Unwinding: A History, a Mechanism, and an Unresolved Argument

Section One: Viola Frymann and the Osteopathic Origin

Viola Frymann began her work in the cranial osteopathic tradition in the early 1950s, after the death of her newborn son led her to a course taught by William Garner Sutherland, DO — the physician who founded the cranial concept in osteopathy, the idea that subtle, rhythmic motion in the bones and membranes of the skull reflects and influences the health of the whole body. From there she became one of Sutherland's closest students, went on to found the Osteopathic Center for Children, and became a founding figure at the College of Osteopathic Medicine of the Pacific. She was eventually awarded the Andrew Taylor Still Medallion of Honor — the highest award the American Academy of Osteopathy gives — and taught the cranial concept on nearly every continent over the course of her career, like most pioneering figures, not without controversy along the way.

She is generally credited, including by Ruth Duncan in the article and comment thread that prompted this essay, with coining “unwinding” as a clinical term — working from Sutherland's premise that the body carries an intrinsic, self-correcting force, and that the practitioner's role is to facilitate that force rather than drive it. Frymann herself described the principle plainly: to place the patient in the position they were in at the moment of injury, and permit the fascia to move through whatever motions are necessary to eliminate the forces the impact imposed. There is footage still in circulation of Frymann demonstrating this directly: an early recording of her working with a whiplash patient, guiding what's described as an “indirect myofascial unwind.” It is one of the earliest documented instances of what would later be called unwinding, predating John Barnes bringing the term into massage and physical therapy circles by decades.

Section Two: John Barnes and the Trauma-Storage Reframing

If Frymann treated unwinding as an expression of the body's intrinsic, self-correcting force, John Barnes recast it as something closer to extraction. Starting in the 1970s and building through a three-part series in Massage Magazine (2006–2007), Barnes described his approach as a triangle with three points: structural myofascial release, myofascial unwinding, and myofascial rebounding, each working toward what he called “complete healing.” Central to his teaching is the claim that the body “remembers everything that ever happened to it” — that fascia doesn't just restrict movement mechanically, it holds emotional and autobiographical memory, and unwinding is the process by which that memory surfaces and releases. In 2020 he put it more vividly still: unwinding, he wrote, “finds positions of past traumas that are somehow embedded in the fascial system,” after which “the subconscious holding patterns start to release their iron grip on our clients, and the straitjacket of fascial restrictions begin to soften, melt, rehydrate and glide.”

Barnes has trained more than 100,000 therapists over the course of his career, and his influence on how a whole generation of practitioners talks about the body is considerable. His framing has also stayed remarkably consistent over time: as recently as 2025, on Sue Hitzmann's Anatomical Gangster podcast, he was still describing fascia and unwinding in the same register he used nearly twenty years earlier — quantum potential, energetic dimensions, deeper states of awareness, and a reference to Russian “torsion wave” physics said to travel faster than light. That consistency says something genuine about the depth of his conviction; it is not a metaphor he later set aside. It also means, in the same spirit of care this essay is trying to bring to every figure in it, that his theoretical framework sits apart from the kind of evidence the rest of this piece is weighing everything against — worth naming plainly, not as a dismissal of his practical contribution, but as an honest account of where the science currently stands.

Section Three: Minasny and the Ideomotor Mechanism

Budiman Minasny's 2009 paper remains the only formal, peer-reviewed attempt at a mechanism specifically for fascial unwinding, and it rests on ideomotor action — the concept William Carpenter described in 1852 to explain the table-turning and dowsing crazes of Victorian England. The modern account of how an ideomotor response is built is more specific than Carpenter's: it begins with what's called “motor babbling” — unstructured, exploratory movement with no goal attached, the way an infant repeats a sound purely to explore what it feels and sounds like, not to achieve anything. Through repetition, an association forms between the movement and its felt consequence. Once that association is set, the sequence can run in reverse: simply holding the felt outcome in mind is enough to trigger the movement automatically, without any intention or plan preceding it.

