What is postediting?
Before answering this question, let me tell you that postediting is the translation of the 21st century, or, in the marketing terms of some translation companies, it’s translation 2.0.
Postediting is a way of spontaneous translation which new generations of translators are taught at universities and, at the same time, the way of translating that the translators of the last century detest.
On the other hand, postediting is a strategy which many companies are currently using to lower translation costs and reduce delivery times.
Índice de contenidos
Index of contents
Index du contenu
I’ll get right down to business and give you a simple definition of what postediting actually is:postediting is the revision of a machine translation by a human translator.
While you are reading these lines, the term “post-editing” may be new to you, but “postediting" has already been used at the end of the 20th century. According to Collins Dictionary postediting is: the act of editing after a piece of writing has been produced or printed by a machine
Postediting is becoming a more frequently offered service among translation companies. This trend can be explained by the quality improvement of machine translations.
If you have used Google Translator since its launch in 2006, you will have witnessed the progression of machine translation yourself. This is particularly true since the introduction in 2017 of a new paradigm: neural machine translation.
Postediting is linked to the quality of machine translation. For years, the poor quality of machine translation made postediting unfeasible, as it was basically a matter of re-translating the text and not just revising it.
As I said, the viability of the post-editing depends on the quality of the machine translation. Therefore, in cases where machine translation does not give us good results, there will be no point in postediting.
In other words, there will be no productivity gains between revising the machine translation or doing a translation from scratch.
As a general rule, we can say that the more creative a text is, the worse will machine translation work. Machine translation is terrible for translating a Bob Dylan or a Joaquín Sabina song.
Nor does it work for translating literary pieces, nor does it work for translating marketing texts or advertising slogans created by inspired creative minds.
In the case of rare, highly specialised texts it usually does not work either, unless we train the machine translation engine. For example, if we have a lot of good translations on that topic, we can feed the machine translation system with that.
To find out more about what we can do to improve the results of machine translation if your company operates in highly specialised fields, you may be interested in reading my blog “Customized machine translation: The right option for my company?”.
If you read the marketing of companies involved in developing machine translation systems, you would think that the translation profession has extinguished. This is still far from the truth.
Nor is it true what many translation companies and professional translators say about how bad and not accurate machine translation is.
The truth is that when machine translation produces good results, there is no point in translating from scratch. Better to have a programme translating and having a person reviewing to refine the style and detect errors.
In other words, in many cases it does make sense to use postediting, as we will be able to reduce costs and reduce delivery times.
However, many translators reluctantly agree to do post-editing jobs, or reject right away. This rejection is partly due to the same reasons I gave in my blog “Why translators don't like revisions”
Basically, there are three reasons: the translator does not know a priori the quality of the machine translation, it is a different job type than translation and, last but not least, the rates per word are lower.
Moreover, something that for years created animosity towards postediting on the part of professional translators was the fact that postediting discounts given to freelance translators by translation companies were not equivalent to productivity gains.
In this way, the translator could charge up to 40% less per word for practically translating a text again.
Postediting, i. e. the revision by a human translator of a translated text using a machine translation system, is now a more frequent and feasible service.
It is possible for your company to benefit from the advantages of postediting. Basically, it reduces costs and result in shorter delivery times.
In the future we can expect postediting to become even more prevalent than it is today, due to professional translators accepting this job type more widely and as machine translation continues to improve.
If you are interested in this topic, I recommend the following blog: “Past, present and future of translation”.
Other articles you may be interested in:
Josh Gambin holds a 5-year degree in Biology from the University of Valencia (Spain) and a 4-year degree in Translation and Interpreting from the University of Granada (Spain). He has worked as a freelance translator, in-house translator, desktop publisher and project manager. From 2002, he is a founding member of AbroadLlink and currently works as Marketing and Sales Manager.