With clinical trials now a global industry, "back" translation requires a closer look.
The expanding global market for pharmaceuticals is requiring drug companies to translate their product information, labels, informed consent forms, and other printed material into various languages to be understood by the end user. Translation is looked upon by those who perform it as an art rather than a science, a subjective process in which there is no right answer. While there are exams and qualifications that can be taken and acquired, there is no plug and play at the end of the translation process in which the translation can be validated against its original source.
As a result, more and more pharmaceutical companies are turning to the use of back translation as a secondary step in the translation process. Back translation is the translation of an already translated text "back" into its original language. It is performed, as with the forward translation, by a professional translator once the material has already been translated. The back translator does not have access to the source text.
Clinical trial managers and other project and product managers use the completed back translation to validate the translation itself by comparing the back translation and the source and looking for discrepancies in the text. While this would seem on a basic level to be a good way to judge the quality of the original translation, this is not always the case. Since translation is so subjective, those reviewing translations must be aware of the potential pitfalls in this process.
Misunderstandings in Word or Meaning Translations
The term "back" is somewhat misleading, as it implies that the text will be translated to exactly match how it was written in the original. This is almost never the case. The primary objective of back translation is to identify problem areas in the original translation by providing a comparison to the original source, in order to identify discrepancies in meaning. While this is certainly a valid quality assurance step, it also invites those comparing the texts to look at more than just the overall meaning of the back translation. Understanding the importance of meaning in the back translation process will help those reviewing back translations.
Back translation hinges on the idea that the back translation and the source can be lined up and compared against one another. However, any translation that is done well will depart from the semantics and structure of the original source language in order to meet the requirements of good style in the translation. Grammar is inherently diverse across different languages, and the translator will retain the meaning of the translation, rather than the original grammatical properties of the source text. A back translation comparison should essentially ignore syntax and focus on the comparative meaning of the two texts. As such, a back translator is usually able to translate more than one dialect into English (Spanish for Mexico and Spanish for Puerto Rico, for example).
The concept of meaning has been long researched and documented by translation theorists and academics. For example, the noun "work" can be synonymous and equivalent in meaning both with "literary masterpiece" (e.g., a work of Shakespeare) and "place of employment" (e.g., workplace). At a word-to-word level, equivalence in meaning can be useless; already we are beginning to see the importance of context in any given text.
Each translation segment, whether it be a sentence, heading or short paragraph, should be considered as an entity. To break these segments down to a word level not only takes the words out of context, but results in the reviewer comparing word against word in a sentence that may be entirely equivalent in both languages (see Table).
With no context, the back translator has no way of knowing which sense the original English had. It's important to remember that almost all translators are given reference material when they are working on a text. Whether this is a translation memory, past translated material, a glossary or pictures, the translator knows definitively how their translation will be used. On the contrary, it is impossible to provide the back translator with as much reference material. Providing a brochure with pictures will also let the back translator see the original text, which is against the rules of back translation, lest it influence the translator's work back into the original language.
Even at the highest textual level, equivalence of meaning relies on not only the context within the translation itself, but among other, similar types of materials, all of which may not be present at the time of back translation review. Nevertheless, if the client is the end reviewer, they will almost certainly benefit from knowing the subject matter and having access to previous translations/back translations so that cohesion across texts is maintained.
In the pharmaceutical industry, differences in the back translation and original English such as "dose" and "dosage" are common, as are the terms "adverse effect," "adverse event," and "adverse reaction." In these instances, it is the reviewer's responsibility to research the terms and find out whether or not the meaning is different enough to warrant a change to the original translation.
When the source language is English (which is very common in the translation industry), there is often discrepancy in the meanings between the original source and the back translation. Spanish, for example, has a notably smaller vocabulary, and many words that are used in English simply do not exist, which can create issues (see Table).
Two separate nouns for "aches" and "pains" do not exist in Spanish, resigning the first translator to use only one noun: "dolor" (literally: pain). As such, the back translator, with no access to the source, is unaware that two nouns exist in this text, and translates simply pain. A comparison of the back translation and the source will reveal a missing noun, which will only become evident at the review stage. Writers of the English source text should be aware of these issues, wherever possible.
The Spanish "afectar" means not only "to affect" but also "to harm, to influence unfavorably," so the Spanish translation above is perfectly correct. The back translator, though, has translated it as "affect," also a correct translation. In this case, though, the difference between the two verbs is important, since we are referring to drug effects. This kind of example will seem on first review of the back translation to be a mistranslation, and must be clarified with the translator during the review.
Should an actual mistake be made, we see where back translation is most valuable. In another example, if the original translator translated "decrease" as "aumenta," the back translation would read "increase." In these situations, back translation is of real value, since the words are antonyms, and there is a definite discrepancy in meaning.
When a back translation is requested, it is built into a project schedule as any other step would be. Professional translators with subject-matter knowledge are assigned to work on the back translation, in the same way as professional translators are the ones translating the original text. Because the back translation cannot begin until the original translation is complete, it adds a considerable amount of time to the schedule, sometimes almost doubling it. The cost of another round of translation is significant, too, and this should be accounted for in the budget.
Alternatives to back translation are usually employed when cost and schedule do not leave room for back translation. Since back translation is used as a linguistic quality assurance step, alternatives to this procedure must involve some level of linguistic knowledge (a formatting or other mechanical QA, for example, is not an alternative). In the translation industry, the main alternative to back translation is an in-country review of the original (or "forward") translation by an affiliate on-site. The purpose of back translation is to determine whether there are any mistakes in the original translation, and an in-country reviewer is undoubtedly qualified to come to the same conclusion.
Back translation should be employed with caution. When comparing back translations against the source material, reviewers should be aware of the importance of meaning versus semantics. Syntactical or stylistic differences such as "November 12th, 2006" and "11/12/2006" should be viewed as having the same meaning. Back translation, a costly and time-consuming procedure, can only be successful if this concept is understood. Reviewers will find that this will help eliminate the back-and-forth to project managers and linguists, and aid in keeping costs and review cycle-time to a minimum.
1. S. Hervey and I. Higgins, Thinking Translation: A Course in Translation Method, French-English (Routledge, New York, 1992).
2. M. Baker, In Other Words: A Coursebook on Translation (Routledge, New York, 1992).
3. P. Newmark, A Textbook of Translation (Prentice Hall, New York, 1988).
4. U. Eco, Mouse Or Rat?—Translation As Negotiation (Orion Publishing Group, London, 2004).
5. Back Translation for Quality Control of Informed Consent Forms, http://www.translationdirectory.com/article1043.htm.
Hilary Davies was most recently Linguistic Services Lead for ForeignExchange Translations, and she has since left the company, which is located at 1001 Watertown Street, Newton, MA 02465. For correspondence, please contact Jason Heaton at email@example.com.