Languages are interesting objects of study not only to linguists, but, to judge from history, also to mathematicians and physicists, probably because their study requires the same kind of rigor and presents the same degree of testability encountered in the latter two disciplines.

Moreover, there are some decidely mysterious elements to language! The power of speech is instinctive- almost unnervingly so! In this sense, it acts as a bridge between the wakeful and the subconscious states of our mind. This psychological aspect makes it even more interesting, doesn't it?!

Natural langauges

Since my childhood, learning new languages has been a constant hobby for me. Over the years I got to learn English, Hindi, Tamil, French, Sanskrit, German, Russian and Finnish. As a corrolory, I can "decode" (with the odd help from a lexicon) texts in Norwegian, Danish, Dutch, Italian and Spanish. And my advice to any one looking for an exotic language: check out Suomea- I mean, Finnish! :-)

I learnt English as part of my regular school requirements. French was the first, and worthy, object of my spontaneous linguistic interests. In this respect, I was really fortunate to have as my tutor the renowned scholar Dr. Sambhasivam.

This started me on a linguistic search that took me to learning German, for which I lucky to find some good "Teach Yourself" books. I started off Russian on my own, but my crude spoken accent was exchanged for the standard Moskowian accent thanks to my excellent teacher Dr. Nadezhda Krashenkina.

Constructing languages

All along, I got into the habit of making up my own constructed languages (or conlangs for short), each one usually reflecting the natural language that was currently holding my attention. In case you think I am weird, only check out here for more information on a good many conlangs by other creators. :-)

My earlier conlanging attempts were mere relexifications of natural languages, gradually giving way to more independent and creative ways to make up a language. I have linked herebelow information to two of my on-going conlang projects:

1. Lin: This is an experiment in creating a very concise language (in terms of computer memory rather than semantics). As a side-effect, it violates a number of "linguistic universals".
2. Hambhukringki. This is a psycholinguistically oriented language, which, if naturally evolved, would have required the speaker's worldview to be quite different from terran worldview!

Information theory, mathematics and semantics

According to Indian tradition, speech (vak) is a many-layered phenomenon spanning various degrees of mental awareness. At its subtlest, it is rooted deep in the subconscious, where it is called para. At its grossest, it is verbal oral manifestation, where it is called vaikhari. The intermediary states in decreasing order of subtlety are pashyanti and madhyama.

Is it possible that mathematical modeling, especially using concepts from group theory, topology, gauge theory and graph theory shed some light on the semantic and psycholoinguistic aspect of language? I have some very incipient and inchoate ideas on this. It's an on-going project for me!

A more limited undertaking, but in which I have been able to do more work, is studying semantics from an information theoretic viewpoint. Here I have argued that to quantify semantic information in a language-independent way, Shannon information is inadequate. A new quantifiication of semantic information, based on redundancies at a part-of-speech rather orthographic level, is explored.

My latest thoughts on these topics can be found here.

Comments are welcome! .