In 1966, G. Devaneyan, also known as Pāvāṇar, published a book entitled The Primary Classical Language of the World..  To quote Wikipedia, "In his 1966 Primary Classical language of the World, he argues that the Tamil language is the 'most natural' (iyal-moḻi) and also a proto-world language, being the oldest (tol-moḻi) language of the world, from which all other major languages of the world are derived. He believed that its literature, later called Sangam literature and usually considered to have been written from 200 BCE and 300 CE, spanned a huge period from 10,000 to 5,500 BCE.”  His method was largely based on his claimed etymologies, which he derived not by the usual methods of comparative linguistics but rather by similarities he saw in various roots and what he claimed were their cognates.  He writes, "One of the fundamental principles of Descriptive Linguistics is, that all languages are arbitrary. This is disproved by the fact that Tamil, the earliest cultivated language of the world, contains no arbitrary word, and is still traceable to its very origin.”  Thus he carefully goes over the Tamil lexicon (including words that comparative linguists would say are of Sanskrit origin like paṭimam < pratimā) and finds what he considers roots of other words.  He uses this to construct an edifice that conforms to his notions of Tamil as a perfect language that is the source of all others.  He believed Tamil was the language of Lemuria, long ago submerged in the sea, and that South India was once part of that huge continent.

His ideas have been quite influential in Tamil Nadu — they were even thought to schoolchildren (and may still be).  Some academics in TN have been deeply influenced by his writing and thoughts.  I once heard Karunanidhi, the previous CM of Tamil Nadu, give a disquisition in front of a large audience in which he advocated Pāvāṇar's ideas and claimed Tamils did indeed come from Lemuria.  Recently on cTamil, a forum on classical Tamil, there was a discussion in which proponents of comparative linguistics debated those who follow Pāvāṇar.  The results were that no one won — it was agreed that the two methods are quite different and that there is no possibility of productive dialogue.

When it comes to Dravidian etymologies, we have fewer resources than for IE, but we still have enough to find many words in the family which are related—the exemplary work of Emeneau and Burrow in the DED collects these.  Emeneau himself said that every etymology is an act of faith, but clearly those in the DED are mostly beyond reproach or question.  As a scholar, I cannot find much useful in Pāvāṇar’s ideas, but many who post on forums such as Indology use Pāvāṇar’s methods to support various theories about the great age of Tamil, the relationship of Tamil with the IV Civilization (still very much unproved, in my opinion), classical Tamil social structure, and the like.  Some of these ideas may have some validity (we certainly cannot disprove that the IE Civ. was Dravidian-speaking), but the etymological evidence cited is rarely more than speculation, in my opinion.

We are fortunate to possess in Tamil an enormous literature (Sangam Literature) that is about 2000 years old and that is not much influenced by Sanskrit — though institutions from North India like Brahmins, Jainism and Buddhism were present in TN at the time of its composition.  This fine literature gives insight into village and city life with a precision and richness that we do not possess in other Indic languages.  We can use the Sangam texts to discover many things about Tamil Nadu, and by extension the Deccan, in the early centuries CE, but it is not much use, in my view, for understanding the ancient past of the north, of the Indus Valley or of history long before the common era began.  It can also be used, I have argued, to help understand features of the Dravidian-speaking culture upon which heterodox, Sanskritic and Brahmanical elements came to be superimposed. George Hart