Humanoid Translation (Machine-assisted Human Translation)

The goal of several linguistic AI programs and developers has consciously or unconsciously been to create a world where everyone has access to the other via language, despite language, and by language. Meaning to say, that people are able to communicate with each other instantaneously. Translation and interpretation software and instant voice recognition software are making things easier.

With each passing day, artificial intelligence engines are getting smarter, better and more efficient, but are not yet perfect. Meaning is still lost in translation. Even when human beings translate, meaning is at risk of being lost. Meaning has several aspects. Besides the semantic, there is also the syntactical, functional and idiomatic sense that is often lost for want of cultural, contextual and subtle distinctive knowledge. What can one then expect of a machine that is most likely based on lexical replacement.

For centuries, how to translate has been an point of theorisation & discussion. It was clear among cross cultural communication-facilitators that translation is the transference of the meaning and sense of words, in both their denotative as well as connotative senses, in the literal as well as figurative senses, and not merely the replacement of words.

In the earlier years, automatic translation was based on?the principle of mechanical replacement of words, often without as much as a syntactical reordering of words. Clearly, such developers have even overlooked the morpho-syntactical variations of languages, how some languages inflect words unlike English or others.

The problem lie in that developers of automated translation software were not linguists and linguists were not developers!!! This disconnect initially made automated translation an exercise in cryptography and the reader invariably became the de facto descriptor. The problem of communication was yet not solved! And is still in the process of being solved.?

And with success! There is no denying that progress has been made. In leaps and bounds. Linguists have joined the bandwagon and work in tandem with developers. Linguistics and translation methodology are becoming part of the repertoire of skills required to be developers for translation software. Interdisciplinary knowledge transfer and collaboration have revolutionised the way in which machines are translating today. Machine translation combined with cumulative and limitless translation memory have facilitated an unprecedented artificial intelligence in the field.

Although we seem nowhere near an infallible technology, the advancements are impressive, and have and will continue to have far reaching implications.

Machine assisted human translation is helpful in increasing productivity and speed. Translators who earlier translated about 3,000 to 4,000 words a day are now able to double their output effortlessly. AI assists in day-to-day interpretation of important texts. For doctors, lawyers, bankers and simply anyone who needs a quick, preliminary gist of a text, this is a boon! Despite all its limitations, machine translation at least provides a basis foundation for subsequent work and possible improvement.

The world’s consumption of translation services has multiplied several times over in the last few decades owing to the expansion of the world-wide-net and knowledge based economies. The global outlook and outreach of the world, the integration of nations under the ambit of glocalization has led to unprecedented requirement for translation and cultural exchange. Machine translation has helped keep up with this increasing demand.

Machines can aid in accuracy and enhance quality of translations. Today’s machines provide high-quality first-draft, which translators go on to edit. Machine generated translations reduce the burden of repetitions within texts. They provide a high level of automated consistency. This makes the job of a translator easier and the output better. Machine translation has therefore graduated to the category of a word processor and translation tool.

Of course, artificial intelligence has yet not replaced the human brain and perhaps never will, considering the complexity of the human brain and linguistic competence. But surely, it has proven its mettle over the last couple of decades and challenged the views of naysayers. It is here to stay. It is here to grow. It is here to assist in creating a world without linguistic boundaries.

Originally published here: