As everyone in this discussion has already noticed, linguistic taxonomy can be an extremely contentious issue, and a lot of language families tend to have multiple organizational schemes depending on which scholar you talk to. Now I don’t have background in Indian linguistics, but off the top of my head I seem to recall that Hindi and Urdu are usually considered to be a diasystem: essentially they started off as a single language, but due to cultural differences and a growing set of dialect-specific vocabulary, it eventually became the case that the language was partitioned on cultural borders. One of the best-known examples from recent times is probably Bosnian: it’s my understanding that fifty years ago, there was a single language called Serbo-Croatian, but today, for various reasons, nobody speaks Serbo-Croatian: they speak Bosnian, Serbian, or Croatian. Despite being mutually intelligible (in general), the speaker population divides itself into distinct languages.
Since it’s my understanding that diasystems are often created and sustained by schisms that stem from a single speaker population that self-divides itself along cultural lines, each dialect (or language) typically begins to pick up an increasing amount of unique vocabulary. In the case of Hindi and Urdu, I’m looking at Ethnologue, which is usually a pretty reliable source for linguistic taxonomy, and it indicates that they are two registers of a single umbrella language called Khariboli. (Khariboli also encompasses Dakhini and Rekhta, but this is my first time hearing of either register.) Apparently literary Hindi tends to feature unique vocabulary drawn largely from Persian and Arabic, while literary Urdu favors Sanskrit terms.
ED: And I see **John Mace **already covered this. :smack: My mistake.