| பதிவின் பக்கங்கள்: [1 2 3 4] > | Saying no to MT இழை இடுபவர்: Phil Hand
|
|---|
Phil Hand 
சீனா
Local time: 12:37
சைய்னீஸ் - ஆங்கிலம்
Sorry to raise again an issue that's been discussed many times.
But I've been thinking about this, and was prompted by a number of post-editing jobs posted recently, and by a comment made to me by a colleague. He said to me (I paraphrase): MT is inevitable, but high quality translators like you and me will be alright.
Now, I've actually expressed similar sentiments before. But the way he put it just instantly brought to my mind that famous poem by Niemoller: First they ... See more Sorry to raise again an issue that's been discussed many times.
But I've been thinking about this, and was prompted by a number of post-editing jobs posted recently, and by a comment made to me by a colleague. He said to me (I paraphrase): MT is inevitable, but high quality translators like you and me will be alright.
Now, I've actually expressed similar sentiments before. But the way he put it just instantly brought to my mind that famous poem by Niemoller: First they came for the communists, and I didn't speak out...
That's a rather extreme comparison, I know, and I don't mean to imply that MT is as bad as the Third Reich! But it made me reflect, and now the argument as it relates to MT seems equally wrong to me.
If we concede now that MT is OK for some purposes, then somewhere down the line, someone will conclude that MT is good enough for the translation that YOU are doing. And then you've got no job.
More than that, MT is a dehumanising process. For those of you who have read any translation studies, the academics all very earnestly tell us that translation is a form of communication. Well, it's not when a computer does it. When a computer does it, translation is a form of calculation. And that's not a good way to interact with people. Human interaction is an inherently valuable thing, and it is our job to support it in our little corner of the world.
Finally, MT isn't very good, and for many many years to come will remain not very good. We can all chortle at the silly mistakes made by Google Translate, where Berlusconi can only ever win elections. But the fact is that certain corporations are starting to rely on computers to translate large volumes of certain types of material. And one of these days, they're going to made a dangerous mistake. Perhaps they already have, and we just haven't heard about it.
I don't want to deny that computers can be of use in translation. Translation memories, termbases, Google and literature searches can be invaluable tools on some projects. But the fact that computers can add up really well doesn't mean that we would ever consider letting a computer do our accounts. Everyone knows that you need a skilled human accountant for that. In the same way, the fact that computers help us translate better doesn't mean that we should let a machine do our translation for us. However, at this point in time, apparently not everyone knows that you need a skilled human translator to translate between languages. They should know, and we should tell them.
And that's my point. MT is a creeping attack on our profession that is 1) wrong-headed, 2) dehumanising, and 3) very preventable.
I would like my fellow translators to help with the beginnings of a resistance to MT. We should tell all clients, loudly and consistently, that MT is not OK; that post-editing will not save them money in the long run; and that we should be valued for the skills, experience and humanity that we bring to the role of translator.
I'm inspired in part by the "No Peanuts" website and movement. Like them, I'm not really sure how to go ahead. I'm not a member of any associations yet, but they might be the places to start: if associations were to state clearly that they do not recommend MT, we could work from there. I don't know if there are any IoL or ATA members out there who would be willing raise such a proposal? ▲ Collapse
| | | | Ty Kendall
யுனைடட் கிங்டம்
Local time: 05:37
ஹெப்ரூ - ஆங்கிலம்
| Over my dead body | Nov 1, 2011 |
Phil, have you been reading my mind?
I recently noticed a MT editing job in my language pair, (also listed a dozen other languages) for a "Potential Job", whatever that is.
Now, I'll be honest, for a split second I considered it. Then I came to my senses.
In my language pair at least, MT is pretty diabolical. (The languages are quite divergent - as I imagine they are in your pair too) So any MT editing job (in my languages + some others) is basically asking for translati... See more Phil, have you been reading my mind?
I recently noticed a MT editing job in my language pair, (also listed a dozen other languages) for a "Potential Job", whatever that is.
Now, I'll be honest, for a split second I considered it. Then I came to my senses.
In my language pair at least, MT is pretty diabolical. (The languages are quite divergent - as I imagine they are in your pair too) So any MT editing job (in my languages + some others) is basically asking for translation, because that's practically what I'd be doing with the gobbledygook that MT spews out.
MT for me is just a glorified dictionary, the time it might save me looking up the odd word is cancelled out by me having to check the syntax and phraseology (MT frequently translates Hebrew negative sentences into positive ones - despite the abundance of negative particles).
Yet, the people who "offer" post-editing jobs think it is justified to promote this practice and charge proofreading prices (or lower)....
