Found in Translation

Helping Rescuers in Haiti
While translation helps prevent some disasters, there are others that are harder
to foresee. In times of such emergencies, spoken language communication is
vital. The ability to say something as simple as “If you hear me, knock three
times” in another language can mean the difference between locating a
survivor or leaving that person behind—especially someone who is unable to
speak due to extreme pain, fatigue, or hunger. Time is of the essence in a relief
or recovery situation, but language barriers slow things down. For domestic
and international disasters alike, language can make or break rescue efforts.
Thankfully, technology is changing the speed and ease with which vital
language support can be accessed.
After the earthquake struck Haiti on January 12, 2010, virtually all lines of
communication between the affected areas and the rest of the world went down,
with one exception: the transmission of SMS, or text messages. The existing
emergency response number failed, but Digicel, the largest mobile
telecommunications company in the Caribbean, immediately made a free
phone number available to a group of organizations that set up an information-
sharing platform. Technical solutions for capturing and channeling the
thousands of messages were quickly in place, but the majority of the messages
were in Haitian Creole, a language unknown to most responders. The
relatively few professional translators in the area were already completely

---

overwhelmed by other responses and were unable to handle this onslaught of
additional translation tasks.
Enter Rob Munro, a linguist and graduate fellow at Stanford who had been
developing methods for processing large volumes of SMS text messages in
less-common languages. He’d also been working on crowd-sourcing projects.
These two distinct specialties became the perfect combination for a new project
called Mission 4636, which was named after the number of the free phone line
the individuals used to communicate. Munro went about setting up a team for
the task. In the first week alone, he assembled more than a thousand volunteers
from a total of forty-nine different countries.
An online chat room served as both the orientation venue for newly joining
volunteers and as a platform for translators to communicate with each other
and their coordinators. The online collaboration was critical. Many of the
messages received were full of colloquialisms that required further discussion,
and many translators had specific local knowledge that they needed to share
with rescuers on the ground. Mapping data were also embedded, so that
translators could apply their local expertise to remotely help responders
generate the exact coordinates for a given (written) location and respond
appropriately at the scene. Here is an actual text message exchange between
translators and rescuers: