I wonder how his system of analysis of the difference between a complex system and a machine would distinguish between a lit candle and an unlit one? (Not trying to be snarky; I'm really curious about that.)
If you disassemble a candle, you can presumably reassemble it, and it remains unchanged.
However, if you disassemble a lit candle, you may "extinguish" an ineffable property of the candle that cannot be reassembled readily.
I don't think it could, because neither candle, lit or unlit, would be a complex system within Rosen's meaning. In general, inorganic systems are not complex systems.
Just thinking through what you wrote, how does one "disassemble" a candle? I suppose one could melt it, or smash it up with a hammer. But to "reassemble" it would not reconstitute the original candle. The only thing about that candle (or any candle) that doesn't change is the "form" or idea of "candle." But coming straight from Plato, "form" (i.e., formal cause) is usually regarded as an unscientific notion of the sort Francis Bacon wanted to banish from science.
Also it seems unscientific to ascribe "ineffable properties" to a candle; i.e., its quality of "litness." Its "litness" has nothing to do with the internal properties of a candle. I.e., it doesn't itself produce the fire. That has to be added to it, from outside the candle system.
In short, I'm not sure what your question is, NicknamedBob. Maybe some of the above ideas might help in refining/reformulating it?
Thank you so much for writing!