...It's a bit below my dignity to agree to that. I'd rather go back to my 9-5 monotonous soul-destroying office job.
P.S. What I also noticed - the company "offering" MT editing - their website has a veneer of human translation (lots of pictures of happy smiling translators) with no mention that a computer will be having first dibs to the client's material.
-They also have a "testimonial" from a translator they have supposedly used, which makes it clear they employ translators who translate into languages they do not have a solid grasp of, i.e. the testimonial mentions translation into English, but then is littered with bizarre English: "Pleasure has been mine working with you. I definitely love to work for you again in the near future". The icing on the cake is the evidently non-English name of the translator just to confirm that they use non-natives (**with a tenuous grasp of English) to translate into English. Hmmm.
(Before I'm lynched by non-native English speakers, ....Yes, they may well be proofread, but by using a translator with such visibly poor English in the first place - doesn't fill me with confidence that any proofreader they employ (if employed) will be of a higher quality than the translator they chose.
The battle of CATtools is pretty much lost (but I don't currently use these either), but this is a battle that can still be won by translators as long as we stop being doormats and stand up for best practice - which is human translation.
In a nutshell - MT bad. Human translation good. ▲ Collapse
| | | | | Out of our hands | Nov 1, 2011 |
I thought I might start a revolution, or at least a trend, when my biggest client was taken over by Lionbridge and I refused to work for them any more, but it didn't happen... I think we have to accept that CAT and MT (and globalisation in the case of translators into English) will gradually reduce the amount of work for conventional translation, and there is nothing we can do about it. It's all about market forces. If MT really was that bad, people wouldn't use it. Clearly these postediting jo... See more I thought I might start a revolution, or at least a trend, when my biggest client was taken over by Lionbridge and I refused to work for them any more, but it didn't happen... I think we have to accept that CAT and MT (and globalisation in the case of translators into English) will gradually reduce the amount of work for conventional translation, and there is nothing we can do about it. It's all about market forces. If MT really was that bad, people wouldn't use it. Clearly these postediting jobs allow somebody somewhere (though obviously not the translator) to make more money, just like CAT does, so these jobs will only grow in number. I actually used Google Translate for the first time yesterday to read a Spanish website, and yes it was crap but it helped me find the information I wanted so it was also brilliant. So MT definitely has a place in the market. End-customers will only pay if the end-product is acceptable to them, so perhaps the issue is not MT but expectations and the bastardisation of English on the Internet. I think most CAT translation and indeed most human translation is substandard, but most end-customers are clearly happy with it. So it all comes down to demand and expectations IMO. I have maybe 20 years left until retirement, and based on how things have come on over the last 20 years (WP5.1 anyone?), I'm now looking to get more copywriting work
In haste... ▲ Collapse
| | | | Keith Boak
போர்த்துகள்
Local time: 05:37
போர்சுகீஸ் - ஆங்கிலம்
+ ...
| Yes but ... it depends on the rate. | Nov 1, 2011 |
I fully agree with the sentiments, and I almost always refuse to edit MT translations, but in reality, there is a price that I will do it for - my standard translation rate (or close to it). So rather than say "no", I often say, OK, but what you have here is very dangerous document full of the usual MT errors, and if you want me to sort it out, you will have to pay my normal rate. Occasionally it works, but more importantly, it helps educate the clients about the minefield they're stumbling into... See more I fully agree with the sentiments, and I almost always refuse to edit MT translations, but in reality, there is a price that I will do it for - my standard translation rate (or close to it). So rather than say "no", I often say, OK, but what you have here is very dangerous document full of the usual MT errors, and if you want me to sort it out, you will have to pay my normal rate. Occasionally it works, but more importantly, it helps educate the clients about the minefield they're stumbling into.
Another side to this is that it's worth the translator trying MT on some projects, then doing the edit/re-translate to see if it is in fact any quicker, and also in order to learn about repetitive MT errors in their language pairs. This helps you quickly spot MT projects which are not openly declared as such. ▲ Collapse
| | |
|
|
|
| Newbies don't know what "translation" means | Nov 1, 2011 |
I just thought I'd mention that --as revealed by the many topics about "MT does not work"/"Google Translate does not work" in the fora recently-- newbies are cheerfully embracing MT for their "translations".
In doing so, they are failing to put in practice whatever they have learned in their translation studies, and are outsourcing the groundwork of their translation to a machine. While I understand that MT could be useful in to-open-the-file-click-Open translations, I think these t... See more I just thought I'd mention that --as revealed by the many topics about "MT does not work"/"Google Translate does not work" in the fora recently-- newbies are cheerfully embracing MT for their "translations".