That is structurally different from how a predictive-processing account works, where a goal comes first, a motor plan is built to achieve it, a prediction is made about how the resulting sensation should feel, and any mismatch between predicted and actual sensation drives correction. Ideomotor action runs the opposite direction: felt outcome — automatic movement, with no plan or goal involved at all. This matters because it means Minasny's own mechanism for unwinding sits closer to the ideomotor, non-predictive side of current cognitive science than to the Haggard/Barrett account often used to explain the same phenomenon.

Minasny proposes unwinding works through three channels: stimulation of fascial mechanoreceptors (pressure-sensing nerve endings in fascia) producing reflex motor effects; subtle guidance of movement direction from the practitioner's technique; and a deep relaxation that “switches off” the conscious mind's usual inhibition of movement. Two conditions have to be met for it to occur: the practitioner needs fine palpation skill, and the patient needs to be able to relax and let go.

Crucially, Minasny's own paper explicitly rejects the tissue-memory theory that Barnes held so passionately throughout his career. His conclusion states plainly that a touch or position triggering a memory does not mean the memory was stored in the tissue itself. He calls this state-dependent memory: the memory remains stored in the central nervous system, and the body position or touch functions only as a trigger, not a container. The paper most often cited to ground unwinding in “real neuroscience” is, at its origin, already arguing directly against the trauma-storage claim that dominates the field's popular language.

Other Proposed Mechanisms

Minasny's ideomotor/autonomic model isn't the only mechanism proposed for involuntary movement in manual therapy — a few others sit alongside it, none formally reconciled with each other or with his account.

Pandiculations. Luiz Fernando Bertolucci first connected involuntary movement in manual therapy to pandiculation in 2011, following EMG recordings he published in 2008 showing involuntary tonic activity in the cervical erector muscles and involuntary eye movements during a technique he called Muscle Repositioning — discovered, by his own account, serendipitously, while working on fellow Rolfers who reported the maneuver felt distinctly different from anything familiar to them. Robert Schleip later (2017) described the same underlying pattern more generally: the stretch-and-shudder reflex seen in yawning, or in a cat waking from sleep, elicited through slow, resisted, alternating movement over sixty to ninety seconds.

Somatic Experiencing. Peter Payne, Peter Levine, and Mardi Crane-Godreau (2015) offer a nervous-system account rather than a fascial one: the sympathetic nervous system can become stuck in a state of excess activation, disturbing the proprioceptive feedback loop that would normally let the parasympathetic system re-engage. In this model, involuntary shaking and postural shifts are the nervous system's own mechanism for resetting that imbalance.

What's notable across these: none of them agree about what unwinding actually is, and none have been tested against each other, or against Minasny's ideomotor account, within the same study.

Pandiculation's Two Lineages

Pandiculation is often treated as a single, settled concept, but it carries two distinct lineages worth separating. Schleip's and Bertolucci's version is spontaneous: the stretch-and-yawn reflex, arising unprompted, elicited but not manufactured.

A second, related lineage comes through trauma therapy rather than fascia research. Peter Levine, developing Somatic Experiencing, drew on Stephen Porges' polyvagal theory (1994) to argue that animals in the wild discharge fight-or-flight activation through spontaneous trembling, and that humans often suppress this same response. David Berceli built directly on this to create Tension and Trauma Releasing Exercises (TRE) — a set of specific poses designed to fatigue the psoas and legs until they trigger what he calls a neurogenic tremor. This is a contrived pathway to a visually similar output: rather than waiting for an unprompted reflex, a specific protocol is used to manufacture one, because the spontaneous version cannot be reliably produced on command.

