How can a computer translate?

Most machine translation (MT) is aimed at generating a first draft translation of text from one language to another. This is achieved by supplying dictionaries and grammars for the source and target languages. Computer scientists call this approach "Knowledge-Based Systems". Where an extensive linguistic database is available for both languages the approach works well.

Unfortunately, most bible translation is aimed at vernacular languages without these resources. In these cases it is difficult to use the typical MT model.

Systems for any language

At BFBS we create systems which can find out linguistic information about languages without the need for prepared databases. Our systems are all designed to be language independent. Starting with nothing more than example text from a language we engineer systems which can learn about the structure of that language for themselves. Computer scientists refer to this approach as "Heuristic Systems" or sometimes "Artificial Intelligence".