It’s easy to think that translation work only concerns the translation of words and sentences and editing text. In a way, that’s exactly what it is, but few outside the industry actually know how much modern translation relies on technology. As this is something that greatly benefits customers, let’s take a closer look at some basic translation tools.
Software that facilitates translation was developed in the 1980s when the need for translated content started to rapidly increase. There were not enough translators to meet the demand, so computers, which were starting to become more common, provided some long-needed efficiency to translation.
Computer-assisted translation (CAT) refers to software or translation tools, such as memoQ or SDL Trados Studio, that professional translators generally use in their work. Translation tools have various features that speed up and facilitate translation, such as a translation memory (TM), term bases and spellcheckers. Keep in mind that computer-assisted translation is not the same as machine translation (MT) or Google Translate; rather, it refers to tools that help translators work more efficiently and better.
How does translation memory software work?
The most important tool available to a modern translator is translation memory software. A translation memory is a database where human translations are stored. A translation memory is not capable of independent learning or thinking. Instead, it is solely based on saved strings of characters called translation memory segments. Typically, segments are words, groups of words and sentences. As the translator works through a text, their translations are saved for later use, segment by segment.
The concrete benefit of translation memory software is that in subsequent translations the software suggests previously translated segments whenever it detects a degree of similarity with something that has already been translated. Complete segments that already exist in the memory are called full matches and segments that are similar to a lesser degree are called fuzzy matches. This helps the translator keep their translations consistent and work faster.
How does a translation memory benefit the customer?
Translation memory software is at its most useful when translating texts that are updated regularly (such as manuals) and when translating larger projects where many translators and proofreaders are working simultaneously on the material. In this way, it is easy to re-use segments that have been previously translated, that repeat within the material or that a colleague has just translated, with no need to repeat the work again.
As the translation memory grows, the translation speed for texts that repeat might double or even triple, while the consistency and quality of translations improve. In addition, the customer benefits from cost savings as the same text or part of the text does not need to be translated again and again. This is why a customer who orders translations often should order all their translations from the same service provider.
Various kinds of term bases go hand in hand with translation memories. Translation memory software usually includes term bases in which translators can save any terms found in their translation. Once the terms have been validated and clarified as the exact terms the customer wants to use, the translation memory software suggests the correct term in the term base whenever it detects it in the source text.
The benefits of the term base are obvious for both the translator and the customer. The translator saves a great deal of time because finding the correct terms is faster. The customer benefits from consistent terminology in all its translations and also saves on expenses when the quality of translations improves and the need for corrections decreases.
Data security of translation tools
GDPR made some customers ask whether it is safe to store translations in a translation memory for a translation agency’s later use. A professional translation agency knows how to store the translation memories created from its customers’ texts, separate from each other and data secure.
Instead of storing all translations in one big translation memory and utilising one customer’s translations in another customer’s assignments, a responsible translation partner creates a separate memory file for each language pair for each customer, which is then only used when the customer needs translations in that language pair. When translators translate directly into these customer-specific databases on data secure servers, the customer can rest assured that their translations will not be stored other customers’ translations.
A good and appropriately maintained translation management system identifies each assignment with the correct customer account and with the correct customer’s translation memory. The memories do not spill over into another customer’s text pool as each memory file is a separate closed file, in which another customer’s source texts or translations cannot appear even by accident. The same goes for term bases: customers cannot see the terms or access them as the term records are customer-specific and protected with access rights.
Translation is specialist work that uses a number of different tools. Both the translator and customer benefit from these. However, the most important and best tool is the translator’s brain — at least for now.