I think what you're looking for is called bitemporality (https://en.wikipedia.org/wiki/Bitemporal_modeling https://en.wikipedia.org/wiki/Bitemporal_modeling).

I think what you're looking for is called bitemporality (https://en.wikipedia.org/wiki/Bitemporal_modeling https://en.wikipedia.org/wiki/Bitemporal_modeling). It allows you to query a system (some databases support this) in a temporal context. This is useful in financial services where you have market data baked into a price you've offered to a client, where the underlying market data might have been later corrected by a vendor. So you can query how the world looked last tuesday 4pm; but also - given what we now know, how ought the world have looked last tuesday 4pm. Interesting stuff.