Sanskrit to English Machine Translation

This work is a the result of years of hard toiling research by the guru - sisya duo. Comprehending the Sastras in god'd own language in difficult and more so in the case of computer. The book is the fruit of the journey of the scholar's doctoral research. We hope that the book will be helpful for U.G., P.G. students and research scholars alike.

Dr. Chitresh Soni

Born and brought up in Rajasthan, Dr. Chitresh completed his graduation and Philosophy from the Univesity College of Social Sciences and Humanities, Udaipur. He did his Masters in Sanskrit from the prestigious School of Sanskrit and Indic studies, J.N. U., New Delhi. After that he got the U.G.C. JRF to persue his research. He wrote his Ph.D. under the able guidance of Prof. Girish Nath Jha and Prof. Naveen Kumar Jha.

Human language is very interesting. The faculty of mind has infinite potential to create sentences. The revered Sanskrit philosopher of language, Bhartrhari, has judged the meaning to be one in mind though it manifests itself in many forms. This meaning in mind is the root of all varieties of behavioral language." Although the meaning stands to be one butit is grasped through expression in vocal language. The meaning is illuminated through the Sphota in mind. The Sphota concept is like universal or linguistic type, as opposed to their tokens." Sanskrit grammarians and philosophers of language alike accepted and configured the principles of the Sphota doctrine ina metaphysical way. They view Sphota as the language capability of a person, revealing his consciousness. This ultimate state of reality, which resides in a person’s mind, is stated as Sabda-brahma or Eternal-Verbum.

This notion of Sphota described by Patanjali is further discussed by Bhartrhari at three levels

Varna-sphota — At the syllable level

pada-sphota — At the word level

vakya-sphota — At the sentence level

A distinction is made by Bhartrhari between sphota, Which is one and indivisible and nada, which is divisible and sequenced. Sphota is the root cause or the intention of the speaker. It also illuminates in the listener, the mental state as of the speaker. This theory of sphota and nada are very relevant to a researcher in MT. It signifies the concept of unified language discussed by ancient grammarians. Till this date Artificial Intelligence’s essential aim is to make computers intelligent so that they can first decode the meaning of a sentence in the SL and encode it in a sentence in the TL. It involves a complex process to understand and internalize the lexical-semantic and morpho-syntactic information and make use of it. This process is useful to understand the Principles that make HCII possible. The zest of the matter is that Artificial Intelligence aims to understand and then implements intelligence using the functions and methods of computation, thus offering a radically new and different paradigm for theory formation. Natural Language Processing is one of the areas of Artificial Intelligence, where attempts are being made for computer to understand and generate human language, either in one language or from one language into another.

Initially MT was considered to be an easy way for deciphering SL and converting it into TL. People thought that if they were successful in mathematical or computer representation of the syntactic structures of the sentences of a natural language, they could quickly convert it to another language using compiled dictionaries. Exact translation would follow up by replacing words in the dictionaries with their corresponding synonyms. After this they could rearrange and reorder the words according to the syntactic structure of the TL. But this work which seemed to be very simple proved to be challenging. It is a particularly difficult task in NLP at present. Both handling the context of the sentences and the mathematical representation of the grammatical structure was considered to be thorny. As a result, the MT program, in the 1960s, went to a sabbatical. Today MT systems focus on post editing or MAHT formulae. Here the role of knowledge representation comes into sight. In many systems knowledge representation techniques such as semantic networks, first order logic etc. are used. In any MT system, issues will always arrive. The best way to find issues and challenges using an MT methodology in a particular language pair is to actually R&D a system. This is called constructive approach in the area of MT. Daksiputra Panini also composed his genius work, the Astadhyayi using the constructive approach. Panini’s Astadhyayi is termed as "the greatest monument of human intelligence" by linguist Leonard Bloomfield.

**Contents and Sample Pages**