Machine translation [Archives:2004/730/Education]

April 19 2004

By Dr. Bashar Ghazi Askar
[email protected]

Automatic translation is a new Internet approach in the field of translation. It is a very attractive idea at a time when the world is witnessing technological and scientific communication boom in the information technology system, with its phenomenal acceleration. There has been great need for immediate transfer of scientific and technological information from one language to another. The first attempts were to feed the computer with matching lists of words in the two languages. Of course, this did not produce language. Languages differ in their syntactic and semantic structures. Translation requires programmed texts with grammar and lexicons of two languages, and an inter-lingual grammar representing equivalences and differences between them. Human intervention and touch is very important in the translation processing. For example, after finding sentence equivalents between the two languages, the computer will match the morphemes of the input sentence with those of the output sentence. In fact, machine translation faces serious problems relevant to word choice for it happens to choose the wrong equivalent in the wrong context. Machine translation might be good in simple texts. Complex texts, particularly those of high cultural connotations found in religious, historical, literary works, still require human intervention.
Translation is the replacement of a representation of a text in one language by a representation of an equivalent text in a second language. Linguistics tells us that total equivalence between any two languages is not possible. Languages differ in form, structure, grammar, and build up. The structures of a language have different meanings. There is no total congruence between the lexical repertoires of any two languages.