There is also a precision worth holding here, since the polyvagal foundation under Levine's and Berceli's account is itself genuinely disputed. The vagus nerve is real, uncontroversial anatomy, and heart rate variability is a validated, if imperfect, proxy for parasympathetic activity — so “touch and movement can influence autonomic state, and there is a plausible way to measure that” is defensible. What is specifically contested is Porges' three-tiered hierarchy of dorsal vagal, sympathetic, and ventral vagal states, and the “vagal brake” mechanism said to distinguish them. A thirty-nine-author critique led by Paul Grossman, published in 2026 in Clinical Neuropsychiatry, argues these specific claims are not supported by the underlying neurophysiology; Porges published a formal rebuttal the same year, in the same journal, arguing the critique misrepresents the theory's actual claims. That dispute is live and unresolved as of this writing — which means Levine's and Berceli's shaking-as-discharge account rests on a foundation currently being contested by vagal physiologists themselves, independent of anything to do with fascia or unwinding.

Taken together, the mechanisms proposed for unwinding, pandiculation, and neurogenic tremor share a pattern worth naming plainly: each rests on a different, separately contested theoretical foundation — ideomotor action, mechanical shear between fascial layers, autonomic discharge, a still-disputed polyvagal hierarchy — and none of them have been tested against each other, or against a shared standard of evidence. The field has multiple candidate explanations running in parallel, each borrowed from a different corner of physiology or psychology, held together mostly by the fact that they all produce movement that looks, from the outside, like the same thing.

Section Four: Two Accounts of What's Actually Happening

Everything traced so far — Frymann's intrinsic force, Barnes' stored trauma, Minasny's ideomotor reflex, Bertolucci's and Schleip's pandiculation, Levine's discharge — eventually runs into the same question from two different directions in current science: does the movement arise because the brain predicted it, or because it was never a prediction in the first place?

The predictive-processing account holds that the brain is constantly generating a model of the body and its likely next state, comparing incoming sensation against that prediction, and adjusting the model — or the movement — when the two do not match. Patrick Haggard's 2017 review is the standard reference for how this produces our sense of agency: the felt experience of “I did that” comes from a match between predicted and actual sensory outcome, not from direct awareness of the decision itself. Lisa Feldman Barrett's account of emotion (2017) extends the same logic to feeling: an emotion is not a fixed program that fires, but the brain's best prediction, built from past experience, about what a given bodily state means. On this view, unwinding is what happens when touch, stillness, and reduced distraction let the brain's ordinary predictive machinery run somewhat unsupervised — sensation feeds movement, movement feeds new sensation, and occasionally a strong prediction error surfaces a memory or an emotional response along the way.

The ideomotor account, sitting inside Minasny's own 2009 paper, runs in something closer to reverse, as described above: motor babbling builds an association between movement and felt consequence, and once that association is set, holding the felt outcome in mind is enough to trigger the movement with no plan or goal preceding it. This means the paper most often cited to ground unwinding in neuroscience is not, in fact, offering a predictive-processing account — it is offering something structurally closer to the enactive side of the argument.

This is where Merleau-Ponty's objection becomes more than a philosophical aside. His argument, in the Phenomenology of Perception (1945), was that the body is not an object the mind runs predictions on from a distance — it is the very thing doing the perceiving and moving, with its own pre-reflective intelligence that a purely computational account risks treating as an afterthought. Francisco Varela, Evan Thompson, and Eleanor Rosch built this into an actual cognitive-science research program in The Embodied Mind (1991): cognition, in their account, is “enaction” — brought forth through an organism's ongoing coupling with its environment, not computed internally and then executed. Evan Thompson extended the same argument into biology in Mind in Life (2007). Maxine Sheets-Johnstone sharpened it further, arguing in “Thinking in Movement” (1981) and The Primacy of Movement (1999) that movement itself is a mode of thought, not something downstream of a plan — and notably, she argues Merleau-Ponty himself did not go far enough, staying too focused on third-person description rather than the first-person, felt sense of movement as it happens.

