Expert tips on style coupled with a comprehensive review of machine translation post-editing issues prepare attendees for success. BY ELENA OW-WING
On Saturday, July 30, the Technical Writing for “into English” Translators & The Art and Science of Post-Editing workshop enjoyed a turnout of 20 participants at the SFSU Downtown Campus. The first presentation, on methods and techniques of technical translation, was made by Karen Tkaczyk, a technical translator and editor with a pharmaceutical industry background. The second topic, dealing with attitudes toward machine translation (MT) and MT post-editing issues, was introduced by Mike Dillinger, Vice President of the Association for Machine Translation in the Americas and Principal at TOPs Globalization Consulting.
Skills can be developed
This was the main message in Karen’s presentation, which she began with the specifics of technical writing and its applications. Karen reminded us that technical writing aims at transmitting information accurately, with an objective tone and a focus on the technical content; therefore, it is important for technical translators/editors to have subject matter expertise, as well as strong research skills and an excellent command of the target language.
Much of the presentation was centered on methods for producing precise, clear, and brief technical texts. Karen emphasized that whereas precision is achieved through correct and consistent use of terminology, clarity and brevity have more to do with how we use the language. To be clear, the technical translator/editor should: avoid long sentences; reorder thoughts for coherence; and use strong action verbs and plain English to prevent misleading information. Techniques for achieving brevity include writing economically, omitting needless words, using the active voice, avoiding “to be” verbs that indicate passive voice, and reviewing texts for excessive use of “of”.
Karen focussed on factors that induce translators to ignore aspects of technical style. Certain texts (e.g., litigation, a patent case) have to be ambiguous, be it a source or a target text. Sometimes the translator/editor can’t access the author to clarify concepts, and that is reflected in the translation. Another problem is carryover: differences in source and target language structures may lead to ineffective translations. In addition, variety in style can be an issue (e.g. 200 ml/mL– small l in EU and UK, capital L for US).
Finally, Karen provided a list of resources and shared tips on developing your own style. She advised picking a style you like best and using it for jobs where no style is specified. It can also be handy to create your own style sheet for larger translation jobs and to refer to it for consistency.
Attitude is everything
Mike stressed this point more than once in his presentation on the controversial topic of post-editing. In his opinion, the problem is not Machine Translation, but attitudes toward this tool. Despite similarities to TM, and the professional recognition of the latter, MT is not widely used. Mike indicated that prevailing reactions to new tools, including MT, are to ignore them or to complain about them. Such attitudes prevent a translator from learning new tools and thus from increasing output, income, or free time. Those who accept MT, and are ready to work around its limitations, must learn how to post-edit MT output.
As Mike explained, post-editing consists of three main elements: accepting good MT segments; fixing those that can be fixed; and rejecting or redoing those that cannot. He advised against using any MT output that takes more than 2 seconds to understand. Mike also provided revealing post-editing statistics: the vast majority of post edits are only 1 or 2 words. Here, MT use makes perfect sense; if you are willing to develop strategies to fix commonly repeated error types.
According to Mike, it is important to know when to reject MT jobs. Ask yourself “Can I post-edit this job fast enough to get paid well?” Look for warning signs such as frequent terminology errors, strange formatting mistakes, unusual characters, etc. Mike made it clear that despite benefits of post-editing other people’s MT, it has its cons too. For those choosing this option Mike recommended to get some training.
This stimulating workshop created an entrepreneurial buzz among the attendees and provided us with many useful insights for developing successful, mutually beneficial relationships with end clients. EOW