Translators and interpreters in conflict zones: how machine translation made its way to replace human presence.
A brief evolution of machine translation and translation needs during conflict
Being a translator and interpreter in a conflict zone is one tense choice of a career. And a highly interesting one for a linguist. There are a few moments more crucial than these where understanding on both sides has a real impact on world peace.
Before, it was a 100% high-risk job. Translation would many times be done in an environment vulnerable to attacks. All people present had to be prepared on how to react under fire or havoc. There are cases where having an interpreter saved soldiers from landmines and dangerous paths.
Luckily, as time went by, machine translation made its way to replace unnecessary human presence in conflict situations. Today, a microphone can automatically translate into a screen or speaker multiple languages in real-time(a technology also used in restaurants in Japan). And if not, an earpiece with someone safely on the other side can do for the rest of the situations. The more is not the merrier in this case and being able to advance in smaller groups alleviates tension. In some cases, when negotiation is adamant, and in person, an interpreter could be necessary. But today this is a choice and not an imposition.
Whatever the means, the purpose is to ensure that the message is conveyed effectively. With most soldiers being unable to speak beyond basic phrases of foreign languages, and a shortage of interpreters, machine translation was a breakthrough.
How does this technology work and how can it improve how we communicate?
How it started
In 2006, IBM took a big leap. It delivered 35 notebook computers with voice recognition software for the medical personnel in US Special Operations forces and the US Marine Corps.
This meant that now medical situations with Iraqi security forces and citizens could be sailed through without major language complications. The technology allowed for effective communication in real-world tactical situations. It quickly translated languages that seemed immensely far away, like Iraqi Arabic.
The speech-to-speech translator had been in development since 2001. And at the time, machine translation was still far from being something that could be present in our daily lives.
So, what is Machine language translation? We’ve been through this in previous articles but we’ll recap.
It is the process of converting text from one language to another through automatic translation software. It’s also a continuous work in progress. It becomes better year by year, but demands also become more demanding. The differences in the syntax, semantics, and grammar of languages around the world make them prone to mistakes. But compared to MT technology in 2006, what we have today is incredible. Advancements have allowed machine translation to pull syntax and grammar at an unmatched speed. Not to mention readily available for anyone, and usually for free.
Regional sub-dialects, minority languages, and the translation of idioms and popular expressions are still a complication. Two different automated translation tools may produce two different results. The parameters that govern the machine translator vary. And some solutions are still hazy around the edges. But here are the three main types of systems:
1. Direct Machine Translation
The most elementary form of machine translation. Probably what that 2006 IBM was aiming at. The machine translator goes word by word, or phrase by phrase, and translates it directly. It may adjust the output based on morphology and syntax. It was a great starting point but has been pushed into the wayside by more advanced techniques.
2. Transfer-based Machine Translation
It organizes the source language’s grammar structure. Analyzes it to identify its rules. Then, the sentence structure is converted into a form that’s compatible with the target language. The result is a coherent final text.
3. Interlingual Machine Translation
Interlingual machine translation translates the text from the source language into interlingua, an artificial language developed to translate words and meanings from one language to another. Through this intermediate representation, it reaches a more accurate translation in the target language.
There are more, of course. Such as massively multilingual ongoing development, but this gives an accurate glimpse.
About the job and why kits are important
On the UN website, a translator for the United Nations, mentions how the job is unique: “The political nature of many texts requires that translators exercise special care when choosing their words since nuances and overtones are particularly important and might have a huge impact.”
Their role has developed considerably in the twenty-first century. Today their skills are razor sharp. More than ever, they are now only present in person when it can’t be avoided. Therefore when it is game-changing.
Meanwhile, associations such as the Saudi Arabian National Research and Development Organization are collaborating on the advancement of Machine-Translation technologies. King Abdulaziz City for Science and Technology (KACST) and IBM researchers are working to perform complex simulations, computational modeling and improve the translation of Arabic to other languages.
This way, the kit people count within conflict zones (and in inevitable cascade, in all other life situations) will be the best at hand. What was a ‘translation kit’ that consisted of notebooks and microphones for MT in 2006 is today software that can be used in almost any device, earpieces that automatically translate, agencies dedicated to remote interpreting, and, only if needed, in-person interpreters with the highest training for crucial situations.