Training health translators from scratch

April 2, 2012, wasn’t just another hot day in Nairobi, Kenya. It was the day Translators without Borders (TWB) started the pilot test site of its Health Translators’ Training Program, to train a team of Kenyans how to translate health information into Swahili. You may wonder “Isn’t TWB the web-based platform that facilitates the work professional translators volunteer to do for humanitarian organizations? So why training?” Good questions, so let me explain.

TWB is indeed running such a platform, and it works extremely well. It was developed, donated and managed by ProZ. At the end of June, TWB had translated a total of well over four million words (with a “street value” of $800,000) since we started counting a year ago. These translations were done by 682 volunteer translators, for 84 humanitarian organizations, and 77 language combinations were concerned. For most European languages we have no problem at all in organizing this free service. However, for many other language combinations this is not the case. In some countries there are hardly any translators around, and those who are available may not be able to afford donating their time. This is especially true in areas that suffer from the unfortunate mix of poverty, a heavy health burden, a shortage of doctors, nurses and hospitals, and no translation infrastructure. In those areas health information is crucial, and simple instructions about, for example, what to do when your baby has diarrhea, can even be lifesaving. However, even if health information is available, it is often in the wrong language. English is a perfect example of such a wrong language. “Why’s that,” you may wonder, “when so many people do speak English?” Well, that is the point: many do speak English, but many others don’t. And for people who don’t speak the language, English health information is as effective as no health information. Simple as that!


Health information in the wrong language

A perfect illustration of this problem is Kenya, a country in East Africa with 43 million inhabitants. The official languages are English and Swahili, and Ethnologue reports a total of 69 local languages spoken there. Around half of the population speaks English, which is great, but it means that the other half doesn’t. Most health information is in English. So, here is the problem in Kenya: there are not enough doctors to treat patients, so health information in such a setting is crucial to bridge that gap. Because health information is available in English, it’s in the wrong language for over 20 million people.

A logical solution would be to go through Swahili, the lingua franca spoken by 60-80 million people across nine countries in East Africa, including Kenya. Kenyans who have gone to elementary school have learned Swahili, and more people speak Swahili than English. So why isn’t the health literacy gap solved by translating all health information into Swahili? Another good question!

The Kenyan Ministry of Public Health and Sanitation estimates that only between 5% and 10% of all health information is available in Swahili. There are several reasons for this. In the past there was clearly a lack of understanding that translation is important, and to a certain extent this is still the case. On top of that there is the eternal lack of budget. And finally, there is the lack of translators. Most professional translators in Kenya translate into languages such as French or Russian, not into Swahili or into their own local language. And maybe they prefer to work for big Western companies and not for the government, fearing that they may be poorly paid, and late. Whatever the cause, the lack of local translators is a problem, and TWB decided that creating translation capacity would help solve this.

Under the motto “We no longer accept that people suffer or die because of lack of translation,” we decided to train people to do this type of work. And it does not concern complex information: no instructions for brain surgeons, no high-end medical equipment or anything like that. No, it is the easy-to-understand information for ordinary people that keeps them from getting infectious diseases, and from unwanted pregnancies or unsafe abortions. Such information is written for the average person, and therefore translators should not have any problems understanding it.


Pilot test

We concluded that the pilot was successful, although we did have some surprises and there were the obvious start-up problems. One of these was that the training location offered to us by the ministry was not ready in time, and by the time it was ready, we could only effectively use it for five of the 20 course days. The location we temporarily used was a run-down classroom that was actually very charming and pleasant to be in, but the downside was that we could only use two PCs rather than the 25 we had locally purchased for the training. 

Another problem concerned the recruitment of trainees, which was done by an organization close to the ministry. Several participants proved to be very talented indeed, but there were others we do not expect to become the type of translator we have in mind. Also, there were a lot of them, over a hundred, which meant that we had to spread our efforts over many people. We gave the course to six groups of between ten and 25 persons. With people who never translated before, you can, of course, barely scratch the surface of what translation is all about when you have only three or four days. One of the lessons we learned in this respect is that it seems to be better to recruit people with strong language skills and work with them for a longer period of time.

One total surprise was that in Kenya people expect to be paid when they follow a seminar, course or training. Apparently this custom was introduced by large organizations such as the United Nations and the World Health Organization. These organizations initially paid the bus fare for participants, then also a bit of lunch, then more money, then a bigger lunch and so forth. The concept of paying people to follow our (free) course was so foreign to us that we initially refused to go along with it. Surely, we were prepared to pay bus fare to people without an income, but not pay people with a job. However, it turned out that not paying the allowance is taken as an insult. Eventually we did pay, but small amounts and therefore not everybody was happy.

