Adding a temporal dimension is a great idea. It would offer a fascinating perspective on the data by allowing us to track the 'semantic drift' of legal frameworks over time.

Brazil is actually a perfect use case for this: the country has had 7 different constitutions since its imperial era (https://en.wikipedia.org/wiki/History_of_the_Constitution_of...). Plotting its transitions through a monarchy, dictatorial periods, and modern democracy—alongside historical texts like the French Republics or the U.S.S.R.—would let us visualize exactly how a nation's priorities shift within the latent space across different political eras.

Thank you for the suggestion! I will seriously consider advancing the project in this direction.