Reflection to the 17th Portsmouth Translation Conference

Beautiful pre-conference morning in Portsmouth old dockyard. (photo Jakub Absolon)

This is the reflection to my participation at the 17th Portsmouth Translation Conference “Translation and Disruption: Global and Local Perspective”.

During the conference, while watching the presentations, from which the most were really of high interest to me, many ideas, questions and answers stroke to me. Maybe my remarks help you to see some topics from other perspective or even gives you an idea to answer your own questions.

Reflection No. 1: Premium translation market, premium services, premium clients

Up to me, there are not only two or three levels of PEMT or translation. It is hard to define strictly what is FULL or FAST post-editing. I think the quality is the term which describes client’s expectations to fulfil their requirements. So, there is countless number of levels which must be explained by project briefs or client’s style-guides. If there are no customised requirements, there is only one standard quality which comes from standardised conventions. If there is no standard yet, recognized practices should be adopted.

Reflection No. 2: Mixing of terminology used for different roles of translator

It was mentioned several times during the discussion there are “dozens” of terms used for persons who are editing MT output. Some researches even try to put equal sign between PE and human translation revisors. At my lessons I am using this explanation which is, of course, only my point of view. There are two types of revisions.

Revision A, when revisor’s role is to revise text after translator whose quality is not considered to be excellent. LSPs use translator’s services for different reasons e.g. translator is domain specialist, there is no other native speaker for this language combination or simply the translator is in “acclimatisation” process when translator’s quality must be proven by real translation jobs. However, neither in this kind of revision a revisor should play the role of translator’s teacher.

Revision B, when revisor knows translator’s work and they are working as a team, not as two separate individuals. This of course requires PMs are actively creating such teams. In this case revisor should adopt the principle as follows:

TRUST BUT VERIFY

In PEMT process PE must be suspicious towards MT output even if its quality is close to human. We discussed it several times at the conference, there are new and dangerous types of critical errors hidden behind NMT “glasses” optics. Output is fluent, but the meaning is different. That is the reason why post-editing work is up to me more exhausting and requires continuous concentration for a long time. That is the reason I believe PE task is less creative but more exhaustive. But I would not call it “repeating automatic activity”. This would be only in case if meaning of translation is correct and PE only polishes the output. And this is not the case, at least I have no experience with such high-quality MT so far. In this case Post-Editor adopts the principle as follows:

BE SUSPICIOUS AND CORRECT

Reflection No. 3: PEMT definitions

There were several PEMT definitions used in our presentations, however, I would like to add the one we use. The main difference is that we focus at and insist on three main features of PEMT: NEED FOR:

SPEED

COMPETENT TRANSLATOR

MT OUTPUT QUALITY

Definition:
Post-editing is rapid translation using automated/machine translation output and subsequent revision and correction by a translator for the final product to reach the desired level of quality.
If you are interested in our brief Post-Editing Guidelines, please feel free to drop me an email to absolon@asap-translation.com and I will send it to you ASAP 😊.

Reflection No. 4: Different ways how translators use MT

Another good point was debate about different usage of MT by translator. I did a very short research on usage of MT by professional translators on Slovak and Czech market, previously Czecho-Slovak market 😊, with 150 respondents in Dec. 2016 / Jan. 2017. Principal findings are as follows:
Participants:

150 participants all together

134 (89.33%) answered: Yes, translation is my main source of earning

16 (10.67%) answered: No, I translate only occasionally, as an addition to my main job or I do it non-commercially

Results: Active professional translators

Yes, I use MT. I pre-translate each sentence by MT first (directly in CAT or outside) – 6%

Yes, I use MT. I often use MT for translation of parts of the sentences – 12%

Yes, I use MT. I use it for translation of some unknown words – 16%

Yes, I use MT very rarely – 31%

No, I never use MT – 35%

Results: “Amateur” translators

Yes, I use MT. I pre-translate each sentence by MT first (directly in CAT or outside) – 6%

Yes, I use MT. I often use MT for translation of parts of the sentences – 38%

Yes, I use MT. I use it for translation of some unknown words – 6%

Yes, I use MT very rarely – 25%

No, I never use MT – 25%