In doing so, they are failing to put in practice whatever they have learned in their translation studies, and are outsourcing the groundwork of their translation to a machine. While I understand that MT could be useful in to-open-the-file-click-Open translations, I think these translators are ruining the whole concept of quality and real internationalisation our profession has invested so much effort and training in.
Newbies --or older translators recently mesmerised by the slight reduction in typing-- limit themselves to accept whatever Google Translate may feed them, with some changes at first, and with a much lower criticism threshold as time goes by ("after all, who cares?"). The result is catastrophic, I am affraid, and not just for their work, but for the idea of translation among their customers. ▲ Collapse
| | | | Phil Hand
சீனா
Local time: 12:37
சைய்னீஸ் - ஆங்கிலம்
தலைப்பை ஆரம்பித்தல் | Not inevitable | Nov 1, 2011 |
Ty - Thank you
Autor - absolutely, educating clients is what it's all about
Chris S - you're exactly the person I'm trying to convince here.
MT is NOT inevitable. The market is NOT always right. The market is nothing more than a collection of fallible human beings making decisions. We can help them to make better decisions.
I know exactly what you mean about using Google translate to navigate around a website. I've done that. But that is a far cry fro... See more Ty - Thank you
Autor - absolutely, educating clients is what it's all about
Chris S - you're exactly the person I'm trying to convince here.
MT is NOT inevitable. The market is NOT always right. The market is nothing more than a collection of fallible human beings making decisions. We can help them to make better decisions.
I know exactly what you mean about using Google translate to navigate around a website. I've done that. But that is a far cry from actually translating a text (or a website).
I'd like to draw a sharp distinction between CAT and MT. I'm not a big CAT fan, but there's a world of difference between using Trados to ensure consistency throughout a big project (and accepting that market forces have imposed a cost for doing so), and getting a computer to translate your project for you.
Like I say, I've been guilty in the past of saying, "I'm a top end translator, I'll be OK. I'll just hang those other guys out to dry." Increasingly, though, I think that's wrong. A lot of the people who chat here seem to be reasonably comfortable translators, with enough work and reasonable rates. It's kind of our responsibility to look out for those who are not in that fortunate position - and ultimately, it will benefit us. If we send a message to our clients that it's OK to use MT for some projects, I assure you it is only a matter of time before they start to use MT for all projects.
We are the experts. We see the problems that our clients cannot see. If we don't educate them, no-one will. ▲ Collapse
| | | | Ty Kendall
யுனைடட் கிங்டம்
Local time: 05:37
ஹெப்ரூ - ஆங்கிலம்
| MT will relegate us to damage-limitation clerk | Nov 1, 2011 |
I've always thought it strange that translators are so willing to roll over and accept practices that are detrimental to their own career and long-term prospects in the name of "inevitability". If we look at the future with those goggles on, shouldn't we just pack up and admit defeat now? Go back to whatever passed as a "career" before we started translating.
Do you think that if doctors were told that soon in the future, most surgery will now be done by barely trained field medics ... See more I've always thought it strange that translators are so willing to roll over and accept practices that are detrimental to their own career and long-term prospects in the name of "inevitability". If we look at the future with those goggles on, shouldn't we just pack up and admit defeat now? Go back to whatever passed as a "career" before we started translating.
Do you think that if doctors were told that soon in the future, most surgery will now be done by barely trained field medics and that their role would be damage-limitation after the fact, paid at half their usual salary, that they would accept that? Of course they wouldn't. But that's what we're being asked to do and we're doing it with a smile. (At least we'll get to keep our careers in some form - line of thinking). Is that enough?
The only inevitable thing I see is the passivity of translators in not preventing this. The market is mentioned, that market forces drive this, really?
I was always taught that markets work through supply and demand....
I concede that the demand is surely there for MT (it seems like a cheaper, faster alternative to Human translation-although this is false logic).
The real problem is that there is an abundant supply of translators willing to do this type of work for peanuts, thus completing the circle and driving the market on. If the supply dried up, the atrophy of the demand would surely follow.
The translator is the loser in all this. It's akin to correcting MT's homework, and then MT passes off the work as its own (at least this is the way agencies/clients will see it - the translator's role is overlooked and MT gets most the glory because no matter how bad the MT original translation is, the translator is thought of as an editor only).
This isn't something you have to accept, it's something that you choose to accept. Would those doctors mentioned earlier accept it on the basis of health care cuts? A concrete reality in this economic climate No. Would translators accept it on the basis of market inevitability.A hypothetical possibility far from a reality Unfortunately yes.
