nucleation

LIT-735: The Epistemic Benefit of Transient Diversity

2 min read

Properties6
StatusActive — read 2026-10-05 (NOTE-564): read in full from the author's posted manuscript (dated 29 September 2009). Worth reading for its conclusion, which Rosenstock et al. (LIT-733) single out as the robust result of this literature: what a community needs is transient diversity, long enough to keep a theory from being dropped too soon and short enough to allow convergence. Sparse communication and stubborn priors each supply it, and together they make diversity too stable and learning fails. The peptic-ulcer history is the motivating case.
Tagsepistemology, network-science, philosophy-of-science
ExtendsLIT-724 — The Communication Structure of Epistemic Communities
Corrected byLIT-733 — In Epistemic Networks, Is Less Really More?
Cited as “paper” byNOTE-564 — The Epistemic Benefit of Transient Diversity
Cited as “source” byTHEORY-148 — A group learning by trying needs diversity in what its members pursue that lasts long enough to survive misleading early evidence and then gives way to consensus; limited communication and stubborn priors each supply it, and together they make it permanent, so the group never settles

Graph View

Kevin J. S. Zollman (2010), Erkenntnis 72(1):17–35 — DOI-10.1007/s10670-009-9194-6

Key takeaways

  • The case. In the history of peptic ulcer disease as Zollman tells it, Palmer’s widely read study turned research towards acid suppression and away from the bacterial hypothesis, until H. pylori was identified (§1).
  • The model carries forward LIT-724: a two-armed bandit with beta priors, simulated for 10,000 rounds, with success probabilities of 0.5 and 0.499. Cycles beat wheels beat complete graphs, and over all six-agent networks denser networks are on average worse (Figures 3–4).
  • Dogmatism as a substitute. As the upper bound on the prior parameters rises, the complete network becomes the best of the three and the cycle the worst (Figure 6). Sparse networks with dogmatic agents often fail to agree on anything within the run (§3).
  • Conclusion. “At the heart of these models is one single virtue, transient diversity.” Limiting information and extreme priors are two ways to get it, but “both together make the diversity too stable, and result in a worse situation than either individually.”
  • Group versus individual. The scientists look epistemically sub-optimal one by one and optimal as a community, which Zollman reads as confirming Hull’s conjecture that the rational features of science belong to groups.

Standing in the record

Filed on 2026-10-05 at the owner’s request, with LIT-724. The account is THEORY-148. For an essay on organizational will, the useful result is the symmetry: the conditions that protect a minority view (low connectivity, stubbornness) are the same ones that, combined, stop a group from ever settling. That has the shape of the essay’s “protective lock-in” (§5.4), a protective regime that comes to sustain itself, in an epistemic setting.

Backlinks