To what extent do translators need to take account of PEMT?
Thread poster: John Fossey

John Fossey  Identity Verified
Canada
Local time: 14:20
Member (2008)
French to English
Jul 25, 2016

According to a recent study by Common Sense Advisory over the next 3 years the volume of human translation is expected to grow from 265 million words per year (mwpy) to 307 mwpy, an increase of 16%. PEMT volume is expected grow from 307 mwpy to 603 mwpy, an increase of 96%, in the same period. The figures represent the average projections of a small number of the largest translation agencies.

So if these figures are correct, how much do translators need to worry about PEMT?

(from "MT Gains Traction at Both Enterprises and LSPs" by Common Sense Advisory Common Sense Advisory)

[Edited at 2016-07-25 17:54 GMT]


 

Luciana Jesus  Identity Verified
Brazil
Member (2014)
English to Portuguese
+ ...
Translation = editing? Jul 25, 2016

The only thing that worries me is that the translation we do will be called "editing" and rates might be lower. I don't believe we will ever be out of work though.

 

philgoddard
United States
Member (2009)
German to English
+ ...
I think those are made-up figures. Jul 25, 2016

A full-time translator should be able to do roughly 400-500,000 words a year. If their figures are correct, and ignoring machine translation, there's only enough work for maybe 500 translators in the world.

 

Sheila Wilson  Identity Verified
Spain
Local time: 21:20
Member (2007)
English
+ ...
I think it depends on their market Jul 25, 2016

John Fossey wrote:
So if these figures are correct, how much do translators need to worry about PEMT?

I'm in the fortunate position of specialising in an area which doesn't lend itself to PEMT: marketing. I'm also 60 so less concerned with the long-term future but I wouldn't regard that as fortunate at all icon_frown.gif. Other areas where style counts for a lot (journalism, literature, etc.) should see little change.

As far as technical and "general" translation is concerned, I'm sure there will be two distinct markets:
Scenario 1) A multinational agency with an enormous data bank will get a massive multilingual order from an end-client corporation. It will run the texts through its own MT software and deliver it to a translation broker, who'll pass it on to more lowly brokers. It will eventually arrive on a freelancer translator's desk as a task to be done yesterday for 3 peanuts per hour. Once the translated text has arrived back at the top-level agency and been declared pretty much unfit for purpose, two of those peanuts will be taken as a "fine" by the agency paying the translator, as they will be fined by the level above, etc. After 30 days the agency at the top will pay the remaining percentage to its broker, and so a little money may start trickling down the pyramid until one day a peanut might reach the translator. By that time he or she will probably have died of starvation, but that's of no worry to the top-level agency that's doing just fine, thank you.

Scenario 2) A suitably qualified translator is paid a suitable rate to translate a new technical document, with the aid of a tailored TM, no doubt. That person or similar translators will be paid similar rates - per hour, that is - to update the document. There will also be a need to translate similar material, and this could well be in the form of PEMT. But this market will pay the same hourly rate for PEMT as it was paying for translation as it recognises the need to attract the same quality professionals to work on the texts.

As both those markets exist today, I don't expect any real change, just a shift as most professional translators try to move themselves up to the N°2 scenario or just drop out of the industry, leaving the N°1 work to the "hobby" translators (students, full-time parents/carers, moonlighting employees, the unemployed, ... who happen to get by in two languages) and those unfortunates who can't find anything better.


 

John Fossey  Identity Verified
Canada
Local time: 14:20
Member (2008)
French to English
TOPIC STARTER
Clarification Jul 25, 2016

philgoddard wrote:

A full-time translator should be able to do roughly 400-500,000 words a year. If their figures are correct, and ignoring machine translation, there's only enough work for maybe 500 translators in the world.


I've clarified the source of the numbers, they are average projections reported by a small number of the largest translation agencies.


 

Peter Linton  Identity Verified
Local time: 21:20
Swedish to English
+ ...
Never out of work - but what work? Jul 25, 2016

Luciana Jesus wrote:
I don't believe we will ever be out of work though.


I agree. I have done some PEMT work, and been reasonably impressed by the MT output. But there are some aspects of language that are no problem for humans but very difficult for computers who lack common sense. For example, there is a Swedish word 'tal' which means number. But it also means speech. The MT text I edited sometimes translated it as number and sometimes as speech, sometimes both in the same sentence.

Another problem for computers is where something has two names in different sentences, such as 'The Prime Minister' and 'Mrs May'.

The context makes the meaning plain to humans, but how do you programme the computer to recognise the context or the name?

So I agree with you about never being out of work, but I think that work will increasingly be PEMT.


 

Frank Zou  Identity Verified
China
Local time: 04:20
Member (2016)
Chinese to English
+ ...
The projections must be for these companies only Jul 26, 2016

John Fossey wrote:

philgoddard wrote:

A full-time translator should be able to do roughly 400-500,000 words a year. If their figures are correct, and ignoring machine translation, there's only enough work for maybe 500 translators in the world.


I've clarified the source of the numbers, they are average projections reported by a small number of the largest translation agencies.


I think the numbers are their own business projections instead of the whole industry. We should expect word count by billions per year, otherwise, a lot of translators would lose their job.


 

Tomás Cano Binder, BA, CT  Identity Verified
Spain
Local time: 22:20
Member (2005)
English to Spanish
+ ...
PEMT is not my market Jul 26, 2016

I think that PEMT is not my market, probably because of my age. I am not too old to learn the technology and PE techniques, but I am too old to accept that PEMT is proper translation. This makes me a bad candidate for PEMT work.

