Online Translation English Portuguese

Automatic translation systems first appeared in the 1940s.

At the time, it was believed that they could eventually substitute human translation, that with the growing complexity of computer systems, automatic translation could also reach a degree of perfection capable of rivalling translations made by humans. It was thought that language could be deciphered by computer logic circuits organised into a form of artificial intelligence that could make the link between language and context required for a good translation. However, language has special features and indeterminacies and it became obvious that the increasing numbers of problems discovered in machine translation work could not easily be resolved.

The current state of technology is still a long way from perfection. We do not know what the future has in store, as we see spaceships piloted by computer and an increasing number of activities that are entirely controlled by artificial intelligence systems. There are machines which can play chess and beat human champions, to quote a simple example.

Let us look at an excerpt from The Times on the Cannes film festival last year, translated by Power Translator 11, a professional programme, and Google Translator.

Original Text:

The apocalypse came early to the Cannes Film Festival this year, filling screening rooms with snarling dogs, bursting bombs, shouting men and screaming women.

Power Translator 11:

O apocalipse veio cedo ao Cannes Filme Festival este ano, enquanto enchendo quartos de blindagem de cachorros rosnando, bombas congestionantes, gritando os homens e as mulheres agudas.

Google Translator:

O apocalipse chegou cedo para o Festival de Cinema de Cannes deste ano, enchendo rastreio quartos com snarling cães, explodindo bombas, gritos homens e mulheres gritando.

Notice how Google Translator has vocabulary problems, not recognising one of the words because it is not in its internal dictionary. The sentence makes no sense to someone who does not have the original.

The professional programme produces results that are not much better. The problems of context are the most difficult to solve, although lexical issues are also raised and interfere with a proper understanding of the sentence.

If automatic translation systems have not developed as far as we had hoped, or they continue to present the same problems, how can they be of use?

Since it will be difficult for a machine to be able to understand alternative meanings of a word according to the context in which it is used, and it will never be able to hold enough information (not due to limitations of space) to be able to complete translations in all fields, the real use of these systems will be to convey some information to people who are not familiar with the source language. They should be able to get a very general idea of the subject of the text by processing it through an automatic translator.

For us, translators of a given language, automatic translators will not be of great use on a professional level. They can serve as little more than a dictionary, and even in this role, they present the limitation of offering only a single translation for each word.

Translation, that complex bridge between two different language systems, will continue to be an activity reserved for the human brain.