Désolé, cet article est seulement disponible en Anglais Américain.
In my view machine translation it makes my work quick as I translate easy. for instance, when i am using Autshumato ITE i am also develop my own glossary for future so that my pace of translating will be fast. Both human and machine scientifically are directly proportional ,entailing that they need each other.
Using platforms such as Linguee.de and Google can be very useful and much quicker than physically opening the dictionary lying next to you. But as in the case of the dictionary, you have to think about the options the platform gives you and decide on a proper, appropriate solution for your text, based on your feel for the source and target languages and the type of text you’re dealing with.
If by machine translation you mean back-translation, this can also be useful to ensure that no elements of meaning have been lost. But, as in the case of dictionaries, CAT tools and other platforms, you cannot translate without the “human intervention element”.
We as Afrikaans translators represent the youngest language in the world which exists next to one of the oldest and most developed languages in the world, English. For the past several decades we had to find countless equivalents for English words and phrases in a tsunami of new technology with all its accompanying terms. Unlike translators in Dutch in the Netherlands, we did not simply incorporate English words into our language, avoiding Anglicisms.
A translator using machine translator will not remotely have the translation and linguistic experience I have, in that I had to look up words on a large scale in a very laborious fashion to reach the levels of expertise at which I am operating. Machine translation can never match the subtle linguistic skills I have, translating according to the demands of different audiences. Extremely subtle creative skills are required in e g advertising, which are mainly creative skills, which no machine can acquire–not even humans can acquire these skills even after they have been doing run-of-the mill translations for half a century. In post-graduate translator schools they only touch on this area near the end of the course–they do not have the skills themselves to impart to others.
I love my Trados software! Yes, machine translation can be terribly incorrect, but I find it easier to fix the machine translation than to do everything from scratch. I can’t imagine translating without my software!
I think it is transforming our jobs, from translators to revisers/editors of machine-translated texts. Because we are always more pressed for time and increased productivity, quality is going down and, to take the example of French-English translation, we end up with a lot of Frenglish. The sentences are not necessarily bad or wrong but poor, and we’ll end up losing linguistic diversity.
Machine translation fits in the output-intensive orientation of a logic geared mainly to cutting costs. Never mind quality, job satisfaction or proper communication.
Yet, machine translation is regularly improving and it can be a great help, particularly with technical, repetitive texts, or very straighforward texts.
Hi, having started my translation career on the late (3 years ago, after 15 years in sales and marketing), I found Word tedious and started using Wordfast pretty much straight away as I needed a tool to ensure consistency of terminology on some large projects, and haven’t looked back since. In many ways, using translation tools compensate for my lack of academic background in translation in terms of speed of work and consistency. I am now looking at using a voice-recognition software to increase my output, so I can tackle urgent projects without having to sacrifice my evenings and weekends.
Machine translation will never replace human translation in my eyes, but for basic translation (i.e. emails), it serves its purpose. The companies that are happy to get edited machine-translated texts would probably only be happy to pay for low-quality high-output translators, whose quality is on par with MT.
Decent translators will always be needed, they might just have to adopt new technologies and give themselves an “upgrade” now and then. 😉
I have been working as a freelance translator for more than 20 years. Since the advent of technology, my rate of production has improved enormously.
I consult several different on-line dictionaries and glossaries, although I have also compiled my own subject-specific vocabulary files. It makes a big difference to be able to instantly check several sources to ensure that the correct term or accurate translation is used.
Without technology, I would not be able to work as I do, anywhere there is an internet connection, which has made such a difference. I seldom if ever meet my clients and colleagues, yet I have a wide supportive network, which would not be possible without the many advances in technology.
Machine translation has been suggested more than once by various friends and colleagues, yet I am reluctant to commit to it. Mainly, I believe, because of the many occasions I have had to proof-read and edit documents which were machine-translated, and much of the meaning has been lost, or the sense of a piece missed altogether. For straight and simple texts, to give the basic sense of a document, it may work well, but languages have so many different nuances, and I feel that a machine cannot possibly pick up the subtleties and convey the true meaning.
Technology certainly has made my professional life easier: Google (Translate) is my best friend, although I do consult the one or other (hardcopy) bi-lingual dictionary. Machine translations save a lot of time, thus enabling me to take on more projects, and I regularly search the Internet to check on certain facts. Yet, human intervention is imperative, so I never solely rely on what the computer gives me and always use my own (or collective) knowledge, especially when it comes to fine tuning.
Hi, everybody! Thanks to FIT and to Aljoscha for the invitation!
My name is Martín Chamorro, I’m an English-Spanish translator from Argentina. I’m a member and collaborator of the Argentine Association of Translators and Interpreters (AATI), member of FIT.
I’ll try to be as brief as possible 🙂
As regards my background as a translator, I got my BA in Translation eight years ago and have been using translation software for around nine years now. During the past five years, I’ve also been training colleagues in the use of CAT tools (in groups and one-on-one, online and in person, for now, from eight countries in the Americas and Europe). I have an extensive, advanced knowledge of the SDL platform and have used several other solutions, such as memoQ, Wordfast, Transit, memsource, smartCAT. I currently work as a translator, editor, proofreader and as a project manager every now and then. While my language combination is EN-ES, I also participate in projects which involve many language combinations, including left to right and languages that are not listed in all platforms. Since I make use of these tools most of the time I’m involved in a translation or localization project, I have gained a lot of knowledge in terms of troubleshooting and software logic/architecture.
Despite my extensive experience with file-based and cloud/server-based computer-assisted translation, I have very little experience using machine translation. However, I feel that it’s about time I started learning more about this part of the process, just to be more equipped and keep up with the times. I hope QT21 proves a good chance to do so. In any case, anything I can say about MT is going to be sort of prejudicial, but at the same time rather unbiased.
As a user ot MT, I find it very useful mainly for two tasks:
• For translation: In my work life, I sometimes resort to MT to untangle very complex or obscure passages or, rather unfrequently, I get pre-translated projects whose textual structure have a strong effect in the final outcome (I always feel that it sounds rather unnatural), which I find somehow annoying. I don’t actively use MT, just because. Outside my work life, I like to teach myself other languages, and for that, I’m proud to say I use Google Translate a lot (I find Hebrew into English as a very optimal combination, with even better results than German into English, for instance).
• As a PM, I have colleagues who systematically (and mostly without consulting) use MT sensibly and others who do so recklessly (lol). For the former, it’s fine: we need to embrace the development of technology and make use of resources that, ultimately, are meant to help us. For the latter (in this case with languages other than my working languages), I have received feedback from proofreaders saying that it really showed that, from the beginning, the material was translated automatically. I’m assuming that, currently, the goal would be to find something somewhere in between these two possible scenarios.
I would like to add that it’s also important to keep in sight the two sides of the coin: on the one hand, how machine translation drives productivity and, on the other hand, how this also puts new demands on all the team. But I guess that would be another part of the matter.
Some time ago, I heard that the letter combinations MT or AT have been sort of taboo in the translation market and networks, but that later these have been better received in the community. In my opinion, this is a good moment to make the most not only of the technological resources and social communication available, but also of the flood of material that comes from us users every minute.
Many thanks for reading through!
Email (will not be published) (required)