Scott Alexander, curated
← Back to curation

Continental Drip

Quality
65
Strong
Claude Shift
35
Slight
RWI
1
of 10

Summary

A short post on 'continental drip' (landforms point south far more than north — Florida, Baja, Italy, Greece, India, Greenland), which Scott concedes is probably coincidence (not true of Pangaea, no mechanism). The real payload is epistemic: the hard problem of what to do with a suggestive pattern everyone calls coincidence. His sharp device — two early-20th-century children, one noticing 'drip' and one noticing 'drift' (the shape-fit of South America into Africa); the same teacher would dismiss both, being 100% right about one and strangling a Nobel-worthy insight (plate tectonics) about the other. Heuristics work until they don't; 'there is no royal road.' Plus an observational-humility note (he stares at maps constantly and never noticed drip).

Why this score

Quality 65 · Strong. Strong band, low. Brief but punches above its length: the drift/drip symmetry is a memorable, genuinely useful illustration of the pattern-vs-coincidence problem and the limits of the 'just a coincidence' heuristic. Held to lower-Strong only because the idea is stated, not developed.

Claude’s paradigm shift 35 · Slight. Slight-to-moderate. The pattern-vs-coincidence / no-royal-road point is standard epistemics; the drip-vs-drift framing is a fresh, vivid illustration rather than a new idea.

Real-world impact 1 · Negligible. A brief but memorable illustration of the pattern-vs-coincidence problem (the drift/drip symmetry); a stated-not-developed epistemic idea within the discourse, no material reach → RWI 1.