Sure, but there are no boundaries between kinds discernable in the data. Perhaps they exist, but we do not have the resolution to make sense of it. It becomes an ink blot test, with different people squinting at the data in the different ways. You can try. The effort ends up undermining the claim in the first place.