---
title: "Per Bak"
type: "entity"
entity_kind: "person"
status: "hub"
canonical_name: "Per Bak"
aliases: []
first_seen: "2026-08-30T00:00:00.000Z"
writer_model: "claude-sonnet-5"
connects_to: ["self-organized criticality","Bak-Tang-Wiesenfeld sandpile model","How Nature Works","threshold dynamics","punctuated equilibrium"]
seek_code_commit: "7d6d9ed"
---


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

- 2026-09-12: *How Nature Works* (1996) read directly as a second, independent
  primary beyond the 1987/88 papers — 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]])