It's not too late to reverse the trend just yet.
Edited for a typo
[Edited at 2011-11-01 09:55 GMT] ▲ Collapse
| | | | neilmac
ஸ்பெயின்
Local time: 06:37
ஸ்பேனிஷ் - ஆங்கிலம்
+ ...
| Colleagues, I salute you! | Nov 1, 2011 |
Here here! It's nice to know I'm not alone in my reluctance to join in the MT scrum.
In fact, owing to the proliferation of dodgy translators/ropey translations and falling rates nowadays, just last week I edited my proz profile as follows:
"Owing to market intrusion, I no longer provide revision/proofing services for texts translated by other translators, or machine translations. My revision and correction services are now limited mainly to academic texts written in ... See more Here here! It's nice to know I'm not alone in my reluctance to join in the MT scrum.
In fact, owing to the proliferation of dodgy translators/ropey translations and falling rates nowadays, just last week I edited my proz profile as follows:
"Owing to market intrusion, I no longer provide revision/proofing services for texts translated by other translators, or machine translations. My revision and correction services are now limited mainly to academic texts written in English by non-native speaker authors or teams." ▲ Collapse
| | |
|
|
|
| Agree to differ? | Nov 1, 2011 |
I really do think it is inevitable, because MT can only get better.
I'm not sure about the relevance of analogies, and my history ain't great, but during the industrial revolution there were doubtless plenty of skilled craftsmen who grumbled about the big new factories springing up, and maybe some of them tried to rise up, but the factories won because they were cheap and the quality was adequate, so most customers were more than happy with their output, and most people were happy/... See more I really do think it is inevitable, because MT can only get better.
I'm not sure about the relevance of analogies, and my history ain't great, but during the industrial revolution there were doubtless plenty of skilled craftsmen who grumbled about the big new factories springing up, and maybe some of them tried to rise up, but the factories won because they were cheap and the quality was adequate, so most customers were more than happy with their output, and most people were happy/willing to work there cos a man's gotta eat. There will always be a market for hand-made jumpers and baskets and what have you, just not a very big one.
I have been gobsmacked in recent years to run across competition in my language pairs (Scandi to English) from Eastern Europe. WTF?! If customers are happy with that, they will also be happy with MT edited by people in Asia with good but imperfect English. While we rich Westerners can get all high and mighty about quality and rates and stuff, you can't expect that in less fortunate parts of the world.
That's not to say that I like it. I have battled for 20 years to convince customers that it is worth paying extra for quality, and it's a battle that is increasingly hard to win, but I will keep fighting! ▲ Collapse
| | | | | There is always hope | Nov 1, 2011 |
It is difficult to convice some clients, but in my opinion a good product/service begins with a correctly written text. Imagine an expensive product's brochure or manual full of errors. That's not good marketing for the company. Some day they will realise too
| | | |
"Owing to market intrusion, I no longer provide revision/proofing services for texts translated by other translators, or machine translations. My revision and correction services are now limited mainly to academic texts written in English by non-native speaker authors or teams."
What a fantastic way to make a point and educate your clients. Kudoz for you!
| | | | Ty Kendall
யுனைடட் கிங்டம்
Local time: 05:37
ஹெப்ரூ - ஆங்கிலம்
| We're not Luddites because we oppose bad practice | Nov 1, 2011 |
Chris S wrote:
I really do think it is inevitable, because MT can only get better.
We keep being told that though, don't we? By everyone. But does it have any basis in fact? I certainly have not noticed any great leaps and bounds in MT recently. I also have some experience working in a Computational Linguistics department at a University and whilst I don't doubt their enthusiasm, their results are quite sketchy. They're big on theory and endless experiments, but results were always thin on the ground, and not particularly impressive when they occur.
When will MT understand irony, or sarcasm? If I write "Yeah, right"** in some contexts, it can actually mean "No, I don't agree with you in the slightest/I think you're totally mistaken" etc. Try getting a computer to understand that. And this is a relatively simple example. The pace of technological advancement has slowed lately (the fact that Apple constantly release new versions of the iphone doesn't really count). So I'm not really convinced MT will get vastly better any time soon. For a computer to really understand and manipulate language in a way that would threaten translators, it would need to get to a point where it could understand the vast array of nuance and subtlety in language. But at the moment, as has been said time and again before, it often can't even get the basics right.
I'm not a Luddite, I'm not opposing technology for the sake of it, I would oppose any tool which is not fit for purpose, and MT certainly isn't fit for the purpose of professional translation. At the moment, it's handy if you want to get the gist of the odd webpage, but that's about it.
