Both camps take the position that you must choose machine translation OR human translation. In reality, machines and humans can work together to deliver a quality translation quickly.
We’re living in a time when a growing number of tasks can be completed by machines. We use ATMs to get cash (unless we use debit cards). We use self-checkout at the store (unless we buy online). We use chatbots to resolve our service or delivery issues (unless the site’s FAQ has the answer). It’s no surprise that translation has become another target for automation.
And why not? Pages and pages of documents can be translated in a fraction of the time it would take a human to do it – and at a fraction of the cost. As language technology evolves, it gets smarter and develops a memory, retaining certain pieces of data that will appear regularly in your content. Plus, it doesn’t need to take coffee breaks or sleep through the night, so you can use it anywhere, anytime.
Although machines can cut translation time, language isn’t so easy to automate. Machines may not recognize the nuances of jokes, idioms or industry jargon. They lack the creative flair of humans, and they don’t always translate a phrase or word in a way that sounds natural in the target language. For all of the benefits of machine translation, it has its limits when it comes to certain types of projects.
Before automated translation came along, humans were doing all the work. People knew how to turn a killer piece of French copy into a killer piece in Portuguese, or Russian, or whatever language they needed to translate into. They knew how to make a joke in English just as funny in Greek. A good translator could make a compelling piece of content in one language just as compelling in another.
Humans can only work so fast for so long. If you need a translation overnight, you must either hire one translator who may be forced to sacrifice quality for speed, or a team of translators who deliver quickly but submits an inconsistent translation because each translator uses different wording to translate the exact same phrase. Moreover, the higher quality of human translation will cost more than what you get from a machine.
Lastly, as machine translation tools become more sophisticated, the debate rages over whether machine translation may one day match or even surpass the quality of human translation.
Machine Plus Human Translation
As you can see, there’s a great case to be made for both types of translation. Machines offer convenience, but humans offer linguistic fluency. When you use both, you get the best of both worlds.
Let’s say you have a huge amount of content to translate, like an e-commerce site or a large volume of scholarly research, and you want it in a matter of days. One human couldn’t translate all those words that quickly, but a machine could.
Then again, the machine you utilize might not catch all the specific needs of your content. It might not know how to localize the creative nuances of marketing copy, or it might not understand the terminologies specific to law or medicine. That’s where a human can help. A human translator, particularly one with a specialized background, will know how to finesse marketing content or translate industry terms so they resonate in both languages.
In short, when you employ machine to translate and a human to review and edit, you receive a high-quality translation in less time and at a lower cost. It’s the best of both worlds.
The choice between machine or human translation is a false one. Combining human linguistic expertise with the convenience of technology allows you to speed up translation projects without sacrificing quality.
Launched in Spring 2020, Memsource Translate is a modular aggregator of machine translation engines. Ubiqus NMT is now available on the interface.
Language evolves all the time. Yet, despite the remarkable frequency with which new words seem to spring up, the vast majority of us scarcely notice the shift.
Shall we talk about your project?