Younger translators, however, might be more prone to accept PEMT as an honourable activity, and could make very good money with proper training and tools. It is important, however, that entrepreneurs also accept the idea that PEMT is a proper profession, and pay "PEMTranslators" a decent rate.


 

Gabriele Demuth  Identity Verified
United Kingdom
Local time: 21:20
Member (2014)
English to German
I am out Jul 26, 2016

I enjoy translating as I consider it to be a creative process and I cannot see how I could possibly enjoy PEMT, therefore I am not really interested.

 

Christine Andersen  Identity Verified
Denmark
Local time: 22:20
Member (2003)
Danish to English
+ ...
Differentiating the market Jul 26, 2016

Translators need to accept that PEMT is not going to go away, and do their best to push the market in the direction of something like Sheila's Scenario 2.

Scenario 1 still leaves the original client without a usable translation, so they too should be contacted and drawn over to Scenario 2.

Just as there are different roles in the profession - translators, interpreters, project managers ... it should be assumed that some will be skilled and contented MT post-editors. Not me, I get a kind of block after half a paragraph of mediocre text, whether it is produced by a human or a machine. I am hopeless at interpreting too.

In short, language professionals will need to find their roles and specialise, and educate clients to understand that a 'translator' is not necessarily also a teacher, interpreter and MT editor. How much individuals will combine or exclude the different functions is up to them.

In short, translators will have to decide whether they want to specialise in PEMT and work with it, or whether they prefer another area, but it should not be the job everyone hates and tries to push over to everyone else.

PEMT should be taught as an option alongside other forms of training, to attract the kinds of people who will be suited to the job. I will really respect anyone who has the patience and can do it well!


 

Tom in London
United Kingdom
Local time: 21:20
Member (2008)
Italian to English
Illness? Jul 26, 2016

I thought PEMT must be some sort of illness. I've read the posts and I still don't know what it is.

 

Ben Senior  Identity Verified
Germany
Local time: 22:20
German to English
PEMT Jul 26, 2016

Post Editing of Machine Translation.

 

Sheila Wilson  Identity Verified
Spain
Local time: 21:20
Member (2007)
English
+ ...
You should meet my husband, Tom Jul 26, 2016

Tom in London wrote:
Illness?

I thought PEMT must be some sort of illness. I've read the posts and I still don't know what it is.

He has a very similar sense of humour. Just occasionally he gives the game away by revealing some knowledge of football, computing, America politics or whatever but most of the time he prefers to act dim icon_smile.gif.


 

Maxi Schwarz
Local time: 15:20
German to English
+ ...
starting with this Jul 26, 2016

John Fossey wrote:

According to a recent study by Common Sense Advisory ....

I've been reading quotes of things said by the CSA for years it seems and was always puzzled that what they said about translation didn't seem to reflect my experience. But a name and title like that lends it a certain aura of authority which we as individual translators can't seem to project. So a while back I took part in a survey, and since I took part, the results and conclusions were sent back to me. My impressions:

When I took part in the survey, the style of questions did not allow me to state my actual opinion or experience. The questions were highly slanted toward large translation companies. To date, my response has been totally ignored.

These large companies and other smaller ones are making money out of translations. The big time expenditure and thus cost for such companies is administration: managing each translation request by each customer, handling dealings with the translator etc. Therefore they like to streamline and automate processes, handle large projects (meaning less individual things to administrate), offer discounts for the large projects etc. It's like the assembly line of translation. And these are the entities whose responses are reflected in these studies. It is not the whole picture.

Going on to MT. The prevailing thing is to have PEMT paid by word at a low rate. Thus you could run a translation through MT at zero cost, hire a translator to post-edit for .03/word and essentially you're getting a translation for a low .03/word price. The translator, however, has double the work when MT is used inappropriately producing a poor product. If he translates his earnings into "per hour" it is slim pickings indeed. And fixing a poor translation cannot give the same quality as doing it properly from the start.

If that same post-editing were charge per hour so that the actual time spent were reflected, it would probably be much less popular among these companies.

All that said, if my impression is correct that the experience of the professionals who do the translations - rather than their customers - are barely taken into account, then how can we consider anything in these reports?

(Sorry for the length)


 

Bernhard Sulzer  Identity Verified
United States
Local time: 16:20
English to German
+ ...
Don't fall for it! Jul 26, 2016

Christine Andersen wrote:

Translators need to accept that PEMT is not going to go away ...


I disagree. Colleagues need to find out what it is. No one has to accept it. PEMT is a misnomer (see second link below) and is simply used to drive down rates for "translations" and dupe new translators into working for a pittance. Surveys claiming otherwise are certainly very questionable.

http://www.proz.com/forum/money_matters/303881-is_post_editing_a_pricing_scam.html

More detailed discussion:
http://www.proz.com/forum/translation_theory_and_practice/282023-post_editing_machine_translations_is_a_misnomer_but_there_are_now_training_sessions_for_it.html

PS: Remember, machines don't translate.

[Edited at 2016-07-26 19:02 GMT]


 


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

To what extent do translators need to take account of PEMT?

Advanced search







TM-Town
Manage your TMs and Terms ... and boost your translation business

Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.

More info »
Anycount & Translation Office 3000
Translation Office 3000

Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.

More info »



Forums
  • All of ProZ.com
  • Term search
  • Jobs
  • Forums
  • Multiple search