Lost in (Machine) Translation

Machine Translation (MT) is perfect for translating a text just enough to gain context or the gist. For many other projects, MT is just plain ineffective. The main issue with Machine Translation (MT) is that it is a non-value added process. You plug in a sentence and a translated response is generated. MT lacks the steps that comprise an effective and valuable translation process; without those steps your global company is sacrificing quality for cost. This will hurt overall brand image and run down global presence.

MT does not account for the following points that could make or break the connection between you and your target.


1. Cultural implications and interpretations: MT does NOT conduct a cultural analysis before translating; this hinders the accuracy of the translation. As a result, members of your audience could be offended by the terminology the machine generated. Audience interpretation should be top of “mind” when translating. Using human translators will help avoid things like colloquialism misconceptions.

2. Dialect: MT only takes into account the original translation from one language to the next. What MT does not do is understand the dialect that your audience utilizes. Understanding the specific dialect of a given language is important and can help avoid message/product confusion with your audience and in themarketplace.

3. Audience background: MTs do not know the specific audience your company is trying to acquire. They do not know the age, gender, education level, etc. This makes it impossible for machines to generate translations for specific targets. The adjustments are costly and time consuming procedures that require the use of a human translator.

Building a foundation on which you can communicate with your target is important to gaining customers and delivering the right message to your global audiences. MT capabilities are not sufficient for building this foundation.

Tell us in comments below how the use of MT has or has not worked for you.