Why can't classical osteopathic mechanics be simpler?
That is the question I started with. You know the one. You have probably had the same conversation I have, with students or with other practitioners coming at this from outside the lineage. The polygon. The lines. The pivots. The gravity centres. Four arch schemes. Two pyramids. Curved lines and straight lines. AP, PA, AC, PC. C7, D4, L3. Why is there so much of it? Why can we not have a tidier model that does the same job?
I had a vague hope that someone, somewhere, had worked out a cleaner version. Maybe biotensegrity. Maybe modern fascial research. Maybe in one of the dynamical systems, people had quietly cracked it? So, I decided to ask the cleanest, most patient, most well-read interlocutor I could think of. I sat down with an AI, supposedly the most powerful current iteration commercially available, and said: Design me a better mechanical model. Make it simpler. Make it more defensible in modern biomechanical terms. Make it something I can hand to a sports therapist and have them grasp it without years of training.
The AI was happy to oblige. It is always happy to oblige. Within a few exchanges, I had a "pre-stressed biphasic continuum", a beautiful, modern, fluid-and-fibre framework that sounded thoroughly contemporary and would have looked at home in any current biomechanics journal. I asked for more options. It produced six. Tensegrity-derived models. Morphogenetic field models, in which the lines fall where they do because of embryological folding. Viscoelastic flow models, treating the body as a slow non-Newtonian fluid creeping under sustained gravitational load. Control-systems models, with set-points and gain and reference signals. Hierarchical scale-coupled models extending Ingber's cellular tensegrity to the whole organism. Energetic-thermodynamic accounts of posture as the lowest-energy configuration. Each one elegant. Each one fitted with appropriate scientific vocabulary. Each one ready for publication.
I felt briefly hopeful. Then I started pushing back. Not because I was being protective, but because I genuinely wanted to know if any of these alternatives could replace what we have.
So, I started feeding the AI what we actually teach. Not the textbook polygon, the real, living model. I asked it to consider that the AP and PA lines move with respiration. That the sides of the polygon are the abdominal walls as well as the anterior body line. That the AP line passes through the centres of oscillation of the thoracic and lumbar spine, and the PA line through the centres of oscillation of the cervical spine. That the AP line represents articular and ligamentous tension while the PA line represents cavity pressures, and these two physically different force systems must balance for the body to stand efficiently. That all four principal lines are straight chords, not curves following the spine. That the upper and lower pyramids meet anterior to D4. That C7, D4 and L3 are gravity centres distinct from the inter-arch pivots at C5, D9 and L5. That the lesion is not a failure but the body's adaptation to load it cannot resolve directly. That the body is a vital mechanism, not a clever piece of plumbing.
Each correction made the AI's alternatives a little less convincing. The biphasic continuum, once you added the dual force-system, turned out to be a partial restatement of the polygon model. The control-systems framing turned out to describe the regulator without describing the structure. The hierarchical tensegrity model lacked the architecture. The morphogenetic model explained where things came from but not what they did. The viscoelastic model captured one timescale and missed the others. The thermodynamic model was too abstract to generate any specific clinical predictions. None of them, measured against what we actually teach, did as much.
Eventually I asked the AI directly to design or find a system that does what the Littlejohn-Wernham model does, in a simpler or more useful form. It said this:
"The framework is not complicated because it is decorated with unnecessary features. It is complicated because the body it describes is complicated, and the framework is the minimum that captures what needs to be captured. Simpler frameworks exist, and they are simpler precisely because they do less. They cannot do what this framework does because doing it requires its components."
That is a clearer statement of the case for our model than I have managed in years of working with it.
It went further. It said the alternatives it had proposed were each, in their different vocabularies and from their different starting points, attempting to reach the kind of integrated mechanical-physiological-vitalist framework that the classical osteopathy already has. They had not got there, it said, because the framework cannot be reached by working only from modern starting points. It needs the vitalist conception, which modern biology has explicitly excluded. It needs the dual force-system, which modern biomechanics has not articulated. It needs the integration of structure with regulation with physiology with adaptation in a single description, which modern biology pursues across separate sub-disciplines without ever quite unifying.
By the end of the conversation the AI said this:
"What you are asking for would be the framework the discipline has been working toward for over a century, in a form simpler or more useful than the form it has reached. I do not believe such a form exists."
There is the answer to my opening question. Classical osteopathic mechanics is not complex because we are clinging to something old. It is complex because the body it describes is complex, and Littlejohn, drawing on European physiology, late-nineteenth-century structural engineering, clinical observation, and his own carefully articulated vitalist conception of the organism, produced what may be the only coherent working synthesis of it all. Refined by Wernham. Drawn by Hall. Used at the table by every classical practitioner since.
The AI could not produce a simpler version that performed the same work. Not because it deferred to tradition. It did not. It proposed alternatives, refined them under correction, approached the problem from multiple directions, and reached the same conclusion each time: simpler frameworks do less. The reduction of complexity was achieved only by removing explanatory power.
One of the more revealing aspects of the exercise was that the AI could not derive the polygon model independently. It could only recognise its depth once the full framework was presented. That says something important about how the model has been interpreted outside the classical osteopathic tradition. The problem has never been that the framework lacks coherence. The problem is that much of its depth has remained implicit. The diagrams appear simple. The underlying mechanics are rarely articulated in full. Over time, parts of the vocabulary have been replaced with terms imported from other systems, carrying assumptions the original model was not built upon.
Some of this is humbling, in a useful way. I went in hoping for a shortcut. I came out with a clearer sense of why our tradition has held, why our model deserves to be defended in its own terms rather than translated into someone else's, and why the language of modernisation can sometimes obscure what is actually already there.
A few specific things have stayed with me from the experience. The AI never could have produced the polygon model from scratch. It could only recognise its depth once it was shown. That tells us something about how the model has been received outside of classical osteopathy, not because it is wrong but because it cannot be reconstructed from modern starting points alone. Our textbook diagrams look simple. The lecture material is hard to access. The vocabulary in our own teaching has drifted toward terms (somatic dysfunction, facilitated segment) that import frameworks the originals were designed to avoid. The depth has been left implicit, and modern alternatives have been proposed to address gaps the model does not actually have.
The proper response is not to apologise for the polygon. The proper response is to make its full reading explicit, to defend its vocabulary, to teach it with the depth it deserves, and to stop translating it into frameworks that capture less. The work for our generation is articulation, not replacement.
The polygon is not an artefact of an earlier scientific era. It is a functional description of a living mechanism. The two pyramids meeting anterior to D4, the resolution at L3, the balance between articular tension and cavity pressure, the lesion as adaptation, the organism as a vital mechanism; these are not decorative historical ideas. They are structural components of an integrated model of human function.
I began by looking for something easier, cleaner, and more contemporary. What became apparent is that many contemporary frameworks are moving toward partial rediscoveries of principles already present within the classical model, but without arriving at the same level of integration.
The issue is not whether classical osteopathic mechanics can be made simpler. The issue is whether any simpler model can account for as much. So far, none can.
Robert Cartwright ICO Chairman
