The Problem With Online Automated Translations

woman playing with VR headset

Even as countries imposed border controls this year because of the pandemic, trade and commerce continue to operate beyond borders.

Products in one country could easily be accessed in another, the efficiency of doing online transactions now even better than before the pandemic. It is even possible for an individual to qualify for a loyalty rewards program from a company several oceans away.

The power of this global market, however, has a stumbling block. There is a persistent problem of miscommunications because of some language barriers. English is considered the universal language, but there are several countries that do not speak it.

Currently, other than human translators, applications have been developed to aid basic navigation of websites and fast online communications. However, although fast, machine translations are prone to several errors that could lead to a breakdown of deals and negotiations. Here are some of the problems with most online translation applications:

Words could have several meanings

It is very confusing when a word in one language could mean different things that are not at all related. For example, the word ‘gongjag’ in Korean could mean either a duke or a peacock. Similarly, the English word desert could refer to a barren landscape or ‘to abandon’. Many of us who are still learning a language get the meaning of an unfamiliar word from the context surrounding the word. But if the context is also confusing because of an unfamiliar syntax, it’s dangerous to assume meaning.

baggage sign

There’s a need for specialized keyboards

It is difficult to simply key in languages that use a different alphabet system unless you have the characters on your keyboard. If the document you need to translate could not be copy pasted, it is impossible to use most of the available translation apps. Word Lens had provided a revolutionary app that allowed users to point their cameras at the foreign alphabet and the system translates based on the image. But in this case, the translation had been per word. It would be difficult to interpret what you are reading if you don’t know the syntax or sentence construction of that language.

Google has also improved its translation services with sound and image translations. However, sound detection is difficult to use in a real-time meeting since when speakers talk fast, the program couldn’t capture the sentences fast enough.

Gender nuances are not always picked up

Gender nuance may not be that important in these quick translations, but when a paragraph is discussing several people, and only pronouns are used it could get confusing. That’s when translating from a foreign language to English. But when translating the other way around, gendered nouns are not always distinguished. It could be understandable, but could also be confusing to the language speakers who are used to the differentiation in the nouns.

Machine translation does not translate the mood and emotions

Of course, machine translations could not be used if you’re planning to read literary works. The beauty of the narrative would not go through in the jumble of words and headache-inducing grammar outputs. The writer’s choice of words to communicate a feeling or an ambiance in the story would not be nuanced by these automated translations.

The future of programmed translations

There is still too much potential in the development of translating devices and programs. The speed at which the programs and apps could capture sounds, images, and transform them into words arranged in a comprehensible order would be important. Grammar, learned for years by humans, is another challenge for programmers. Although readers could understand ungrammatical sentences, very bad grammar could also lead to misinterpretations.

Programs also need to be updated regularly because language is very dynamic. In many languages, there are several ways to construct a sentence. Elaborate sentences could be shortened. Words also gain new meaning, are used differently depending on the developments in society. For example, the word ‘capacitation’ has for a long time been unaccepted as a noun but has become part of the jargon of non-profit organizations. Countries that speak the same language may also have a different understanding or use of a word. Some programs could already differentiate British, American, or Australian English. This could be done for other languages, like widely-used Spanish.

As we are integrating further, interests not only in businesses but also in cultures and traditions expand. Currently, we rely on automated translations for quick references. This could, however, be the extent of the service of these programs. The most they could offer is a basic understanding of what’s being stated.

Communication as a whole is a nuanced field. Even speakers of the same language could misunderstand each other when they have different contexts. Mediated communication using the same language is also vulnerable to misunderstandings because of missing context clues like body language, tone, and facial expressions. In the end, interpersonal communication still needs the skills of human translators who could convey the subtle emotions and deeper meanings of sentences.

Share this on

Share on facebook
Share on twitter
Share on linkedin
Share on pinterest
Share on tumblr
Share on reddit

You might like

Scroll to Top