I want to pluck out of the previous post on the Dravidian language and the timing of ANI-ASI population genetic admixture in Indian using LD methods (based on an article that is also widely discussed elsewhere) to focus on one key conclusions:
While the Dravidian languages likely have a source that pre-dates Indo-Aryan (i.e. Sanskrit derived) linguistic expansion in India, all of the modern varieties of that language likely derived from an expansion that followed the first wave of the Indo-Aryan invasion of India (at least linguistically). Essentially all other non-Austro-Asiatic, non-Tibeto-Burman languages of India were wiped out at that time.
This lack of basal diversity in the Dravidian languages is one of the factors that makes this language family hard to assign to any macro-family of languages. The Dravidian language family was basically one proto-language around the time of the golden age of the classical Greeks.
Of course, it also doesn't help that the Harappan language is lost and that there are no attestations, even snippets of phrases from any languages of the Deccan Pennisula prior to about the 7th century BCE. All of the intermediate steps that gave rise from something else to Dravidian have been erased from the annals of history.