There is no doubt that machine translation services are one of the greatest advances of man in recent years. With open automatic translators, it seems that a new world opens up for us: English, German ... Who has never tried to put ‘Hello’


‘Merry Christmas’ and see how it is said in other languages?
Who has not used the Google translator (even if we don't recognize it) to try to understand texts in other languages? It is true that without it we would never have understood that article in Arabic without translation in just a few seconds.

Gradually, thanks to advances in artificial intelligence, translation systems have been sophisticated and increasingly able to give more reliable results in translations. In an increasingly computerized future and in which machines replace jobs, can a machine fully replace translators?

Machine translation

Translation is the process by which a text is understood in a given language and its same meaning is embodied in another language, which generates an equivalent text in another language. Computer translation uses logarithms to perform this same process.

Translation techniques

Although it seems simple, translation is a much more difficult subject than it seems. To perform it, there are different techniques, such as:

Adaptation: the adaptation process uses the replacement of cultural equivalents of the original text with others of the recipient culture. An example of this type of translation would be, for example, replacing rugby with a football match in a translation from English to Spanish.

Linguistic expansion consists of the addition of linguistic elements to be able to resolve the ambiguity of the original text.

Omission or reduction: in this technique, redundant elements are eliminated to produce a more concrete and natural translation.

Compensation: when it is difficult to find a suitable correspondence to the original phrase, this technique is used, through which the expression is modified by a similar one and an attempt is made to preserve the impact of the original phrase.

Calco: it consists in the translation of new terms following the structure of the word in the source language.

Seeing all these techniques, it is easy to ask the question about how a computer translator can apply them through algorithms. Can a machine analyze the cultural references of a text and adapt them to another language and its corresponding culture?

Can you substitute an automatic translator for a human?
The mechanization of translation has great advantages, since the society in which we live continually demands immediacy in the transmission of data and information and the "human" translation times cannot respond to this demand for times.

Currently, translation (view) systems offer good results in Romance languages, such as Galician, Spanish, Italian, Portuguese or French. However, these results are not satisfactory enough to replace the translator's figure. You always need a person who knows the terminology or has access to terminology tools, that fixes the flaws of the translation program and offers a consistent final style.

When a human translator faces a text, he must analyze and interpret all the possible meanings of it in a very thorough and methodical way. In order to do so, the translator must have very high control of the grammar and semantics of the languages between which he is performing a translation. Also, in order to correctly translate a text, the translator must control the phrases made and cultural differences between the speakers of the language to which they want to translate.

When a translator does his job, he often faces complex problems, which can be cataloged into different groups:
Lexical-semantic problems: they are those referred to synonyms, antonyms, neologisms, semantic gaps ...
Grammar problems: referred to the construction of the phrases, which varies depending on the language.

Syntactic Problems

Rhetorical problems: the typical resources of poetry, such as metaphor, simile or oxymoron, are hardly transferable to another language.

Pragmatic problems: expressions, sayings or phrases made are difficult to explain in others languages.
Cultural problems: such as holidays, years, cultural references, meals ... For example, in a translation of a Korean text that includes names of personalities from the country to another language such as Spanish, the translator must document and explain by means of a translator's note who It's those personalities, so that the Spanish reader understands.

To this day, it seems difficult to think that a machine is capable of performing this whole process and facing all these problems correctly. However, advances in artificial intelligence suggest that one day it may happen: what will happen to translation companies?

In short, no matter how much automatic translation is helpful, the human component cannot be lost for now. For that there are professional agencies such as Pangeanic, a world-class Valencian company that has offices in several continents.

In particular, in addition to its parent company in Valencia, the agency has offices in Madrid, London, Boston, Houston, Los Angeles, San Francisco, Miami, Chicago, Hong Kong, Shanghai and Tokyo.