The Load-Bearing Nervous System

Rob Merivale
18 Dec 2025 • ~28 min read

A Thermodynamic and Engineering-Aligned Model of Human Behaviour Under Stress

Rob Merivale
18 December 2025

Correspondence

Academic or technical correspondence regarding this paper may be directed via:
science@robmerivae.com

This paper is a preprint and has not yet undergone peer review. It is published to invite critique, clarification, and interdisciplinary discussion.

Abstract

Current models of stress and trauma describe behavioural correlates but lack a mechanistic framework capable of predicting stability, collapse, and recovery under changing conditions. In contrast, engineering disciplines routinely model how complex systems behave under increasing load using principles of energy cost, entropy, nonlinear failure, and hysteresis.

This paper proposes a load-bearing model of the human nervous system grounded in thermodynamics, engineering load theory, network neuroscience, and stress physiology. Behaviour under stress is reframed as a load-dependent state expression rather than evidence of damage or defect. A formal persistence model is introduced, integrating coherence, stability, and entropy into a single dynamical expression that generates falsifiable predictions.

The contribution is theoretical and conceptual. It does not propose a governing law, universal optimisation principle, or normative prescription. Its aim is to restore mechanistic clarity to the interpretation of stress, reactivity, collapse, and recovery by aligning psychological explanation with established physical constraints.

1. Introduction — Why Behaviour Under Stress Lacks a Mechanical Explanation

Despite decades of research, psychology still lacks a mechanistic account of how the nervous system behaves under increasing stress. Existing frameworks describe symptoms, diagnoses, and developmental associations, but they do not explain why behaviour degrades, destabilises, or collapses in the characteristic ways repeatedly observed under pressure.

In engineering, these questions are not controversial. Load-bearing systems are analysed according to how performance changes as demand increases. As load rises, handling degrades, energy consumption increases, tolerances narrow, and failure probability accelerates nonlinearly. None of this implies damage. It reflects physics.

No equivalent framework governs mainstream interpretations of human stress behaviour.

As a result, behavioural changes under stress are frequently moralised or pathologised. Reactivity, volatility, shutdown, and dissociation are treated as signs of weakness or disorder rather than as predictable outputs of an overloaded system. This interpretive gap obscures mechanism and produces systematic error.

A core problem is the failure to distinguish architecture from state. Neural architecture — network organisation, regulatory capacity, baseline coherence — is relatively stable absent injury. Behaviour, by contrast, is highly state-dependent. When load-driven behaviour is mistaken for trait, explanation collapses.

This paper argues that these failures arise from the absence of a load-bearing model. When the nervous system is treated as an energy-constrained, entropy-sensitive dynamical system, stress behaviour becomes predictable rather than mysterious.

2. Why Existing Trauma and Stress Models Fail Mechanically

Most trauma frameworks frame stress responses as pathology. While clinically pragmatic, this obscures a critical fact: many trauma-associated behaviours are functional adaptations to excessive load, not evidence of structural damage.

A second failure is state–trait conflation. Behaviour expressed under load is routinely interpreted as enduring personality or disorder despite overwhelming evidence of context sensitivity.

Third, psychology lacks the core concepts required to model stressed systems:

• Load magnitude
• Load distribution
• Stability margins
• Energy cost
• Nonlinear failure

Thermodynamic constraints are acknowledged but not formalised. Entropy is treated as metaphor rather than mechanism.

As a result, existing models cannot explain:

• Abrupt collapse after gradual stress accumulation
• Rapid improvement following modest load reduction
• Hysteresis in recovery
• Domain-specific vulnerability shaped by development

These are not anomalies. They are the defining signatures of load-bearing systems near critical thresholds.

3. Load-Bearing Systems: The Missing Template

All load-bearing systems obey the same constraints.

Architecture defines capacity. Load degrades performance. Load distribution determines stability. Energy consumption rises nonlinearly. Sustained demand produces overheating. Environmental unpredictability narrows safety margins. Recovery occurs through load reduction, not redesign. Operating beyond tolerance produces collateral damage.

If these principles are accepted anywhere in physics or engineering, they cannot be arbitrarily excluded from biology.

The nervous system is not exempt from energy limits, entropy, or nonlinear failure. It is a load-bearing system operating under the same constraints as any other complex, energy-dependent structure.

