Computers and Language Use

Due to Esperanto’s logical grammar, restricted word meanings and limited number of synonyms, it is superior to other languages when it comes to translation by computer. Translations in the EU and the UN could be done according to the following model. A text in say Romanian is translated via computer to Esperanto. Here, the computer needs the help of a person to determine which alternatives of the Romanian words are to be used and in general correct the translation. From Esperanto the computer can then easily translate into any other language.

Only 28 Sounds

Esperanto is superior to any other language when it comes to the ease with which computers can understand the spoken word. This is because all letters in Esperanto are always pronounced the same way, all the letters in a word are pronounced and all words are spelled phonetically. The computer only has to be able to separate 28 sounds in Esperanto. In all national languages there are more sounds than there are letters. Despite having worked on the problem for at least 45 years it has not yet been possible to create a device that can understand ordinary, relaxed, human speech. There are computers that have been programmed to understand the speech of a certain person, but then it is necessary for that person to speak slowly and clearly. Yet mistakes are still made. This is hardly surprising though; a person may know 50 000 different words, but he or she may not be able to say them in 50 000 different ways.

There is no national language that is even close to Esperanto when it comes to the suitability for use in synthetic speech.

© Hans Malv, 2004