Pāṇini: Catching the Ocean in a Cow’s Hoofprint

A PDF version of this overly long post is available here .


This is a companion discussion topic for the original entry at http://blog.granthika.co/panini/

Overlap of concerns with semantic web, NLP, ML and knowledge-base engineering is remarkable. I take it Panini was motivated to get his Sanskript expressive model right (semantically tight?) so humanity could be right/tight with the divine; we’re merely motivated to get it right for a better Internet… and profitability, of course. So different from Biblical and other sacred texts.

Pāṇini’s text is remarkably free of any divine purpose or any religious intent, so it’s hard to tell with any degree of certainty what his intentions where. The first chapter doesn’t have the usual prefatory verse or two exalting a divinity or thanking one’s gurus. Pāṇini just starts straight in with the rules: 1.1.1 is something like, “The term vṛddhi is introduced as a name (saṃñjā)…”

Later grammarians and philosophers argued that Pāṇini’s vision supported the notion of śabda-brahman, as in the Vedic sources. But the man himself is remarkably non-prescriptive, and a main reason for his algorithm being so complex is that he wants to allow for all kinds of local usages, corner cases, etc. If he wanted to create a completely artificial language, he could’ve made his rule-set much more simple. It’s an argument that continues, but I’m on the side that supports the notion that he was doing descriptive linguistics.

Nice article. Here’s some other material that is related to this fascinating topic:

Agnicayana and fire altars: https://arxiv.org/pdf/0902.4850.pdf

Panini and Sanskrit: http://www.ece.lsu.edu/kak/bhate.pdf

@subhashkak1: I’m glad you liked the article. And thanks for the links–I’ll read.

And, are you the Subhash Kak who wrote Computation in Ancient India and other books? If so, I’m a big fan. Pleasure to meet you.

Thanks.Yes, I’m the guy who has written several other books. Glad to meet you also.