What problem are you trying to solve?
Language ability and access to life-saving information are intrinsically linked: Speakers of marginalized languages are less able to receive and understand information directly.
Through its work and research, TWB has found that life-saving information is often quite literally ‘lost in translation’ for vulnerable individuals in humanitarian situations, especially women. They are less likely to read or speak a lingua franca, access technology, or communicate directly with humanitarian organizations: Gender, education, age, and disability are all contributing factors. For example, TWB research in Nigeria found that just 9% of uneducated female minority language speakers in internally displaced people’s (IDP) sites could understand written Hausa or Kanuri, the two languages chiefly used by responders. In the Rohingya refugee camps in Bangladesh, fewer than 10% of women are literate.
What is your solution to this problem?
TWB created Words of Relief in 2013 to improve two-way communications between affected populations and responders during and after emergencies; in 2017, TWB received a scale grant from the Humanitarian Innovation Fund to increase services and reach. The innovation solution includes:
- Building knowledge: Research and comprehension testing; sharing the findings to raise awareness of communication needs of affected people
- Building capacity: Training new translators for oral and very marginalized languages, training field staff in the basics of interpreting, supporting enumerators to gather data with regard for language, developing multilingual terminology and other resources to support two-way communication on disease prevention and provision of services.
- Facilitating communication: Translating humanitarian information, education and communication materials in a wide range of text and audio formats to reach the widest possible audience in local languages.
What is your latest update on your innovation?
In the past year, TWB has integrated Words of Relief to support two major humanitarian responses: the Rohingya refugee response in Bangladesh and the IDP support in northeast Nigeria. In those programs, more than 277,000 words have been translated from English to Hausa, Bangla, Kanuri, Rohingya and Chittagonian, and sector-specific glossaries have been developed to support interpreters and partner programs.
In the coming year, TWB plans to take the next big step by building and using machine translation engines in marginalized languages, truly giving communities control of information access through text and audio technologies that enable faster, self-directed communication in the right language and format. This step will harness the expertise of commercial text/audio machine learning and apply it to marginalized languages spoken by millions of people. It will make response more effective and accountable by enabling two-way communication and content delivery, online and offline.