7月 24, 2019 |
Machine translation has had a significant impact on the language services industry. However, many people continue to treat it as a sort of witchcraft. Some overrate the abilities of the software and are ready to bet a company’s brand reputation entirely on it. The other camp wants nothing to do with CAT tools, AI, and the like–and keeps asking for “100 percent human translation” services.
Machine translation myths don’t make things any easier for language services providers. Here are five of the most common beliefs about machine translation and the truth behind them. Which one of these did you believe in?
You’ve probably heard this one before. “Machine translation” has somehow become synonymous with Google Translate which automatically makes anyone with a connection to the internet a translator. Google’s translation software is the best-known tool so far. But that doesn’t mean all professional translators use it for work. In fact, most of them don’t.
Google Translate is like the free version of every software, video game, or app out there. It can only take you so far. The software’s failings have gone viral more than once, so you can’t expect its translations to be accurate.
The language services market relies on a wide range of CAT translation tools. Among the most used in the industry are SDL Trados Studio, ProZ, MemoQ, Zanata, SmartCAT, and Microsoft Language Portal.
These may be lesser-known machine translation tools, but they give better, more consistent results than Google Translate. Moreover, many of these CAT translation tools come with specific features that enable translators to customize the software for every new translation project.
This myth is probably as old as the machine translation itself–and it’s also the most persistent. Moreover, the rise of neural machine translation (NMT) has increased the number of people who believe that software can do the job without linguists.
The opposite is true. Machine translation can achieve performances that a human translator can’t, such as scanning large documents in seconds and providing an understandable version in the target language. However, software alone can’t reach the accuracy of the human mind.
Computers are now able to deliver translated content that is fluent in the target language. But it’s still can’t overcome the significant challenges in translation, such as understanding humor or sarcasm.
Moreover, it’s hard for software to adapt the new version of the content to specific audiences, based on their interests or cultural background.
As with most technologies, human translators are an essential part of the translation process. Machine translation only helps people evolve and improve their services. It can’t replace the human factor in communication. Not for now, at least.
This myth is partially correct, in the sense that translators who use CAT translation tools have lower costs than professionals who do everything manually. That also means that reduced expenses could generate a lower price for translation services.
Unfortunately, the story in which you skip language professionals and use some premium version of software to cut down translation costs is far from the truth.
Machine translation is practical and very useful in the right hands. However, without a linguist in charge, your translated content will likely be weak and barely intelligible. Bad content won’t convince your business partners about your abilities and it certainly won’t sell your products.
Grammar still matters in all languages. Whether you’re translating content for your marketing brochures, website, or internal communications, there’s a high chance you won’t get the action you are calling for if your message is poorly constructed.
Bad translations can affect your international reputation and marketing strategies. You need to express yourself clearly to get positive responses from your audience.
This myth appeared in the team that advocates against machine translation. It’s the old “man versus machine” story that has been fascinating people for over a century. According to this misconception, professionals shouldn’t take the easy way and should continue to translate in the labor-intensive old-fashioned way.
The reality is that when used right, software can only benefit translators. In today’s world, and probably in the future, humans and computers need to work together. This approach enables linguists to translate content faster and with fewer human errors. At the same time, CAT translation tools make translations cheaper and consistent.
When translators use machine translation, they have faster turnarounds as they can translate lengthy texts in less time. The software allows them to use integrated glossaries of terms that make translation easier than ever before.
The concept of “100 percent human translation” is obsolete and highly inefficient. It only slows you down and doesn’t guarantee better results.
Linguists have all the information and terminology they need right at their fingertips. No more going through industry-specific dictionaries to find the exact match for a technical term. Everything is already inside the computer, ready to be used in seconds.
In short, you combine the speed and consistency of machine translation with the language skills and subject matter expertise of a human translator. It’s the best of both worlds, at an affordable price.
There’s no doubt; the best machine translation tools out there come from tech giants like Microsoft, Caterpillars, or IBM. However, it doesn’t mean you should evaluate language service providers only by the tools they use.
While the performance of CAT translation tools is important, it doesn’t affect your bottom line as much as the experience of the translators. Excellent software alone may deliver above-average translation. But a translator without the right skills or with no experience in the niche can cause harm, even with the best neural machine translation tool.
The benefits of using machine translation are hard to ignore. From reduced costs to faster and more consistent results, the collaboration between humans and machines can make the difference to your translation project.
What matters is that you see machine translation for what it is. It’s not a substitute for humans, nor a guarantee for failure. It’s just an excellent tool in the right hands.