- Posted by Stephen Whiteley
- On 29/08/2019
The Risks of Automatic Translation
Advancements in technology have always been a blessing, but what are the risks of automatic translation? From AI running our blogs to website translation using machine language, it is just so amazing. But, are we not becoming overdependent on this bots? They are machines, after all, they only do as they are told Right?. Machine language made it such that almost every internet user can access most contents in various languages, the Google translate feature. But, it is worth noting that this feature comes with its risks.
A real example
Take this as an example, you visit a french site, and thankfully, Google Translate was on deck to help you translate every word and sentence on the site. On navigating through the site, you noticed that the words “nuts” and “nuts and bolts” have two separate words in french as opposed to the singular meaning of both in English. This kind of issue can be detrimental to the website’s success, imagine confusing someone with a nut allergy, that could be pretty serious. This could look funny on the outside, but it is pulling your potential visitors away from your content, not just the English-speaking visitors. Now the questions are, how efficient are the machine translators? and can their effectiveness outweigh the use of human translation? The answer to these questions is evident in the fact that machines can never fully replace humans.
Speed or accuracy?
In terms of speed and accuracy, a human translator will cover more critical areas of a website as opposed to a machine translator. Here some other apparent reasons:
- It is not possible to program a machine to understand culture, considering how vast our lifestyles are. Different culture has individual lexical items specific to them. Machines cannot deal with the complexities that come with understanding slangs, idioms, and names. A native in-country speaker of such a language will understand the full meaning of such slangs and will be able to provide its equivalent in other target languages. Different communities have a different accent, which means their pronunciation is different even if the words mean the same thing. Only a human translator can tell the difference because the machine translator will do a direct word to word translation.
- As mentioned earlier, machines cannot relate words to context. There are words with dual meanings; translating such words will be a problem for a machine translator.
- Another issue is localizing machines for different languages. New phrases are developed every day; it will be challenging to program a machine translator to implement such changes when they come up. The machine will have to be continuously programmed to adjust to these changes and considering the algorithms that make up these programs, and it could take forever. In another hand, a human translator will recognize and implement these changes hassle-free. Consider a situation where there is no suitable translation for a particular phrase, what happens then?
A machine language cannot understand the flexibility between style and tone. Every content has a unique style. In translation, it is difficult for machines to understand these essential features.
Conclusion – risks of automatic translation
A translation can never be complete without a human touch. All the risks mentioned above should not shy us away from using machine translators but serves to help us effectively combine machine translation and human translation. This gives our contents the clarity they deserve. If we can combine both effectively, then this risks can be reduced to be a minimum.