This is one of the main question of certain strands of linguistic typology. I don't think there is a definitive list of impossible and obligatory coocurrences. The closest is probably the Universals Archive, which lists 32 absolute, mutually implicational achronic universals here. Note that these are ultimately hypotheses to be falsified.teotlxixtli wrote: ↑18 Aug 2022 03:01 What kind of linguistic features always co-occur? Which ones are mutually exclusive? Although I’m sure it exists I can’t think of a language off the top of my head that has both tone and grammatical case (for example)
As for tone and case, this is really just a rough tendency. Some Nilotic languages, like Kalenjin and Turkana, even mark case through grammatical tone. The same can probably be said about a lot of Grassfields Bantu languages that mark at least some adnominal possessors tonally. Similarly, other languages of West Africa, like Kanuri, often have case as well as tone. I suspect that there are also some undocumented cases in the Amazonas area. Koasati is a counterexample in Non-South America but there are probably more, especially if you also consider complex pitch accent systems and the ongoing documentation going on for many tone languages in Mesoamerica. Kewa is an example from Papua and again there are probably more due to the sheer diversity and ongoing documentation.