In many economic and geographic systems, the way we rank and order things — such as income distribution, or population number — is abstract and transitory. Michael Batty, director of the Centre for Advanced Spatial Analysis at University College London, has constructed a visual model, called a rank clock to track the rise and fall of cities and civilizations over time (see page 592).

Batty's interest in urban dynamics dates to the 1960s, when he studied architecture. But it was rekindled three years ago when he visited colleagues at Ann Arbor who were interested in the laws of rank–size distribution, which look at how amounts of resources, people and products help determine a city's composition. Batty subsequently examined data for US cities. His former PhD student, Naru Shiode, now at the State University of New York in Buffalo, suggested that what was needed was a new method of visualization.

But the final push for Batty came from learning about Doug White's work at the University of California, Irvine. White was working on the very-long-term dynamics of cities. When examining these data with rank clocks, Batty found an “amazing volatility” over a 2,500-year period, a pattern that conventional models failed to replicate, and that his model takes into account.