has a solid piece of wood or sturdy wall. Now if we imagine that Ushahidi is the Swiss Army knife of communication and information sharing, one of the most important components of this ecosystem is trust.  Who do you trust to deliver information, where do you go to share it, and what mechanisms (oral, written, SMS) of transmission do you rely upon?  If any one of these links in the chain - the mechanism for transmission, the individual/institution delivering the information - isn't trusted, the tools that enabled the info sharing will not fulfill their purpose - and may themselves become distrusted parts of the whole. Last Wednesday morning, John Etherton (this project's technical support manager) and I met with members of the Carter Center and IREX to discuss this delicate balance of trust, new technology and conflict mediation.  Lofa County was our focal point, as new reports are now being published on the complex series of events that led to widespread violence last February.  A recap: What began as the murder of a young girl led to an angry mob seeking vengeance, a mob that blamed the local imam with strong conviction and little evidence; with plans to storm the imam's mosque, the crowd was deterred (Pakistani peacekeepers were nearby) and decided instead to call (that's right, on the mobile) one of the mosque's regular attendants to tell him, 'we are burning down your mosque.'  The beginning of a rumor.  Without verifying if this information was true (and it was not), the receiver of this call formed a retaliatory mob, that actually did burn down nearby churches and Christian businesses.  In the process, several people were killed.  Preexisting ethnic and religious tensions (the Lorma people are primarily Christian; the Mandingo, Muslim) cannot be underestimated as fuel for the fire.  What we also agreed during our Wednesday meeting was that the use of mobile phones played a critical part in this information-sharing wildfire. What we need, said one member of our meeting, is a counter-narrative; there has to be a trusted source of information that is providing factual information to counter the rumors.  Who would this be, and how would they deliver their message?  This is where Ushahidi could come in, using Frontline SMS to send messages to members of the community when conflict is brewing - the message containing either a statement from a local authority figure or instructions on where to look for accurate information.  But we hit hurdle number 1: many of the citizens who live in these areas of Lofa County are illiterate.  Okay, written SMS is out.  Next up: what about a recorded voice message from the local chief?  Hurdle number 2: is the local chief the most respected authority figure?  Would an imam or priest be more likely to carry weight in these communities?  The latter may be more relevant in the Lofa context, but because of religious divisions it would not fly in the Christian community to receive a voice recording on crisis information from an imam (hurdle number 3).  Okay, we thought, what about joint voice recordings - both the imam and priest speaking together to the community, symbolizing harmony and cooperation?  There were joint task forces and peace committees formed by imams and priests in some parts of Lofa in response to the violence, so this option is not unprecedented.* But we were forgetting to ask one crucial question about the social environment - one that would surely affect how useful any tool would be: are religious or tribal elders the authority in these communities?  Youth in Lofa were said to have kidnapped an imam and other traditional leaders during the bursts of violence; young people may not historically be authority figures, but power can be and often is usurped during periods of instability.  Where do youth fall in the delivery of this counter-narrative and within the larger power structure? The realities of using a tool out of the box often means elaborating on the clean order of its imagined purpose in the hot mess that is the changing context.  In Liberia, there are multiple languages, limited literacy, spotty cell phone coverage, chronic ethnic tensions and religious divides, and an evolving tangle of traditional leaders, emerging authority figures and the ever-present usurpers of power.  To work, Ushahidi will have to be a smart tool - in other words, one that learns (via its human operators) how to gain the trust of its users. *What I'm sharing here about the Lofa incidents is admittedly a piecemeal version - one translation - of what happened; my sources have been many over the last several months and only now are documents being released detailing what actually took place.  I will be attending a conference on the Lofa County conflict this week and intend to update any information I've written here that I later find to be inaccurate. This post has been updated.