| Pages in topic: [1 2] > | Poll: Are you doing post-editing? Thread poster: ProZ.com Staff
|
|---|
| | neilmac
Spain
Local time: 18:36
Spanish to English
+ ...
Not for agencies anyway. Most of my revision (or proofreading) jobs are academic papers written by non-native authors, and nowadays they are obviously leveraging MT software like Google translate or others. It's no skin off my nose.
[Edited at 2019-11-27 08:17 GMT]
| | | | | Just out of interest | Nov 27, 2019 |
in view of the discussions on here about post editing, rates being pressured etc. I was curious to find out how many at least on this site are doing post editing or might.
I haven´t done any yet.
| | | | Samuel Murray 
Netherlands
Local time: 18:36
Member (2006)
English to Afrikaans
+ ...
Janet Muehlbacher wrote:
Are you doing post-editing?
Sometimes. But it is my impression that most of what agencies currently call "post-editing" is simply machine-translation assisted human translation with a higher words-per-hour expectation. A good client of mine recently started offering some post-editing jobs, claiming that they are using a custom machine translation engine, but... their "custom" machine outputs almost exactly what Google Translate's old-style statistical engine puts out. When a client offers you post-editing, and you're willing to consider it, assume the worst.
I think what puts many people off from post-editing, apart from the lower rate, is the fact that most translators have never learnt how to deliberately vary the quality of their translations. Such translators lack the skill of knowing when a poor quality translation is still of an acceptable standard and when it becomes unacceptable. Some translators believe that only the best is acceptable. With post-editing, the required knack is to lower the final product quality while still achieving what the client considers acceptable.
Clients usually give translators an indication of the required quality. It's usually in the per-word rate. If a client usually accepts 10c per word for translation, but wants to pay you 5c per word for post-editing, it means that he is expecting you to output twice as many words per hour as you would normally output in an hour. So, for the translator, it's important to know how many words, on average, he can translate (and self-proofread) up to final-delivery standard in 1 hour.
Janet Muehlbacher wrote:
In view of the discussions here about post-editing, rates being pressured, etc...
Long ago, I participated in a very large translation project that used a client-supplied CAT tool, that involved many translators who had never used a CAT tool before. There was an outcry when the final payments were calculated, and the agency got extremely bad reviews from those translators, because they did not understand fuzzy discounts (even though it was specified in the PO beforehand), and they felt that they were being paid less. In reality, the project had a very fair fuzzy discount scheme and good quality matching, so at the end of the day they must have earned more than they would otherwise have. These translators kept track of how much they were going to earn by tracking the wrong figures (i.e. they kept seeing the total word count, and casually multiplied that by their per-word rate in their heads).
I think a similar thing is true of people who complain about post-editing rates: there is no reason why you should earn less per hour. If you do earn less per hour (and you're not simply inexperienced), then it means that you're putting too much work into the final product.
[Edited at 2019-11-27 08:58 GMT]
| | |
|
|
|
I haven't been asked... yet.
| | | |
your remarks are very interesting, in particular with regard to deliberately lowering the quality of the end product. Most translators would indeed try to do their best instinctively, I think. Do you think the customer would notice the difference?
| | | | | No, but it is probably coming | Nov 27, 2019 |
Samuel Murray wrote:
I think what puts many people off from post-editing, apart from the lower rate, is the fact that most translators have never learnt how to deliberately vary the quality of their translations. Such translators lack the skill of knowing when a poor quality translation is still of an acceptable standard and when it becomes unacceptable. Some translators believe that only the best is acceptable. With post-editing, the required knack is to lower the final product quality while still achieving what the client considers acceptable.
Clients usually give translators an indication of the required quality. It's usually in the per-word rate. If a client usually accepts 10c per word for translation, but wants to pay you 5c per word for post-editing, it means that he is expecting you to output twice as many words per hour as you would normally output in an hour. So, for the translator, it's important to know how many words, on average, he can translate (and self-proofread) up to final-delivery standard in 1 hour.
These are the critical points.
After years of trying to explain to clients that quality is all-important, and training, checking and struggling for perfection, it is not easy to accept anything less than the best.
However, a lot of daily communication is simply not great prose. It is repetitive and predictable, and a message may only be useful for a few minutes before it is mulched or deleted. It is in fact a waste of time to agonise over every little comma or odd formulation.
That is what I find difficult!
| | | | | MTPE work better than no work | Nov 27, 2019 |
Translation-related work is my main source of income, so a fat NO can't apply before I find a new "revenue stream".
My experience is limited to customer-trained MT engines in specific subjects with controlled language source, in the ubiquitous EN>FR combination. I have no working experience with GT, DeepL and other free (or paid) tools and MT plug-ins.
My first encounter with MTPE was in 2000 or 2001 with an agency I knew well. Some 8-9 years later, I embarked on quite ... See more Translation-related work is my main source of income, so a fat NO can't apply before I find a new "revenue stream".
My experience is limited to customer-trained MT engines in specific subjects with controlled language source, in the ubiquitous EN>FR combination. I have no working experience with GT, DeepL and other free (or paid) tools and MT plug-ins.
My first encounter with MTPE was in 2000 or 2001 with an agency I knew well. Some 8-9 years later, I embarked on quite a bit of MTPE for an agency, more because I knew them well than sheer excitement.
I regret to report that both these experiences weren't a disaster, including economically. But I realised how boring and tiring MTPE gets after a few weeks.
On the third encounter 6 years ago, a first (and last) small MTPE job reminded me that I should always stick to the rules I have set to myself and committed to comply with, one of which being : avoid massive global translation groups like the plague.
