It's inferable from the fact that the writer saw both chains and produced a new node that merges these heads. So that writer "resolves" the conflict - according to whatever strategy it is programmed to use. (It might be as simple as just storing a JSON that says {"conflict": true, "values": [4, 6]} , and the user will have to pick.)
If it's possible to model operations in a commutative way (eg. instead of assigning values to keys one just stores differences), then the conflict resolution is mathematically proven, just apply all operations in whatever order, they're commutative, great. Of course it doesn't help with "real world data", but that's where mathematically we can use and oracle (the user, or whatever linearizer service we choose).
If it's possible to model operations in a commutative way (eg. instead of assigning values to keys one just stores differences), then the conflict resolution is mathematically proven, just apply all operations in whatever order, they're commutative, great. Of course it doesn't help with "real world data", but that's where mathematically we can use and oracle (the user, or whatever linearizer service we choose).