RIP Polyvagal Theory?

What the Polyvagal Debate Actually Means for Practitioners

There is a particular kind of discomfort that comes from discovering that a framework you’ve been using – one that genuinely helped people, that gave language to experiences previously without language – may be built on foundations that don’t hold.

To me, this doesn’t feel like the discomfort of being wrong about a fact. It’s something more unsettling: the possibility that the map was useful without being accurate, and that you now have to decide what to do with both of those things simultaneously.

That is roughly where the publication of Grossman et al.’s (2026) “Why the Polyvagal Theory Is Untenable” leaves practitioners who have incorporated Polyvagal Theory into their practice. Thirty-nine experts in vagal physiology, autonomic neuroscience, and vertebrate evolution signed their names to a paper arguing that the theory’s core physiological claims are not just contested but contradicted by decades of evidence.

Porges responded at length. And now those of us who work with actual human beings in actual rooms have to figure out what, if anything, to do with this.

The temptation (very understandable, very human) is to pick a side quickly and move on. Either the critique is an overreach by physiologists who don’t understand clinical application, or it’s a decisive demolition of pseudoscience that should have been challenged sooner. Neither of these responses quite fits what’s actually happening in this literature. Sitting with the real complexity is harder and, I think, more useful.

Let’s dig in, shall we?

Image from https://news.mit.edu/2021/neurons-humans-mammals-1110

What Grossman et al. Are Actually Arguing

It’s worth being precise about what the critique says, because it’s often mischaracterised in both directions.

The paper is not arguing that the nervous system is irrelevant to psychological states, that embodiment doesn’t matter in therapy, or that relational safety is a fiction. In actuality, it is arguing a very specific thing: that the physiological machinery Polyvagal Theory invokes to explain these things doesn’t work the way Porges claims it does.

The critique identifies several distinct failures:

First, respiratory sinus arrhythmia (the measure of heart rate variability that PVT has elevated to central importance as a readout of “vagal tone”) turns out to be influenced by so many competing factors (breathing rate, tidal volume, sympathetic activity, ageing, local cardiac influences, arterial CO₂ levels) that treating it as a clean index of central vagal output is, in the words of the authors, simply not defensible. RSA tells us something, but it doesn’t tell us what PVT says it tells us.

Second, the claimed hierarchy of autonomic responses (ventral vagal safety, sympathetic mobilisation, dorsal vagal shutdown) is not supported by the evidence on how these brainstem nuclei actually behave. The Dorsal Motor Nucleus of the vagus, which PVT assigns responsibility for the shutdown state Porges associates with dissociation and freeze, has been repeatedly shown to have limited and modest effects on heart rate in mammals. The evidence that it mediates the kind of profound bradycardia PVT describes is, at best, weak and species-variable. “Dorsal vagal shutdown,” as applied to human dissociation, has no established physiological basis.

Third, the evolutionary story (the idea that myelinated vagal fibres and the social engagement system represent a uniquely mammalian innovation) is flatly contradicted by comparative physiology. Myelinated cardiac vagal efferents exist across vertebrate groups including fish, reptiles, and birds. Respiratory heart rate variability appears across vertebrate classes. Non-mammalian animals display forms of social behaviour that complicate any clean story about mammals uniquely evolving a nervous system for connection. The ladder of phylogenetic progress that PVT leans on doesn’t exist in the form the theory requires.

Porges’ rebuttal is sophisticated, and it makes some genuine points. He argues, with some justice, that critics keep evaluating PVT as if it were making naive one-to-one anatomical claims, when it is actually a systems-level framework about functional organisation. He contends that RSA is meant as a pathway-specific index of ventral vagal cardioinhibitory influence, not a blunt measure of total vagal activity, and that this distinction matters enormously. He points out that finding respiratory-cardiac coupling in reptiles doesn’t falsify a claim about the integration of that coupling with social communication circuits in mammals specifically.

These are not trivial responses. But they run into a problem that is hard to dismiss: if the theory’s correct interpretation requires this level of nuance, and if that nuance has repeatedly been articulated in response to critics over twenty years without changing what gets taught, practiced, and published under the PVT banner, then something has gone wrong. A theory that is only correct when its author is present to interpret it has a robustness problem. The gap between PVT as Porges defends it in academic rebuttals and PVT as it circulates in clinical training, supervision, and psychoeducation is not a minor implementation issue – it is the theory’s most pressing problem.

What This Means for the Applied Claims

Here is where I think practitioners need to think carefully, because the failure of PVT’s physiology doesn’t straightforwardly invalidate everything built on top of it… but it doesn’t leave everything standing, either.

The psychological constructs that PVT popularised (safety, neuroception, co-regulation, state-dependent functioning, the body’s role in trauma) are largely real and important. But Grossman et al. are correct that almost none of these constructs originated with Polyvagal Theory.