4. The Nervous System as a Load-Bearing Architecture

Neural architecture comprises relatively stable features: connectivity patterns, regulatory pathways, and baseline signal coherence. These do not fluctuate moment-to-moment.

Behaviour does.

Under increasing load, attentional bandwidth narrows, inhibitory control degrades, flexibility declines, and error rates rise. Predictive systems become noisier. Self-referential narrative increases as signal integrity falls. Panic, shutdown, and dissociation emerge as overheating responses.

Development matters because load distribution matters. Early unpredictability shifts thresholds permanently, producing domain-specific sensitivity rather than global fragility.

Egoic narrative under stress functions as mechanical squeak — noise produced by strain, not insight.

None of this requires pathology. It follows directly from operating near capacity.

5. A Minimal Thermodynamic Persistence Model

If behaviour reflects load rather than defect, persistence must depend on three factors: coherence, stability, and entropy.

This yields the minimal form:

P = C × S² × e⁻ᴱ

Where:

• P = behavioural persistence
• C = coherence
• S = stability
• E = entropy

Entropy degrades performance multiplicatively, justifying the exponential term. Stability compounds regulatory capacity, justifying the quadratic term.

Coherence decomposes into:

C = C_structural × C_functional

Structural coherence is trait-like. Functional coherence is load-dependent.

Entropy decomposes into:

E = E_external + E_internal

Environmental unpredictability amplifies internal narrative noise.

Given these constraints, collapse is inevitable once entropy overwhelms stability margins.

6. Collapse, Recovery, and Hysteresis

Collapse is a phase transition, not gradual decline. Recovery exhibits hysteresis: the load required to recover is lower than the load that caused failure.

Recovery occurs through annealing — graded load reduction. Abrupt unloading risks rebound instability.

Overloaded systems also cause damage externally. Behavioural collisions — relational harm, misattunement, conflict — are predictable consequences of operating beyond capacity.

7. Worked Example

Two individuals share identical architecture.

Individual A
C = 0.9
S = 0.8
E = 0.5
→ Persistence maintained

Individual B
C = 0.9
S = 0.6
E = 0.9
→ Collapse

Architecture is not the explanatory variable. Load is.

8. Predictions and Falsifiability

The model predicts:

• Nonlinear recovery from entropy reduction
• Quadratic gains from stability improvement
• State-dependent behaviour with trait-stable architecture
• Early warning signals before collapse
• Hysteresis in recovery trajectories

These predictions are falsifiable.

9. Empirical Alignment

Thermodynamic discontinuity methods identify collapse thresholds in biological systems. EEG seizure onset correlates strongly with entropy markers (reported r ≈ 0.935). Patterns align with critical transition theory.

10. Cross-Domain Implications

• Clinical psychology reframes reactivity as load
• Neuroscience aligns with dynamic connectome models
• Developmental science foregrounds predictability
• Artificial systems show analogous noise saturation
• Institutions benefit from entropy-aware design

11. Limitations

The model abstracts molecular detail, relies on indirect entropy measures, and requires empirical parameterisation. Narrative processes lack a perfect mechanical analogue. AI parallels are analogical.

12. Paradigm Comparison

Conventional Models

• Pathology-focused
• Trait-centred
• Descriptive
• Linear recovery
• Moralised

Load-Bearing Model

• Mechanistic
• State-dependent
• Predictive
• Hysteresis-aware
• Non-pathologising

13. Conclusion

The nervous system behaves like any load-bearing system. Behaviour under stress reflects load, not damage. Entropy reduction restores stability. Aligning psychology with engineering and thermodynamics restores explanatory power and predictive clarity.

Scope and Limitations

This paper is a theoretical synthesis. It does not propose a governing law, universal principle, or normative framework. Its scope is limited to mechanistic interpretation of stress behaviour using established physical constraints.

References

Bonanno, G. A. (2004). Loss, trauma, and human resilience. American Psychologist, 59(1), 20–28.

Carhart-Harris, R. L., & Friston, K. J. (2019). REBUS and the anarchic brain. Pharmacological Reviews, 71(3), 316–344.

McEwen, B. S. (1998). Protective and damaging effects of stress mediators. New England Journal of Medicine, 338, 171–179.

Scheffer, M., et al. (2009). Early-warning signals for critical transitions. Nature, 461, 53–59.