Rank-size laws and heavy tails

A handful of huge cities, a handful of works copied thousands of times, a handful of hoards worth a fortune -- and then a long tail of small ones. This page gathers conjectures that treat that lopsided shape itself as measurable evidence, not just an observation about inequality. Three well-known statistical patterns -- Zipf's law from linguistics, the Pareto distribution from economics, and Gibrat's model of proportionate growth -- all say that the exact steepness of that long tail encodes something real about how the underlying system is organized. The prediction here is narrow and testable: not that a few outcomes dwarf the rest, which is true of almost anything, but that the ratio between them falls inside one specific, calculable range.

Brick-stamp Gibrat and Manuscript Yule process both test whether production processes -- stamped bricks, copied manuscripts -- grow the way Gibrat's proportionate-growth model predicts, purely from their own accumulated scale. Heaps' law in clay and Zipf in the archive look for the same vocabulary-growth and word-frequency laws linguists find in modern text, this time in a clay-tablet archive and in a citation record. LiDAR rank-size politics uses LiDAR-mapped settlement sizes to test whether a region's political integration shows up as a specific, textbook rank-size exponent rather than a looser hierarchy.