• Safety, in the clinical sense, goes back to Rogers.

• Co-regulation and attachment originate with Bowlby.

• The body’s involvement in trauma and freeze responses has been documented in the psychophysiology and trauma literature for decades.

What PVT provided was a compelling neurophysiological narrative that appeared to ground these ideas in hard biology. That narrative has now been substantially dismantled.

The question is what happens to the clinical ideas when the biological scaffolding comes down. I think the honest answer is: it depends on how you’ve been using them.

If you’ve been using PVT vocabulary (ventral vagal, dorsal vagal, ladder of responses) as a loose metaphor to help clients make sense of their experience, as a way of communicating that their nervous system is doing something adaptive rather than evidence of personal failure, then the collapse of the specific physiology matters less. Metaphors can be clinically useful without being biologically precise. The problem arises if you’ve been presenting these as established neurobiological facts, because you’d then be providing clients with an inaccurate account of their own bodies – and Grossman et al. make the ethical point explicitly that this is not a neutral error.

If you’ve been making clinical decisions based on RSA measurement, treating it as a reliable readout of a client’s regulatory state, or using “vagal tone” as a guide to intervention, then the measurement critique lands more directly and requires revisiting.

If you’ve been using PVT to help clients develop a non-pathologising relationship with their physiological responses, to tolerate uncertainty in their own experience, to feel less shame about freeze or shutdown – that work rests on pre-existing foundations in trauma theory, attachment, and relational therapy that remain sound. It was never really PVT’s to claim.

The Problem with Porges’ Defence

The rebuttal is at its weakest, I think, when it becomes more concerned with the structure of the critique than its substance. Porges argues at length that Grossman et al. are attacking a straw man, that they’re conflating anatomy with physiology, that they’re applying the wrong level of analysis. Some of these methodological objections are legitimate. But they occupy so much of the rebuttal that the actual empirical questions (does RSA reliably index what PVT says it indexes? does the DMV mediate shutdown in humans?) end up somewhat sidestepped.

There is also something revealing in Porges’ appendix documenting “recurrent misrepresentations” – a table showing that many of Grossman’s objections have been raised before and previously addressed. He frames this as evidence that the critics keep failing to update. But it could equally be read as evidence that the theory’s clarifications haven’t resolved the underlying problems, merely responded to them. When the same empirical objections return decade after decade from different researchers, the burden falls on the theory to explain why, not on critics to explain why they keep raising them.

What an Intellectually Honest Clinical Response Looks Like

None of this means PVT-informed practice needs to be abandoned wholesale or that clinicians who’ve found it useful should feel duped. But intellectual honesty requires some adjustments.

It means being precise with clients about what we know and don’t know – saying “the nervous system seems to be involved in how we experience safety” rather than “your dorsal vagus is creating shutdown.” The first is defensible; the second isn’t.

It means taking the psychoeducation framing seriously as framing, not fact – which changes how confidently we should present it and how much we should invite clients to take ownership of the interpretation rather than absorbing it as a given.

It means being willing to hold frameworks lightly, to tell clients “this is a way of thinking about it that some people find helpful” rather than “this is what’s happening in your nervous system.” This may sound like a small shift, but it’s an ethically significant one.

And it means staying curious rather than defensive about where the science actually goes from here. The autonomic nervous system, embodiment, the relationship between physiological state and psychological experience – these are rich and legitimate areas of inquiry. The failure of one framework doesn’t close the inquiry; it redirects it toward accounts that are better evidenced and more careful about the distance between mechanism and metaphor.

The Harder Question

Underlying all of this is something that the debate surfaces but neither paper fully addresses: what does it mean to use a biological story therapeutically?

Polyvagal Theory succeeded partly because it gave clients a story, a narrative of their nervous system that was compassionate, that removed blame, and that made the body feel comprehensible rather than threatening. That is genuinely valuable clinical work. The question the Grossman critique forces is whether the specific story matters, or whether any sufficiently compelling and non-pathologising narrative would do the same work.

My suspicion is that the story does matter, but not primarily for the reasons clinicians typically assume. It matters not because clients need accurate neurophysiology (they mostly don’t) but because the kind of story we tell shapes the kind of inquiry we invite.

A story built on mechanistic certainty (”your ventral vagal brake is disengaged”) forecloses curiosity. A story built on honest uncertainty (”your body is doing something; let’s try to understand what”) opens it. On that measure, the post-Grossman world might actually be more therapeutically honest than the PVT world it’s unsettling (if we’re willing to let it be).

The challenge for practitioners is to resist the twin temptations of defensive loyalty on one side and wholesale abandonment on the other, and to occupy instead the more demanding position of someone who found a framework genuinely useful, takes seriously the evidence against it, and remains committed to following the question wherever it actually leads. That’s harder than having a settled view. It’s also, probably, what good clinical thinking looks like.