Who was the first person who caused the division of Sanskrit versus Dravidian? When William Carey, Thomas Colebrooke, Charles Willies believe that every Indian language has derived from Sanskrit (1700s),
there was a person called A.C. Campbell who was based in Telugu regions. He wrote a book called Telugu grammar (1816). His primary aim was to disprove these guys, to show that No, there is one more language family in India and it is not all from Sanskrit.
His mentor was somebody called Francis Whyte Ellis he he wrote a 40 page foreword to this Tamil Telugu grammar and that is come to be called as Dravidian a Proof. He compared a Sanskrit word list to Kannada, Tamil and Telugu with a prefix suffix pattern and what meter is used and he proposed that there is something called a Dravidian Language. His works were inherited by two missionaries, one is G. U. Pope and the other Robert Caldwell. And ever since it became the common way of looking at Southern Indian languages.
Michael Witzel (1999) says that there are no Dravidian loan words in oldest Vedic layers, they appear later, only in words of commercial transactions.
There is more Sanskrit in Dravidian languages than the other way around. M B Emenrau in 1956 said that Dravidian loan words ceased towards common era implying exit of Dravidian from Ganga valley. This is debated by K. Elst (2003).
In India today the academia uses something called linguistic area, where many languages co-exist. They use Bayesian Phylogentics in order to see the roots of so called Proto-Dravidian. How are all these Tulu, Malayalam, Kannada, Telugu related to each other and where is their ancestor. And they come to 2500 BCE is where the ancestral Proto so called Dravidian languages. However, the criticism to levy against this is that why would not they also investigate relationships with existing Indian languages, Sanskrit and other such languages. So if they embedded the relationship and looked for a common root we would have found different things. So they have already used Ellis Whyte’s idea that there is a Dravidian language and other North Indian languages and use this as a constraint to build these mathematical models and use that in a circular logic to try and prove whatever they want to prove. This is the fallacy of these studies.