Apparently the only necessary condition for two observers who are "independent" so they communicate two events, is that they perceive the same causal structure and the same kind of events generally expected to have a lifetime in it. So that they must also have a protocol or means of communicating with each other, so they commute the message each sends and "understand" it, there must be not only the same kind of dynamic is part of the same structure but a way to talk to each other about it. So encoding is then any language that recognizes and conveys events between observers, for which we generally assume a minimum of two - although you need three observations to triangulate over a sphere. However observers o and o' need the same three things for communications to occur, which is G,A the structure with algebraic states (messages, measures) in it, and the third is "will" or the encoding and transmission. We wrap information in packages and distribute it, because there is no other way to do it. So ultimately, we slice space with time, or we "time" spaces; alternating with we slice time to distribute it over space. Observers can assume they see independent events and (all they can do is) communicate. Kant, Spinoza, in fact said much the same thing, we can only look and agree, and so learn or attribute. We are "channels, with memory". There is no other result for the question "why" than "because it's all we can do", and for "how" than "because we took it apart and looked". Godel used Boole's product: you cannot prove propositionally that a proposition is universally true or false, against you cannot prove a lie is false, to show that the universe is Godel-finite, or c-complete I think he termed it. And we can't halt because we can't prove that we can.