Benefits of a back translation

Benefits of a back translation

The term “back translation” can seem nonsensical if it’s outside the context of a “forward translation.” Technically, every translation is a “forward translation,” but if a forward translation workflow also contains a process to translate the target document back into the source language, then that part of the process is called a “back translation.” In other words, a back translation only makes sense if it’s part of a forward translation process and, ideally, both parts happen back-to-back in the same project. The back translation process is perhaps the most valuable of all translation processes because it allows for unique benefits:

  • It doubles the amount of linguistic quality control (LQA)
  • It allows a monolingual source text author to personally validate translation quality
  • It includes additional robust documentation that can be used for compliance or legal purposes
  • It even works to evaluate how word play, puns, or heavily connotative text has been rendered in the forward translation.

For all these reasons, a back translation is one of the best ways to ensure the highest quality translations.

Double the LQA

This is what a typical ISO17100 translation process looks like:

Benefits of a back translation

This is what BURG’s back translation process looks like:

Benefits of a back translation

The red denotes LQA activities. What is important to note is that the back translator is “blinded” from the source text. This means they cannot see what the original source text said and “cheat” to use it to guide their own translation. Just like in English, where there is more than one way to say the same thing, there is more than one way to say the same thing in other languages. As a result, the output of the forward translation and back translation, even when perfectly accurate, will not be identical word-for-word. To validate the accuracy and to reconcile the two outputs, the second editor does a “reconciliation” of the two outputs. This second editor doubles the LQA because they not only edit the back translation, but they also compare it to the forward translation. The origin of any defect found in either translation (forward or back) is located and fixed. This is called reconciliation. However, the benefits of the second editor are only realized if the forward translation and back translation happen simultaneously in the same project. If the forward translation and back translation are split into two different projects, then the second editor can’t leverage the original source text used in the forward translation, which results in the two projects equating to two forward translations, but in opposite language-pair directions. Therefore, if you request a back translations, it’s critical that it is requested at the same time as the forward translation.  

Personally validate translation quality

The most obvious benefit of a back translation, and the reason most clients request it, is so that they can see for themselves that the translation meets their level of satisfaction. It bears repeating that, “Just like there is more than one way to say the same thing in English, there is more than one way to say the same thing in other languages.” This means that clients should not expect the back translation to read word-for-word what was written in the source text. This is particularly true when a degree of transcreation was required to translate the source text.  Nonetheless, personally appreciating the quality of a translation can be extremely satisfying to clients.  

Robust documentation

The back translation process produces unique documentation besides just the final translation outputs. The second editor, who does the reconciliation, performs the reconciliation by viewing a trilingual table. This table has the source text, target text, and blinded back translation text as columns. Each row in the trilingual table contains a segment of text from all three. The editor then reviews the table, row-by-row, to confirm that both the forward translation and back translation are accurate. This document is the proof that clients need in order to validate the accuracy of the translation.    

Works with transcreated text

Sometimes, clients ask how back translations could possibly work with transcreation. The common thought is that since languages are different, and sometimes there is no identical way to say something in another language (true!), then how can the back translation possibly be accurate or valid? The answer is deceptively simple: use comments to explain what is not apparent in the translation. For example, for transcreation projects, the trilingual table is the default output of the process, however, it comes with an additional column – the comments column. In a transcreation project, the comments column is crucial for explaining the rationale of the translation precisely – because the words depart so much from the source.  In a standard back translation project, there is no comments column because, usually, there is no need. However, if an editor feels the need to explain a translation, they can add a comment to that row.  

Summary

The back translation process is a unique workflow designed to allow non-bilingual people to read and understand translated documents of any language. When done properly, it doubles the level of linguistic quality control undertaken during the translation process. The outcome is a nearly perfect translation that can be read by any person who can only read the source language text. Furthermore, its robustness extends to highly connotative transcreation text.  

If you’d like to learn more about how BURG Translations can help you with your back translations, contact us today.