Knowledge Representation in Sanskrit and Artificial Intelligence
by Rick Briggs, Roacs, NASA Ames Research Center
(published in AI Magazine, Volume 6, Number 1, Spring 1985)
There have been suggestions to use Sanskrit as a metalanguage for knowledge representation in e.g. machine translation, and other areas of natural language processing because of its highly regular structure (The AI Magazine, Spring, 1985 #39). This is due to Classical Sanskrit being a regularized, prescriptivist form abstracted from the much more irregular and richer Vedic Sanskrit. This levelling of the grammar of Classical Sanskrit occurred during the Brahmana phase, after the language had fallen out of popular use, arguably qualifying Classical Sanskrit as an early engineered language.
— from Wikipedia entry, Sanskrit