Earned vs Borrowed Compression

The most consequential distinction Miller's framework draws is between compression that was built through effortful recoding and compression that was received as a finished product. Both kinds of chunks occupy the same amount of space — one slot in working memory. Both enable the holder to function effectively under routine conditions. But they are not the same thing. An earned chunk preserves the structural knowledge of its contents. The person who built it can decompose it when necessary, reach inside, understand why it works, and reconstruct it for novel conditions. A borrowed chunk carries the surface representation without the internal structure. It is a label occupying a slot that looks, from the outside, exactly like the corresponding chunk in an experienced practitioner's mind. The difference is invisible under normal operations and catastrophic under abnormal ones. When a system built with genuine chunks encounters unexpected failure, the developer can decompose the relevant chunk, inspect its sub-components, identify the inconsistency, and repair it. This is what