CORDIS - EU research results
CORDIS

The INTERnAtional network on Crisis Translation

Periodic Reporting for period 2 - INTERACT (The INTERnAtional network on Crisis Translation)

Reporting period: 2019-04-01 to 2020-03-31

Access to timely and accurate information is known to be of crucial importance for disaster prevention and management. It has been proposed that timely and accurate information in humanitarian response is a human right and as important as medical supplies, food and water. There is awareness of the fact that disasters are frequently transboundary in nature and that affected populations may be multicultural and multilinguistic. Yet, the importance of the role of translation in providing this information is understudied and under recognised. The negative consequences of poor or no communication in disasters may include increased injury or death. Lack of communication can also mean increased vulnerability of minority communities. Recognising the need for translation will reduce these negative consequences. By carrying out collaborative research and training, the International Network in Crisis Translation sought to address this gap so that translation of information is a default feature in disaster prevention and management.
"To address the objectives, five main areas of collaboration have been identified: Policy, Simplification, Machine Translation, Training and Ethics.

Our work on “Policy” has involved an analysis of national and international policy frameworks for recognition of good practice and gaps in relation to the provision of translation and interpreting in all stages of the disaster cycle. We have published a journal article on one part of this analysis and we have produced a public report with recommendations, available on the project website.

Simplification refers to the option of not translating content, but making it more ‘plain’ or ‘simple’ so that it can be understood by limited proficiency readers. This work has a particular focus on health communication, the processes and technology involved in simplifying it, as well as measuring the impact of simplification by focusing on readability, comprehension and machine translatabilty of the simplified content. We have produced several conference presentations, a journal article and have published some recommendations on simplification on the project website.

Machine Translation (MT) is an improving, but not flawless technology. In disaster settings, the challenge is that there may be a sudden onset of the disaster, for example in an earthquake, the language requirements are not known in advance, and often include translation between languages that are not typically the focus of MT development (e.g. English to Haitian Creole). Our work in this area has involved researching techniques that might be useful to help overcome such challenges. In particular, we are interested in the effect of pivoting from one language to another and then into a third language. We have produced conference papers and prototype MT engines for research purposes. We have produced conference papers and a journal article on this topic.

In any disaster, volunteers play an invaluable role. This is true also of linguistic volunteers who may offer to translate in response to information needs. Such volunteers may not have any formal translation training. We are producing materials to help train ‘Citizen Translators’. This material is made up of short, 101 type content covering the basics of translating and of editing machine translated content. We have produced articles evaluating the course content and have made several presentations on this topic. We have delivered open course content called ""Translation 101"" and ""Machine Translation and Post-Editing 101"", which is available via YouTube.

Every aspect of the work mentioned above raises ethical issues: For example, what are the ethical implications if national and international entities continue to ignore language as a human right in humanitarian and disaster settings? Is simplification of content adequate, or does it still exclude some communities? Is it ethical to broadcast potentially poor quality MT in disaster response? Is it ethical to encourage untrained translators and interpreters to contribute to the production of crucial disaster-related information? These questions are just the tip of the iceberg when it comes to considering the ethical dimension of multicultural and multilinguistic disaster communication. We have addressed some of these questions and have one published journal article and public recommendations on ethical considerations."
"The topic of translation of disaster information is under researched and under recognised. By focusing on policy, simplification, MT, training and ethics, the International Network in Crisis Translation has raised much needed awareness among governmental and non-governmental entities, professional fields such as translation and emergency response, as well as across established academic communities such as Translation Studies, Natural Language Processing, Disaster Studies and Ethics. This is evidenced by the number of conference presentations we have made, the number of invited lectures, an edited volume and special issue of a journal on the topic and, not least, our engagement with public stakeholders, such as Oxfam and Red Cross. Awareness raising is an important aspect for a topic that has received little attention to date. The INTERACT project as raised awareness globally of the importance of translating essential information before, during, and after a crisis. For example, we worked with the New Zealand Red Cross to provide training to community translators who translated their ""Earthquake Readiness Guide"" into 15 additional languages. New Zealand Red Cross also approached members of the project team more recently, asking us to provide more training for community translators to respond to the Covid-19 crisis. This additional training is under preparation as we prepare the final project report (but lies outside the timeframe for the project). Furthermore, the team was contacted by officials in Wuhan asking for advice on the provision of translation to foreign nationals at the outbreak of the Covid-19 emergency. Though we cannot claim a direct link, the addition of a line into Ireland's Risk Management Strategy in 2018 to take account of foreign nationals living in Ireland hopefully also raised awareness and led to the provision of Covid-19 information in multiple languages by the Irish government. We contributed to lobbying for inclusion of Irish sign language in all Irish government broadcasts since the Covid-19 emergency commenced. The project has produced open course material for helping to train community translators and post-editors to contribute during crisis-related events. There have been over 700 views of this content since its launch in March 2020. We continue to promote our recommendations for policy makers as well as our recommendations on ethics. The most important results from this project are the following:

- an increased awareness of the need for multilingual information in crisis communication, from North to South
- an established academic field called ""Crisis Translation"" that did not exist prior to this project
- creation of a cross-disciplinary network of expertise that did not exist prior to the project
- multiple academic and public reports on the topic"
INTERACT team visiting New Zealand Red Cross for Work Package 5