Your comment makes me think about statistical mechanics and microstates. That is to say ... in a complex system with properties that are a function of microstates, whether the internal structure of the microstates that correspond to a given property matter can depend on your point of view or interest in the system.
Heat, for example, is a statistical property of a system, and a given temperature can correspond to a vast number of possible microstates of the system. For some purposes, you care precisely which microstate the system is in; for others, you do not, and the temperature property is entirely adequate to describe the system.
Rules may describe the microstate, but may be (depending on your POV) be irrelevant to the property.
Using Wolfram's model of the world, there may indeed be a cellular automata following rules that underlies the property, but there may be no reason to care about it in a given instance; instead you're interested in the "evolution of states" (i.e. values of the property).
Some complexity scientists are quite taken with this idea of not needing to care about the lower levels of a system when consider higher level behavior. In their view (and rightly so, IMO [0]) you don't always need to consider the rules that drive (say) physics when considering (say) psychology.
[0] except that I think that Hofstadter's "heterarchy" idea is likely to be even more accurate - interesting systems are the ones in which there are complex feedback systems between different levels of the system.
It seems pretty clear to me that this desire for "perfect" layers of abstraction is something we strive for due to our own intellectual limits, and that in reality all abstractions are lossy to some degree. Heat as a single integer in Degrees F is good enough most of the time but when you're designing CVD for a Silicon Fab you might actually care about the positions and orientations and vectors of the gas molecules.
Heat, for example, is a statistical property of a system, and a given temperature can correspond to a vast number of possible microstates of the system. For some purposes, you care precisely which microstate the system is in; for others, you do not, and the temperature property is entirely adequate to describe the system.
Rules may describe the microstate, but may be (depending on your POV) be irrelevant to the property.
Using Wolfram's model of the world, there may indeed be a cellular automata following rules that underlies the property, but there may be no reason to care about it in a given instance; instead you're interested in the "evolution of states" (i.e. values of the property).
Some complexity scientists are quite taken with this idea of not needing to care about the lower levels of a system when consider higher level behavior. In their view (and rightly so, IMO [0]) you don't always need to consider the rules that drive (say) physics when considering (say) psychology.
[0] except that I think that Hofstadter's "heterarchy" idea is likely to be even more accurate - interesting systems are the ones in which there are complex feedback systems between different levels of the system.