In the modern world, with smartphones and apps to help with everything from hailing a taxi within 30 seconds and counting the calories of your latest lunch, it is easy to just type or paste some words into Google Translate when you need a translation and away you go. Seconds later you can be asking the German woman in the ticket office how to get into the centre of town, or asking for food recommendations from your Japanese host. Though Google Translate and other machine translations are a useful tool for single and simple use and conversational phrases with slight errors thrown in, translation as a human process is a skill to be cherished and utilised where necessary.
Specialist translation experts have built up a long-standing skillset to understand exactly how to translate between languages based on their own experiences and memory. But all expert and professional translation companies worth their salt will utilise human knowledge alongside CAT, or Computer Aided Translation, tools in order to be as efficient and productive as possible during each translation project they are working on.
The tools that make up Computer Aided Translation are based on translations that have been made through human process. So over time a Translation Memory (TM) will be loaded with new translations until there is a bank of memory to fall back on. This is a vital and efficient support system for translators to help them check on the consistency of certain terminology. When there are huge quantities of text to translate, or terminology relating to support, services and products for business purposes, it can come in very handy. Professional translators can utilise CAT tools to make the most out of translations that have been human made in the past, speeding up the whole process without diminishing the quality of the end product.
A machine translation, including Google Translate, isn’t entirely mechanical in its translation process. The very first translation in any system will have come from an actual human being. After this though, the initial translation is used as a base from which the machine can learn how to translate any new text that is inputted. This is processed through a statistical approach, where information is taken from various sources and mashed back together. That is why that although in essence you’ll find a machine translated text is close to how it should read, quite often there can be certain words missing, or nuances and dialect left out completely that leave the text either completely nonsensical or with a meaning a bit different to how it was originally intended from the first language.
If you do have a translation project at hand, it is always more fruitful and cost effective to hire specialist translators. They will not only have the memory and expertise to translate manually, but also have the aid of CAT tools to help make the whole process a smooth and accurate one.