It would be great if every project followed a linear workflow from beginning to end, so that after feeding the source file in at the beginning, the final file at the end pops out as a perfect specimen, with every step having somehow brought the quality closer and closer to perfection. But the real world isn’t so simple and our workflow is designed to take advantage of the linear aspects of the work, while also dealing with the surrounding complexity.
Failure points in the process cause many of the errors. I’ve found some in the software of some of today’s leading companies. Keeping these pitfalls in mind, we bring together the highest-level technical expertise, English fluency and industry experience with the best English>Korean review talent, to maintain a reasonable rate structure across the range of project types and sizes that we handle, while also delivering our best work.
Hybrid CAT-tool workflow
Because we often work with subject-matter experts who are not CAT-tool savvy, I handle most technical processing internally, with the actual translation and proofreading steps processed by my team on RTF files that I prepare with the relevant terminology, context and TM content for their reference. I have been working hard on these processes for several years now and they include proprietary procedures that I have developed and maintain internally to raise quality.
Language pair-specific quality assurance
One aspect that sometimes affects the quality of translation into Korean is the lack of native-speaker fluency in English among the pool of Korean translation resources. This is a lesser concern on technical documents, where standard language is often used, but it does crop up. Yet one more role I perform in the quality assurance process is to review source files at the beginning for potential linguistic issues that could trip up the team, and then spot-check the translations at the end for common problems in the English>Korean language pair. I’ve written about a lot of these in my Korean Translation Tips series, and am increasingly integrating these into my quality assurance work.
Client-side induced errors
Unnecessarily complicated and unclear instructions, work processes and file formats, as well as process disruptions (such as late client instructions, source files changes), lead to a disproportionate number of errors. This also applies to poor client-end translations and layout which we are asked to review and where we end up fixing things that shouldn’t have needed fixing, leaving less available attention/time for productive editing work. This is exacerbated when the budget is not adequate for the effort required. I try very hard to get all this worked out in advance, and to shield my team from distracting factors.
It is important to keep in mind that only so much can be expected at each step of the process, and with two languages as mutually different as English and Korean, this is even more true. While we’d like to believe that a translation performed by a competent translator will be perfect the first time every time, this is simply not the case.
In fact, each step in the process (file preparation, translation, proofreading, automated and manual QA, content-context integration, etc.) adds value… at least, it does if the scope for each step is designed properly and adequate value added by each participant.
Our process lets the translator and proofreader focus only on linguistic issues by separating out the technical file handling, project management and context integration into discrete steps that I manage at the beginning and end of the project. Furthermore, creating conditions for high linguistic quality from the translation and proofreading team means that QA can then focus on final polishing, while also resolving other technical complications.
In our process, the translator is the lowest function and this person not take any leadership role. My colleague who does nearly all of my proofreading (and selects the translator, thus having a motivation to keep the translator in line) would be closer to this role. But frankly, in the end, it's me doing the dictating of terms and ensuring adherence with instructions, style, etc, and ever so occasionally, going back strategically with mistakes I found to let my colleague know I'm still paying attention. We don’t generally send edits back down the ladder for review and approval, nor do we rub other team member’s noses in every single mistake that we can fix at a higher level.