Our Linguistic Services Manager has submitted the first myth below, based on her experience of working as a translator for many years and helping to optimise the quality process at SimulTrans.
Localization myth #1 ‘Stylistic corrections are purely preferential - they don't affect the translation quality.’
Life would be so much easier for translators and clients alike if this statement was always true! But like so many other issues affecting translation quality, it’s not that simple.
Client reviewers often request changes in phrasing and terminology to translated texts, particularly in the early stages of a translation project. Terminology changes are easy to implement and learn from: our translation teams simply add the preferred terms to a project glossary and then run quality checks before delivery to ensure that all approved glossary terms have been used.
Stylistic changes are another matter. First there is the question of whether they are really necessary, whether the original translation is acceptable as is (and if so, who will pay for the additional effort to implement changes?) and how will they be taken on board in future. No wonder the phrase ‘Stylistic changes, purely preferential’ is heard so often in our industry, to minimize the fuss.
There are cases when style is purely a matter of preference – when it comes to literature, Carver fans may not appreciate the style of Anthony Burgess or Flaubert. And on a more mundane level, it generally doesn’t make a difference if you write ‘in order to’ or ‘to’ in English texts. Just as the style of source texts can vary, so too can translations!
However, inappropriate or poor style can impair translation quality, preventing the translation from performing its function and sometimes – in the worst cases - distorting meaning. Many of the test translations submitted to SimulTrans fail to pass our quality criteria on this score.
Several factors need to be taken into account when evaluating the impact of stylistic ‘infelicities’.
- What is the text type?
- Who is the target readership?
- Is the primary function of the text to inform or persuade?
Generally, the more persuasive the text, the more important the role played by style. This is why awkward or non-idiomatic style can have a fatal impact on translation quality in a marketing brochure. In the following example (a press release), clunky sentences like the one below require stylistic editing for the text to work.
Example A: ‘In 2012, for the 2000 interviews that were conducted within the group, 93 % of these were concluded, a record rate that was attained thanks to the excellent internal communication within each of our divisions, particularly by our HR teams.
Most English speakers will find version B to be more readable:
Example B: ‘Of the 2,000 appraisals that were conducted within the group in 2012, 93% were successfully completed, a record figure that was made possible by the seamless internal communications within our divisions, and particularly by our HR teams.’
In the French example below, the translation is accurate, but the literal style jars.
EN : The carrier will attempt delivery again the next day.
D Le transporteur essaiera une nouvelle livraison le jour suivant.
C Le transporteur se présentera à nouveau à votre domicile le jour ouvré suivant.
3 ways that SimulTrans can help you avoid the prickly issue of stylistic preferences:
1. We include 1 free round of corrections as standard.
Our word rates include translation and revision, plus one free round of corrections. Typically translations are sent to client reviewers once translation and revision is complete: client reviewers then highlight any changes required. Our translators implement these changes and update the Translation Memory (TM) for future projects. Receiving and implementing client feedback is one of the best ways we can customise our processes to suit client needs.
2. We help you choose your preferred translators – if that’s what you want!
Our dedicated Resource Management team recruits and assigns the best qualified translators from our established network to your projects. However, for some projects, where a particular tone and style are required (marketing materials in particular), we encourage clients to participate in selecting translators, if they wish. Here’s one process that has proven highly successful for one of our key clients in the area of statistical software:
- Client selects a representative sample text for translation – approx. 300 words.
- SimulTrans selects 3-4 qualified translators to provide translations of the text, based on their expertise, track record and project availability.
- SimulTrans eliminates the usual revision process and additional QA checks for this sample, since the purpose is to identify the original translator’s command of the register and style.
- SimulTrans labels the completed samples and ask the client to pick the preferred translation(s). We usually recommend choosing at least 2 translation samples (primary translator and backup), provided the quality meets expectations.
- SimulTrans assigns the client’s preferred translators to all future projects.
3. Integrate client reviewers directly in the translation process.
It’s vital for any client changes to be saved in the translation memory so they can be re-used in later translations. For smaller projects, we usually request that translators implement changes manually. For larger, ongoing projects, we can set up a process to allow client reviewers to make edits to translations directly within the translation environment – without having to own or run specialised translation software.
One of the tools we commonly use for this purpose is WorldServer, which is a browser-based system. For client reviewers who prefer to work offline and still have their changes implemented directly into the project TM, other solutions are possible:
The EMEA marketing director of one of our key software clients notes: ‘As I’m often on the move, I prefer to conduct reviews offline – I cannot always rely on Internet access on the train or plane. The team at SimulTrans adapted the review process for our French translations so that we could review bilingual Word files: I’m happy with the process so far as the translators can clearly see and learn from my changes, and these are saved for the future!’.
Style matters: Despite many a reference within our industry to ‘preferential changes’, it’s simply not true that stylistic corrections don’t affect translation quality. Since our definition of translation quality at SimulTrans is ‘meeting or exceeding client requirements’, we also need to pay attention to our client preferences, whether these concern terminology, formatting or style. Let us know if you agree, disagree or have any questions – we’d love to hear from you at firstname.lastname@example.org.