Presentation Framework PALM Articles Books Research Tools Contact
Back to articles
2026

The structural mechanism described by Holling

Rigidity, brittleness, unexpected collapse.

Holling

The triad rigidity, brittleness, unexpected collapse is one of the most robust invariants in complex systems ecology since Holling. Yet it remains very poorly integrated into real-world practices.

1) What Holling actually describes: an observable dynamic mechanism

Holling does not describe a psychological metaphor or a simple management style. He describes a dynamic mechanism, observable in ecosystems, but also transposable to organizations, institutions, and socio-technical systems.

Three operations chain together.

Variability reduction

Managers eliminate fluctuations deemed inefficient or dangerous. Examples: suppression of natural fires, procedure standardization, elimination of local margins.

Progressive loss of redundancy and functional diversity

The system becomes cleaner, more optimized, more apparently performant, but also more fragile. Real resilience decreases even as apparent performance increases.

Silent accumulation of vulnerabilities

The system becomes brittle: it functions as long as conditions remain close to those anticipated, but it no longer absorbs deviations.

KEY POINT

Rigidity is not a state. It is a self-reinforcing process.

2) Why this mechanism is systematically underestimated

The blindness is structural, not merely individual.

Immediate performance bias: what is measured short-term is valued. Resilience remains invisible until it has been lost.
Confusion between stability and resilience: prolonged stability is read as proof of solidity, when it may be a symptom of advanced rigidification.
Command-and-control pathology (Holling and Meffe, 1996): all variability is treated as a threat to eliminate, which accelerates resilience loss.
Expertise illusion: many resilience discourses remain linear. They ignore thresholds, regimes, tipping points, and therefore the real relevant metrics.

3) Typical consequences by domain

Policy

Public policies centered on apparent stability rather than adaptive capacity. Low tolerance for local experimentation. Loss of institutional diversity.

Management

Continuous optimization, lean, KPIs, standardization, variability reduction, fragilization. Unexpected collapses that were structurally predictable.

Education

Elimination of error, exploration, uncertainty. Production of individuals performant under nominal conditions, less adaptive under perturbation.

Resilience experts (institutional sense)

Vocabulary adoption without mental model transformation. Resilience reduced to business continuity or crisis management, whereas real resilience is an emergent property of a diverse, flexible, and redundant system.

4) Resilience is a dynamic regime

A system is not resilient or non-resilient as a fixed property. It is in a resilience regime, or in a brittleness regime. And the transition between the two is often:

Silent — nonlinear — difficult to reverse in the short term — detectable mainly through weak signals.

5) Operational reading: detecting the shift toward brittleness

One can monitor cross-cutting indicators, valid for both organizations and ecosystems. The important point is the combination, not an isolated indicator.

Structural indicators

Declining functional diversity (fewer alternative roles, fewer bypass paths). Redundancy erosion (elimination of duplicates, fat-cutting, weakened fallbacks). Reduced local margins (decisions and resources recentralized, lower autonomy).

Dynamic indicators

Increasingly uniform responses to different perturbations (loss of adaptivity). Recovery time lengthening after small incidents (resilience fatigue). Growing intolerance to uncertainty (need for control, decline in experimentation).

Narrative and cognitive indicators

Internal narrative centered on mastery, zero defect, compliance rather than learning, adaptation, controlled variation. Systematic disqualification of weak signals (deviations, anomalies, field alerts).

Summary

Holling's mechanism is not an abstract theory. It is an observable, measurable, and above all avoidable process, provided one monitors margins rather than nominal performance alone.