During the recruitment for the August sessions we made it clear that we offer a free course and that this means “You don’t pay us, but we don’t pay you,” apart from bus fare and lunch, and that worked out fine. Trainees saw the value of the course, and the opportunity it offers to enhance their skills and learn a new profession. During the pilot in April and again in August, working with highly talented and motivated trainees was a great pleasure. In April, one of the highlights was the three-day training mission in Olosho-Oibor, a village in Massaai land. We worked with a dozen school teachers, a social worker, the dispensary manager and a few others. The dedication of this team was overwhelming and this group now takes care of the Massaai versions of health material that TWB in Kenya is asked to translate. In September, our course instructor will go back there to continue their training.

The structure of the course and the course materials proved to work in April, although when preparing for the August sessions we revised some of the materials and we added an extensive two-day module on using translationmemory (TM) tools, as well as a few other modules.


The training program

Our course program starts with a half day introduction to translation, followed by a series of medical modules, translation exercises to be done in small subgroups, and class-wide discussions about these exercises. In certain respects it follows the concept of the MediLingua course Medical-Pharmaceutical Translation, although it’s simplified for people without experience as translators. Local course instructor Paul Warambo and I each present parts of the introductory module, which includes translation methods, tips and tricks, an introduction to TM tools, and how to build and maintain glossaries. We also discuss the difference between translation and interpretation, give an introduction to subtitling, and provide basic instructions about word counting and spell checking, as well as on how to Skype and how to use search engines. Part of the course is dedicated to quality: how to write clear and error-free sentences, and how to check correct punctuation. I present the medical modules and Paul, who holds a master’s degree in Swahili and translation studies from the University of Nairobi, focuses on issues around Swahili. In April he did a great job in leading the evaluation discussions.  

The medical modules focus on a range of Africa-relevant health problems. These are mostly disorders, such as pneumonia, diarrhea, malaria and cholera, but also social health issues, such as malnutrition, unsafe abortion and female genital mutilation. Each of these modules take 30-45 minutes to teach and are followed by an exercise: participants translate a few sentences from a health information sheet and the results are projected on the screen and then discussed by the whole group. In April this proved to be a very powerful method and participants really seemed to learn a lot from these discussions. It was also remarkable how people who had never translated before started behaving like typical translators during the first exercise, having heated debates about the meaning of a specific word, or the proper location of a comma.

For the current edition of the program we have added a much more extensive TM tools module, which was not yet relevant in April, and we also expanded the module on subtitling health videos. During the April sessions we coached the translators in subtitling a health video on the prevention of cholera into Swahili, Masaai and the Sheng dialect widely spoken by young people. We used the dotSUB technology for this, which was generously donated to us, and what came out of it was the first-ever health video to be subtitled for Masaai and Sheng.

Somehow our training program attracted quite a bit of interest from the media, which can bode well for a start-up program. The Voice of America created a news item about our training program, and The Guardian carried an interesting article about us on their website, as well as in the paper edition. An enthusiastic BBC World Service crew followed us for a day. On June 6 at Localization World in Paris, a 13-minute video was shown which can be seen on the TWB channel on YouTube. In July and August we returned to Nairobi to continue the program and to work with a group of 20 people, and coach them into becoming members of our local translation team. We recruited participants with excellent English and Swahili skills. One third had a language background (recent Swahili graduates from the University of Nairobi), one third had a recent degree in Health Promotion (who are trained to generate health information) and one third had varying backgrounds. These participants are extremely motivated, and the complexity level of the exercises could already be increased after week 1. We expect that many of these trainees will be able to function as a translator on our team. That is good, as some humanitarian organizations and health information publishers already expressed their interest in working with this trained team, and we have been discussing partnerships with several organizations.

For the July/August sessions, we had moved our center from the compound of the ministry of Health to the campus of the East Africa Bible Translation Center of BTL, who are also hosting SIL (the publishers of the Ethnologue). Everybody on the BTL campus is somehow involved in translation, which makes it a rather unique location for us to be, and we feel very much at home there. During the guest lectures that several BTL and SIL staff have given, it was illustrated that the challenges of reaching equivalence between two language versions of the Bible or of health information are actually quite similar. For the rest of the year, the TWB Healthcare Translators’ Training Center will stay on this campus. Our training concept is working, and we are looking at making it available to other regions of the world that, like East Africa, have poor public health, an inadequate health system and hardly any translators around.