Comparison
Theories of emotional escalation: from amygdala hijack to polyvagal theory
May 9, 2026 · 4 min read
Why do people lose their cool? Four research traditions answer that question differently — and the differences matter for what you do about it. The popular “amygdala hijack” frame is widely cited; it’s also widely critiqued by neuroscientists as oversimplified. Below: the four perspectives with the most reach, what each one gets right, and where each one falls short.
Goleman — Amygdala Hijack
The work: Daniel Goleman, Emotional Intelligence (1995). Coined “amygdala hijack” to describe moments when the emotional brain overrides the reasoning brain.
What it explains: Why a heated argument can flip someone from articulate to incoherent in seconds, and why “just calm down and think” rarely works in the moment.
What it misses or oversimplifies: Later neuroscience criticized “hijack” as a metaphor that overstates the disconnect between brain systems. The amygdala doesn’t really commandeer the cortex; processing happens in parallel, and the cortex isn’t bypassed so much as briefly outpaced.
LeDoux — Low Road / High Road
The work: Joseph LeDoux, NYU Center for Neural Science. Distinguished two pathways for threat processing. The low road runs through subcortical circuitry: quick, imprecise, focused on detection. The high road runs through cortex: slower, refined, focused on context. Both run; the low road just fires first.
What it explains: Why initial reactions to threat are often crude and disproportionate, and why the more measured response arrives a beat later — but does arrive.
What it misses or oversimplifies: Qualifies Goleman without replacing him. The model describes threat circuitry but doesn’t fully account for how prior beliefs shape what registers as a threat to begin with.
Porges — Polyvagal Theory
The work: Stephen Porges, The Polyvagal Theory (W. W. Norton, 2011). Locates conflict response in the autonomic nervous system, with the vagus nerve as regulator. Introduces neuroception — the body’s unconscious read of safety vs. threat in the environment.
What it explains: Why tone, eye contact, and body language affect de-escalation more than the words; why purely-cognitive reasoning sometimes fails to reach a dysregulated person.
What it misses or oversimplifies: Operates at a different layer than Goleman or LeDoux — autonomic, not amygdala-centric. Its physiological claims are also the most seriously contested of the four theories here. In February 2026, Paul Grossman and 38 co-authors published “Why the Polyvagal Theory Is Untenable” in Clinical Neuropsychiatry, arguing that the theory’s core premises are not defensible on existing evidence; Porges published a detailed response in the same issue. The dispute is about plumbing — whether the two brainstem vagal pathways are as functionally separate as the theory holds, whether heart-rate variability measures what the theory says it measures, and whether the evolutionary framing is accurate. The critics grant what the clinical world mostly took from the theory anyway: the emphasis on safety, co-regulation, and the body’s role in regulation predates polyvagal theory and does not fall with it. Neuroception — the unconscious read of safety versus threat — is not among the challenged premises, and the underlying observation it names is independently supported by the threat-detection work in LeDoux’s section above.
Lazarus and Frijda — Appraisal Theory
The work: Richard Lazarus, Nico Frijda, and colleagues, reviewed in Emotion Review (2013). Treats emotions as cognitive evaluations of meaning, not automatic responses. Anger isn’t triggered by threat per se; it’s triggered by the appraisal that the threat is unfair, that you’re helpless to stop it, or that someone else is to blame.
What it explains: Why the same situation produces different emotions in different people; why reframing an interaction (“they’re stressed, not hostile”) can defuse it; why apologies that name impact tend to do the work of repair.
What it misses or oversimplifies: Shifts the unit of analysis from biology to meaning. Less useful in moments of full physiological dysregulation, when no amount of reframing arrives in time.
What they share, where they disagree
All four converge on one observation: in moments of high emotion, the slow, deliberate, perspective-taking parts of cognition aren’t fully available. Whatever you call it (amygdala hijack, low-road processing, autonomic dysregulation, mismatched appraisal), the practical reality is the same: you’re not arguing with someone at full capacity.
Where they disagree: each locates the mechanism at a different level. Goleman in the amygdala; LeDoux in parallel pathways through it; Porges in the autonomic layer below it; Lazarus and Frijda in the cognitive layer above it. The disagreements imply different intervention points: body, breath, body language, meaning. None of those invalidates the convergence.
Related: Why logic doesn’t work in arguments · How to de-escalate an argument
From the essay: Read the full piece →
Want more like this? Subscribe to the newsletter.