My own observation: between my first and last attempt at PEMT, I haven't noticed any improvement in raw MT on online help and, more generally, on accidentally or purposefully MT-friendly texts. I remember being positively impressed with EN>FR raw output as soon as 2000 or 2001, and it feels like nothing has happened since.
Anyway.
Global players can spend as much as they want on MT improvements, announce break-throughs every day and try to convince us that working on a word-factory line is hip, the translation world still has a lot of room for the real thing. It's just called transcreation nowadays.
Philippe ▲ Collapse
| | |
|
|
|
Mónica Algazi
Uruguay
Local time: 14:36
Member (2005)
English to Spanish
"Global players can spend as much as they want on MT improvements, announce break-throughs every day and try to convince us that working on a word-factory line is hip, the translation world still has a lot of room for the real thing. It's just called transcreation nowadays."
Long live TRANSCREATION, then!
| | | | | I've given up | Nov 27, 2019 |
I took on a few post-editing jobs a while ago and the quality of every text was appalling. I ended up spending more time correcting the 'translation' than I would have done had I prepared the translation myself from the outset without a CAT tool. Some of the terms used were inconsistent (which I fail to understand as I thought that the whole point of a CAT tool was to ensure consistency) and various sentences had no meaning whatsoever in English. For the small fee generally on offer for post-edi... See more I took on a few post-editing jobs a while ago and the quality of every text was appalling. I ended up spending more time correcting the 'translation' than I would have done had I prepared the translation myself from the outset without a CAT tool. Some of the terms used were inconsistent (which I fail to understand as I thought that the whole point of a CAT tool was to ensure consistency) and various sentences had no meaning whatsoever in English. For the small fee generally on offer for post-editing jobs, I would rather spend my time marketing for translation work, so no longer take on post-editing jobs. ▲ Collapse
| | | | | This kind of links up | Nov 27, 2019 |
with the poll "where is our industry heading" although it´s not strictly representative, but it seems the industry is putting so much pressure on translators that they really will adapt and take PE because it´s better than nothing.
| | | |
and hated it. It's like trying to fix someone else's mess.
| | |
|
|
|
Samuel Murray
Netherlands
Local time: 18:36
Member (2006)
English to Afrikaans
+ ...
Janet Muehlbacher wrote:
Your remarks are very interesting, in particular with regard to deliberately lowering the quality of the end product. Most translators would indeed try to do their best instinctively, I think.
Not all translators have all the skills that the perfect translator has. I myself lack certain skills that other translators do have -- sometimes because I haven't had much practice in those skills and sometimes because I'm simply not intelligent in that particular way.
Not all translators (and proofreaders/reviewers) are, for example, capable of distinguishing between objective edits and preferential ones. Some translators will always, when asked to proofread/review, make many edits that they believe "improve" the text, even though those edits may actually be preferential. Some translators are simply incapable of fixing only certain classes of errors, and will always try to deliver text that they want to be proud of (even if the client specifically tells them not to), because they feel a moral obligation to do so.
It is a skill to distinguish between necessary edits and... oh, I don't want to call it "unnecessary" edits... erm, well, "non-required" edits.
Do you think the customer would notice the difference?
Yes, I think that if the client understands the target language and has a high expectation of the translation, he'll definitely notice the difference. But that's his decision, if he chooses (possibly after receiving bad advice) to get a translation done using a method that produces poor translations.
I certainly am upset about project managers and translation agency owners who tell clients that they'll get the same quality for cheaper if they use post-editing. The only way clients can get the same quality with post-editing is if they pay per hour, and if they don't limit the hours. This does mean, however, that sometimes post-editing is going to cost more than if a translator had simply translated from scratch -- something that project managers and agency owners don't want to admit to their clients.
I know of colleagues who accept post-editing jobs (along with reduced rates) and then simply translate from scratch (or translate using their preferred machine translation provider), because actual post-editing would take too long.
[Edited at 2019-11-27 14:13 GMT]
| | | | Tom in London
United Kingdom
Local time: 17:36
Member (2008)
Italian to English
| Alexandra Hirsch (X)
Austria
Local time: 18:36
English to German
+ ...
| Companies don't budget for it | Nov 27, 2019 |
@Christine
I think you are right about most communication not being worth fretting over if it's less than perfect. However, -- and here comes my defence of the translator's guild -- it can rankle with readers if texts are a complete nonsense (e.g. Amazon product listings), read like the writer has a stick up *** (marketing copy) or doesn't know the language (emails sent out by electronics giant Canon), and it would deter me from buying a product (well, most of the time anyway).
... See more @Christine
I think you are right about most communication not being worth fretting over if it's less than perfect. However, -- and here comes my defence of the translator's guild -- it can rankle with readers if texts are a complete nonsense (e.g. Amazon product listings), read like the writer has a stick up *** (marketing copy) or doesn't know the language (emails sent out by electronics giant Canon), and it would deter me from buying a product (well, most of the time anyway).
I think most large companies using MT are doing so because they had to cut costs (to keep product prices low) and the translation office was part of the general cull. Why would they spend more money on having someone edit the stuff? ▲ Collapse
| | | | | Pages in topic: [1 2] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Poll: Are you doing post-editing? | Pastey | Your smart companion app
Pastey is an innovative desktop application that bridges the gap between human expertise and artificial intelligence. With intuitive keyboard shortcuts, Pastey transforms your source text into AI-powered draft translations.
Find out more » |
| | Wordfast Pro | Translation Memory Software for Any Platform
Exclusive discount for ProZ.com users!
Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value
Buy now! » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
