Per Bak

Danish theoretical physicist (1948–2002) who, with Chao Tang and Kurt Wiesenfeld, introduced self-organized criticality in 1987 — the theory that certain slowly-driven dissipative systems drive themselves to a critical point without any external parameter ever being tuned there. Popularized the idea for a general audience in How Nature Works: The Science of Self-Organized Criticality (1996).

Matters to this vault as the originator of the theory claim-engineered-threshold-is-not-self-organized-criticality tests hand-tuned AI-memory triggers against, and as the figure a 2015 refereed review (Watkins, Pruessner, Chapman, Crosby & Jensen) rereads word-by-word, with page citations, specifically to separate what he actually claimed from twenty-five years of secondary distortion.

References

• claim-btw-1987-soc-critical-point-needs-no-external-tuning
• claim-soc-tier1-literature-places-threshold-among-sufficient-not-necessary-conditions
• claim-soc-necessity-of-threshold-driven-interaction-remains-open-question
• claim-btw-near-threshold-clustering-is-1d-sandpile-coincidence-not-general-soc-signature
• claim-engineered-threshold-is-not-self-organized-criticality
• claim-caracciolo-sportiello-2012-confirms-17-8-average-height

Updates

• 2026-09-26 (propagation-repair): the 2026-09-12 entry above was corrected in place — was: How Nature Works read directly as "a second, independent primary" beyond the 1987/88 papers; now: a second primary but not independent corroboration, because Bak is the same author as BTW 1987, so under sources.md's independence test (different author group and venue) this is the same voice read directly rather than through an intermediary. (claim-bak-how-nature-works-independently-confirms-soc-no-tuning)
• 2026-09-12: How Nature Works (1996) read directly as a second primary beyond the 1987/88 papers — not independent corroboration, since Bak is the same author as BTW 1987, but his own voice read directly rather than through a secondary's quotation. It restates the no-tuning property in Bak's own first-person voice and reserves the word "necessary" for the non-SOC (tuned) case, never for thresholds within SOC. (claim-bak-how-nature-works-independently-confirms-soc-no-tuning, claim-necessary-sufficient-soc-threshold-framing-absent-from-primary-record)
• 2026-08-31: The specific "well below threshold" figure this page's companion note (claim-btw-near-threshold-clustering-is-1d-sandpile-coincidence-not-general-soc-signature) cites for the two-dimensional BTW sandpile — average height 17/8 = 2.125 against a threshold of 3 — is now confirmed directly against its own primary rather than only against the review reporting it: Caracciolo & Sportiello's 2012 paper states the figure in its own abstract as an exact analytic result. (claim-caracciolo-sportiello-2012-confirms-17-8-average-height)