The sharpest, most current version of this disagreement comes from Shaun Gallagher (2022), who argues predictive processing and enactivism are not two compatible descriptions of the same event but are, on cases like this one, genuinely incompatible. Predictive processing is a framework built to minimize surprise. Genuinely spontaneous, unrehearsed movement — his example is musical improvisation, but the same logic extends to unwinding — is by definition a surprise, even to the person producing it. A framework designed to reduce surprise runs into a structural problem explaining an event whose defining feature is that it was not predicted. Enactivism, which treats the body-environment coupling as the source of novelty rather than the source of error to be corrected, handles this case more naturally. This is not settled in either direction — researchers are actively working on both sides of it — but it means this essay's central tension is not manufactured. It is a live fault line inside current cognitive science itself, and Minasny's own mechanism for unwinding sits closer to the enactive side of that fault line than the predictive-processing framing usually laid over it.

Section Five: Mechanism, Meaning, and the Question of What's “Deeper”

A natural response to all of this is to ask whether there is something more going on — whether sensation, movement, prediction, memory, and emotion interacting, however accurately described, is really the whole story. That question deserves to be taken apart rather than answered in one direction, because “deeper” can mean at least three different things, and the field routinely runs them together.

It could mean there is an additional hidden cause — fascia literally storing an event the way a recording stores a performance, waiting to be played back. This is the traditional trauma-storage claim, and it does not hold up: there is no known mechanism by which connective tissue stores an experience, and Minasny's own paper explicitly rejects it in favor of state-dependent memory.

It could mean, instead, that a complete and accurate account of the mechanism still does not tell us what the experience means to the person having it, or what it is worth in a therapeutic relationship. This is not a rival explanation to the neuroscience — it is a different question altogether, one a mechanism was never built to answer. Explaining how a rainbow forms does not make it stop being beautiful to look at; explaining unwinding as sensorimotor prediction does not settle whether, or how, it matters to the person on the table.

And it could mean something sharper still: that the five-part decomposition itself — sensation, movement, prediction, memory, emotion, treated as separate components that then “interact” — already assumes something that may not be true. This is closer to Gallagher's and Sheets-Johnstone's actual objection: not that a sixth ingredient is missing, but that carving a single, undivided lived event into five parts and recombining them may be a distortion introduced by the act of explaining it, rather than a discovery about how the event was actually structured to begin with.

Held apart like this, the field's usual move — reaching for something “deeper” in the first sense, a hidden mechanical or energetic cause — looks like exactly the wrong response to what is genuinely under-examined in the second and third senses. The honest position is not that unwinding is “nothing but” prediction, and not that something mystical remains unexplained. It is that a good mechanical account and a good account of significance are different projects, and conflating them is most of what has kept this field circling the same unresolved argument for over a century.

Section Six: What's at Stake

None of this is only an argument about which neuroscience paper is more correct. How a practitioner understands unwinding shapes what they say to a client mid-session, what they imply about where a movement “came from,” and whether a client is left believing something was extracted from their tissue or simply that their own nervous system, in a moment of safety and stillness, did something it is always quietly capable of doing.

Frymann's language — an intrinsic force, facilitated rather than driven — and the enactive account converging on it from the opposite, contemporary direction both point toward a posture of following rather than retrieving. Barnes' language, and the trauma-storage framing built on it, points toward a posture of extraction — something is in there, and the practitioner's skill is what gets it out. These are not neutral descriptions. They carry different implications for how much authority the practitioner claims over the client's experience, and how much the client is left believing about their own body's reliability.

The honest place to leave this, at least for now: unwinding is real, in the sense that people reliably experience it and it can matter to them. What it is remains genuinely unsettled — not because the field lacks theories, but because it has too many, running in parallel, rarely tested against one another, and rarely held to the same evidentiary standard the rest of manual therapy claims to aspire to. That unsettledness is not a reason to retreat into mysticism, and it is not a reason to overclaim certainty in either the predictive-processing or the enactive direction. It is a reason to keep asking the question with the same care this piece has tried to bring to it.