Rising intonation with in questions is so widespread that there is no historical evidence regarding pre-modern English. The best that we can do regarding explaining the observation is (1) to say more precisely what seems to be true about human language and (2) to conjecture about functional explanations. It seems to be true that polar questions are more associated with a conventionalized intonation pattern with distinctive pitch rise towards the end, than with distinctive pitch fall (although more accurately, there is most often a longer rise and a short final fall). When declarative and interrogative sentences are distinguished intonationally, languages most often do so with rising intonation for questions and some other pattern for declaratives.
Why then would it not be the case that declaratives generally have rising intonation and interrogatives have falling intonation. I should also point out that rising and falling intonations are not exactly symmetrical, in that rising intonation typically reaches a higher pitch that falling intonation starts from. As is the case with many "why" questions in linguistics, the answer requires seeking functional the functional underpinnings that can lead to language evolution. There is a social (pragmatic) difference between questions and declarations, that a question is a request made of another person, a request for information. A request is generally understood to be subject to the willingness of the requestee, though of course you can demand an answer, that is not a typical question.
People tend to react positively to small things, such as children, puppies and kitties. Small things generally have higher pitch in whatever acoustic outputs they produce (for direct physical reasons). A higher pitch is therefore more likely to engender a positive response than a negative response (tendencies, not an absolute law of human behavior). Higher pitch in questions is therefore more likely to generative a positive reaction and compliance with the request than low pitch is.