Beyond Translation Memory: Possible Benefits of Different Functions of Translation Environment Tools

Translation environment tools (TEnTs) are relatively common and easy to access nowadays. Their use is virtually inevitable in many fields, as translators have to create quality translations of large documents in relatively short turnaround times.

By Henrietta Ábrányi

Many researchers in the field of translation studies deal with translation memories or terminology databases from different points of views, and only mention other functions of TEnTs quite briefly. As a result, there are still many questions surrounding them.

As a translation tools teacher at a Hungarian university, I often find that students have high expectations at first, which then transform into a kind of disappointment when they realize that the tools are not something they can learn easily without practising (a lot). After a few classes, many of them still tend to think that these tools just hinder the translation process, especially if technical problems occur. Another concern among students—and even fellow teachers—is that, in their opinion, these tools are not worth the time and money if one translates in a variety of fields or deals only with general texts.

Yet, in my opinion, the use of TEnTs is not an option anymore. It is inevitable if a translator wants to keep up with the high expectations of the translation market. I would like to highlight some benefits of the lesser known functions of the tools available from a freelance translator’s point of view, when working with texts of different lengths in a variety of fields.

What is a translation environment tool?

Computer tools that aid the work of translators were defined and classified differently in the past, which has led to rather misleading and ambiguous terminology in the field. The term ‘computer-assisted/-aided translation tools’ (CAT tools), for example, has been used in two different ways. First, it means any computer tool designed to help the translator’s work in any way, including online or offline dictionaries, glossaries, grammatical aids, parallel texts, utilities for project management, etc. Others use the term as a synonym for ‘translation memory tools’ which covers only those tools that work with a translation memory at a minimum (storing translations for later reuse). Nowadays, researchers suggest using the term ‘translation environment tools’ instead of CAT tools, as it covers more than just programs working with a translation memory, and instead of ‘translation memory tools,’ because that expression focuses only on one function which, while of central importance, is not the only useful component.1

Recent TEnTs contain, for example, a terminology component, an alignment module, concordance search, a QA component, spell-checker and different analysis features, and they can deal with a variety of different file formats. A relatively new component is what is called ‘sub-segment matching or leveraging’ (see below).

Why use TEnTs?

The use of TEnTs is in fact inevitable if a translator works with long and highly repetitive texts using a simple and consistent style, sentence structure, and consistent terminology (e.g. updated or revised manuals, where only the new or modified parts of the document are to be translated).2 TEnTs speed up the translation process because the translator is able to filter empty or changed segments. That way s/he does not have to translate the whole text from scratch, and the translation will be consistent in terms of style, sentence structure and terminology.

But, could TEnTs be useful for translators who do not work with the above types of texts? If a translator deals only with general texts, for example, it does not matter how much s/he works, it is not likely that any matches will come from the translation memory. However, there are other functions of TEnTs that might be useful, even for them.

Terminology

For most translation tasks, terminology work is necessary, even when dealing with general texts. Yet, not only can terms, in a strict sense, i.e. defining a particular concept and thus having a specific meaning in a specific field of expertise, be added to a terminology database, but also any words/expressions that a translator often encounters. However, a ‘bulk’ terminology, consisting of entries on different topics, does not have to be created because separate databases can be created (e.g. ‘general’, ‘geography’, ‘history’), which can all be assigned to the same project if the translator so wishes. So, if historical expressions are found in a geographical text, they do not have to be checked again if they are already in another database. Moreover, definitions, sources or examples can also be added to the entries, which help translators decide which expression to use in a particular context (e.g., in Hungarian we have different equivalents for captain referring, for example, to the head of a company in an army, or to the commander of a ship). All of this speeds up the translation process and helps the translator create a consistent translation and choose the appropriate expressions.

Alignment

The second useful component is alignment.This feature allows translators to store their previous translations—available electronically but created without translation tools —in a translation memory for reuse within the translation environment. This means not having to search through documents to find a previously translated part of the text, which may be very time-consuming if the document is large or if there are several documents to be checked. In one of the most recent translation tools, translators now have the opportunity to create ‘corpora’ where—besides aligned documents—monolingual texts can also be stored and used for reference. Thus, everything can be stored in one place and there is no need to open the stored texts as they can be searched with the concordance search feature, which looks for (parts of) words, expressions or even clauses and full sentences in translation memories and corpora and finds not only the word in question, but also the context in which it appears. Such tools may speed up the translation process, help the translator create consistent texts (e.g. in terms of style) and, if authentic target language texts are added to the corpus, it may help translators distance themselves better from the original text and its sentence structure or wording, and thus avoid creating a text that reads like a translation.

Although translators around the world usually translate only into their native language, in Hungary this is usually not the case, as many translators have to translate into their first (or even second) foreign language. The above function is therefore even more useful because, as a non-native speaker, it is very difficult to create a text that sounds authentic to the target reader.

Sub-segment matching

Finally, another useful function for any translator is what is called ‘sub-segment matching.’ In short, the translation environment tool monitors the selected resources (e.g. translation memory, term base, corpora or dictionaries) and gives the translator suggestions, while s/he is typing. Suggestions appear after typing the first few characters in a list or in the segment and they can be inserted in a second. Suggestions may include words, expressions or clauses. Thus, even smaller units that otherwise would not appear among translation memory matches when the percentage is too low can be reused. This is another function that speeds up the translation process as translators do not have to type each character and can avoid spending long hours searching. It also contributes to consistency as suggestions are provided based on previous work (e.g. when translating recipes with instructions formulated differently, such as imperatives or ‘we/you can decorate’).

In conclusion, translation environment tools can help any translator speed up the translation process by avoiding long searches and providing ‘containers’ to store everything used in the process. While doing this, the tools may even help create better quality translations as they contribute to a more consistent and correct style, terminology and phraseology. Moreover, translation tools can even help translators distance themselves from the original text and avoid creating texts that read like translations. All of these features are even more useful when several translators are working on the same large translation task. With the help of TEnTs, differences, such as in style, may be reduced to a minimum and the text can form a consistent whole.

1 Zetzsche, Jost. The Translator’s Tool Box: A Computer Primer for Translators. International Writers’ Group, LLC, 2008.

2 Feder, Marcin. A Tentative Proposal for Machine Assisted Human Translation (MAHT)—Tool-Specific General Text Typology. Linguistica Antverpiensia, New SeriesThemes in Translation Studies 36(1) 365–374, 2002.

Henrietta Ábrányi is a freelance translator and a translation tools teacher at the Translator and Interpreter Training Centre at the Eötvös Loránd University in Budapest, Hungary. She is currently writing her PhD thesis on the impact of translation environment tools on the translated text.