Jeremy Coombs, senior vice president of operations at MultiLing, shares new ideas on how to measure translation quality in a contributed column to IP Frontline. This leading IP portal helps the IP community share information and also provides original reporting on IP news each day.
Read the complete article from Coombs here.
In his article, Coombs presents a quality definition developed by Dr. Alan K Melby of Brigham Young University’s Translation Research Group, through his work with the American Society for Testing and Materials (ASTM) and International Organization for Standardization (ISO) specifications. His definition addresses all the facets of the process that influence our perception of quality, such as client requirements, necessary accuracy and fluency, and understanding of the end-user’s needs.
Melby’s definition of quality reads:
“A quality translation (1) demonstrates required accuracy and fluency (2) for the audience and purpose and (3) complies with all other negotiated specifications, taking into account end-user needs.”
Coombs shares a discussion where translators and editors were asked to define quality. “Unsurprisingly, their answers followed with the expected ‘correct grammar,’ ‘correct terminology,’ ‘good formatting,’ etc. as would be dictated for the standards of their language. Thus a prescriptive approach to defining translation quality is fairly direct.
“Now, for the sake of argument, what if the requestor of the translation specifically requires a level of translation that will certainly contain errors or inconsistencies? For example, Daniel is an inventor of a product and would like to know if a similar product in another market exists already. If he finds a prior patent or some other article in another language, he might need a translation that allows him to know the content. This is commonly referred to as a ‘gist’ translation.
“For the sake of cost and time, the translator would most likely be instructed to translate at a rapid rate and would probably skip edit and revision. Daniel would receive a rough, first pass translation of the document. Would this document adhere to all grammar rules, be typo free and perfectly consistent with established terminology? Most likely not. However, would this translation fill his requirements of cost, speed and legibility? Yes. So where on the scale of Quality does this translation fall? Is it poor quality because it fails to meet certain prescriptive rules which constitute good language? Is it medium quality, not perfect but fulfilling the requirements?”
Coombs argues that by adopting Melby’s definition of quality, we’re able to capture the intent and requirements of the requestor considering the translation’s end-user. Thus, any translation that meets the scope of the work and best serves both the requestor and the end-user can, and should, be considered a quality translation.
In fact, looking back at Daniel’s example above and applying this definition, the translation neither is low quality (due to potential errors) nor is it even medium quality. As it meets all of the needs of the requestor, it should be considered high quality.
The beauty of Dr. Melby’s definition is the fact that it embraces the subjectivity of language, combines it with the prescriptive ideas of accuracy and fluency, and wraps it all up in a framework based on the scope as agreed upon between the requestor and the translator.