We shouldn't even consider employing substandard tools as replacements for human translation, much less embrace it as inevitable. I'll happily quit translating and go back into teaching or office work when a computer can produce a flawless translation in my language pair. Until that day......
**Reminds me of this linguistic joke:
A linguistics professor was lecturing his class the other day. "In English," he said, "a double negative forms a positive. However, in some languages, such as Russian, a double negative remains a negative. But there isn't a single language, not one, in which a double positive can express a negative."
A voice from the back of the room piped up, "Yeah, right."
| | |
|
|
|
neilmac
ஸ்பெயின்
Local time: 06:37
ஸ்பேனிஷ் - ஆங்கிலம்
+ ...
Chris S wrote:
If MT really was that bad, people wouldn't use it.
I'm afraid I have to disagree there. Just as many shop in the UK at Lidl or Aldi because they can't afford Waitrose or Claridge's every day, many clients will go for the cheapest option available, especially in times of recession (or whatever you wish to call the current quagmire).
| | | | Jaroslaw Michalak
போலந்து
Local time: 06:37
உறுப்பினர் (2004)
ஆங்கிலம் - போலிஷ்
SITE LOCALIZER | What's the difference? | Nov 1, 2011 |
neilmac wrote:
Chris S wrote:
If MT really was that bad, people wouldn't use it.
I'm afraid I have to disagree there. Just as many shop in the UK at Lidl or Aldi because they can't afford Waitrose or Claridge's every day, many clients will go for the cheapest option available, especially in times of recession (or whatever you wish to call the current quagmire).
But what are you disagreeing with? If some clients are satisfied with Aldi-quality translation, that is what they will pay for. Even if we keep telling them it is not up to our standards, it has no chance for a Nobel prize and it is a disgrace in general, that's what they want - substandard quality at bottom-rock prices. And there will be someone willing to provide it.
Actually, it has always been that way - remember those 0.03 jobs everyone is moaning about? Those will be the first to go the MT route.
This will certainly impact the market as a whole. It does not mean that everyone's rates will go down... I am certain that these days the tailors, shoemakers etc. earn more on the average than they did before. The only problem is that there just a few left! The same will be with translators - specialized, high-quality translations will be always in demand and they will be paid well. You just have to make sure you position yourself at that end of the market... Bottom feeders won't have a chance - if one can get bad translation at a low rate or bad translation almost for free, the choice is obvious.
| | | | Phil Hand
சீனா
Local time: 12:37
சைய்னீஸ் - ஆங்கிலம்
தலைப்பை ஆரம்பித்தல் | We should fight for the 0.03 jobs | Nov 1, 2011 |
First, Chris S:
You're fighting for quality, so I regard you as on my side already. We may disagree about the ultimate fate of the translation market, but if we agree about what we should be doing now, then that's much more important. What are we doing now? Educating clients, making clear that MT is inadequate, demonstrating the many positive aspects of human translators.
(For example, I don't know about you, but many clients of mine don't have proofreaders for their or... See more First, Chris S:
You're fighting for quality, so I regard you as on my side already. We may disagree about the ultimate fate of the translation market, but if we agree about what we should be doing now, then that's much more important. What are we doing now? Educating clients, making clear that MT is inadequate, demonstrating the many positive aspects of human translators.
(For example, I don't know about you, but many clients of mine don't have proofreaders for their original documents. As a translator, I perform that service as an easy value added. Just a note in the text saying, this figure looks wrong; you have a typo here; etc. Machines don't do that.)
Jabberwock:
This argument hurts to make, but we have to stand up for the "bottom feeders". First of all, they're not all bottom feeders. In my pair, interpretation can be quite well paid, but even good translators in China can't make much more than 0.02 USD per word for written translation.
Secondly, you're right, some of those very low-end jobs might go. But what I'm calling for here is that all translators tell clients consistently: if you've got a document that needs translating, it is *always* better to get a human to do it.
I can't guarantee that this message is the absolute truth. But I think it's a much better approximation to the truth than whatever money-grubbing compromise "the market" is going to come up with. And if we translators say consistently, human translation is always better, then people will listen. Of course they'll listen to us. Who else knows better than us?
Human translation is always better. ▲ Collapse
| | | | | பதிவின் பக்கங்கள்: [1 2 3 4] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Saying no to MT | Trados Studio 2022 Freelance | The leading translation software used by over 270,000 translators.
Designed with your feedback in mind, Trados Studio 2022 delivers an unrivalled, powerful desktop
and cloud solution, empowering you to work in the most efficient and cost-effective way.
More info » |
| | 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 » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
Translation news
