A different alphabet for the same field.
At close range, every small fluctuation has a name. Step back and a new account becomes useful: averages, regions, relationships. Coarse-graining discards detail deliberately. It asks which large-scale behavior remains describable after that loss.
Renormalization develops this question through changes of scale. Near certain critical points, very different microscopic systems can share large-scale behavior. Resemblance emerges under conditions; it is not permission to declare every object identical.
Here the cells are averaged in blocks. The instrument is a toy illustration of information loss, not a simulation of a physical universe.
repeat the question, not the conclusion →