Appendix D · A Formal Model of Orbital Dynamics

This appendix develops, at a level a reader with some exposure to dynamical systems can follow, the formal model sketched in The Formal Turn. It is written entirely in the fifth register: everything here is a proposed model, not a validated one. Its value is that it is precise enough to be wrong, and it names, at the end, the structural predictions by which it could be. The model does not invent its mathematics; it applies the established formalism of coupled nonlinear systems — the same used in computational neuroscience, mathematical ecology, and the network theory of psychopathology — to the specific architecture this book has proposed. The synthesis is the application, not the apparatus.

D.1 State, Regulation, and Coupling

Represent the momentary condition of a self as a state vector x(t) = (x₁, …, xₙ), one coordinate per function (Sun through Planet Nine, plus the somatic and lunar variables). Each xᵢ is the current activation of that function, scaled so that sᵢ denotes its resting set-point. The governing equation for each function is a leaky, coupled, driven integrator:

dxᵢ/dt = −αᵢ(xᵢ − sᵢ) + Σⱼ wᵢⱼ φ(xⱼ) + Iᵢ(t) + ξᵢ(t)

Here αᵢ \> 0 is the self-regulation (leak) rate, returning the function toward its set-point; W = \[wᵢⱼ\] is the coupling matrix, with wᵢⱼ \> 0 for facilitation and \< 0 for inhibition; φ is a bounded, monotonically increasing gain function (a sigmoid, e.g. φ(u) = tanh(u) or the logistic), which prevents any influence from growing without limit and gives the system its essential nonlinearity; Iᵢ(t) is exogenous input (circumstance, task, another person); and ξᵢ(t) is stochastic drive (noise). This is the canonical form of a rate model — formally identical to a Wilson–Cowan neural-population network, a generalized Lotka–Volterra ecosystem, or a linked-sector macro model — and adopting it is the model's first commitment: the self belongs to the same formal class as those systems, and inherits their mathematics.

Two structural claims follow immediately. First, the diagonal terms αᵢ formalize what the book called self-regulation: a function with healthy α returns to baseline after perturbation, one with α → 0 drifts or latches. Second, the off-diagonal structure of W — the pattern of who inhibits and who facilitates whom — is the person. Two individuals with identical function-repertoires but different coupling matrices are different selves, which is precisely the book's thesis that identity lives in the relationships between functions, not in the functions themselves, now stated as a property of a matrix rather than a sentiment.

D.2 Health as Attractor Landscape

A nonlinear system of this form does not have "a state"; it has a landscape of attractors — configurations toward which nearby trajectories converge — separated by basins. Formally, the long-run behavior is organized by the fixed points x\ where dx/dt = 0 and by their stability, given by the eigenvalues of the Jacobian J = ∂(dx/dt)/∂x evaluated at x\: a fixed point is stable when every eigenvalue has negative real part.

This yields a precise restatement of psychological health, and it is not the absence of conflict. A healthy self is one whose attractor landscape is metastable — populated by several functional attractors that are each stable enough to hold a shape against ordinary noise, yet shallow enough that the system can transit between them as circumstances demand. Rigidity is a landscape with one attractor so deep the system cannot leave it (the perfectionist who cannot loosen, the depressive locked in a single basin). Chaos is a landscape too shallow to hold any shape (the ungoverned self of the introduction, driven by whichever input is momentarily largest). The good life, formally, is not a fixed point. It is a well-shaped landscape — and this is why the book has insisted throughout that the aim is not to reach a settled state but to become the kind of system that can hold, and change, its states.

D.3 The Five Collisions as Dynamical Motifs

The collisions of Part IV, informal there, are here specific low-dimensional motifs with characteristic dynamics. Each is a reduction of the full system to the two (or three) functions that dominate it, holding the rest as slowly varying context.

Mobilization vs. constraint (Mars–Saturn): the mutual-inhibition deadlock. Let M and S be mobilization and constraint, with action output A = φ(M − θS). Take

dM/dt = −αM + a·drive − b·φ(S),   dS/dt = −αS + c·threat − d·φ(M)

with mutual inhibition b, d \> 0. When drive and threat are both high (the situation matters and is risky), the stable fixed point has M and S both large while A ≈ 0: maximal opposed activation with cancelled output. This is the deadlock, derived rather than described — the engine and the brake both flooring, net motion zero. The subjective correlates the chapter named (dread, bracing, exhausting stasis) are the felt side of a system burning energy at a fixed point that produces nothing.

Expansion vs. structure (Jupiter–Saturn): the density fold. Let expansion E (commitments, vision, reach) be a fast variable and density D (competence, integrated substance) a slow one that accumulates only under sustained pressure P: dD/dt = k P(Dₘₐₓ − D) − δD. Fragility is F = E − D. When E outruns D (inflation), F crosses a threshold at which the system's load exceeds its structural support and undergoes a fold — a discontinuous collapse. Because D is slow, recovery is slow: the hallmark hysteresis of a bubble.

Disruption vs. structure (Uranus–Saturn): the relaxation oscillator. Let charge Q (the disruptive potential the book called voltage) accumulate, dQ/dt = r − ℓQ, where r is the charging rate and ℓ a dissipative leak. With no legal channel (ℓ ≈ 0), Q rises to a critical Q\_crit and discharges catastrophically, then resets — an integrate-and-fire, or relaxation-oscillator, dynamic. The eruption "from nowhere" is formally a threshold crossing in a slowly charging variable with insufficient leak; the intervention "upgrade the wire" is, exactly, increasing ℓ.

Imagination vs. articulation (Neptune–Mercury): destructive interference. Let the generative process G and the articulating process C share a channel whose throughput is T = C / (1 + κ·G·C): when both run at once, the cross-term κGC destroys throughput (the fog). Sequencing — time-multiplexing G and C so they are not simultaneously large — removes the interference term. The collision is not a shortage of either capacity but a scheduling failure, and its fix is a scheduling discipline.

Output vs. replenishment (Sun–Moon): the depleting reservoir with a fold. Let reservoir R deplete with output O and refill only when output is low and the receptivity gate γ is open: dR/dt = −O + ρ·(1 − O/Oₘₐₓ)·γ. Sustained high O drives R toward zero; below a critical reserve the system's output-generating capacity itself fails, a fold, after which recovery is slow and requires O ≈ 0 with γ genuinely open — the reason, in the model, that "rest" while still monitoring (γ closed) does not refill the reservoir.

D.4 Stress as Control Parameter; Collapse as Bifurcation

The motifs share a deeper unity. In each, a slowly changing quantity — load, stress, sustained demand — acts as a control parameter that moves the system through a bifurcation: a qualitative change in the attractor landscape. Rising coupling gain tips the mobilization–constraint system from a monostable, mobile regime into the bistable regime that contains the deadlock fixed point; rising E/D ratio carries the expansion system over a fold; rising charge with low leak drives the disruption system into a limit cycle. This is precisely the structure catastrophe theory (Zeeman) and, more recently, the network theory of psychopathology (Borsboom, Cramer) propose for mental disorder: symptoms as nodes in a dynamical network whose collective state can undergo sudden transitions between a healthy and a disordered attractor as a control parameter drifts.

That connection yields the model's most important and most testable structural claim. Systems approaching such a bifurcation exhibit critical slowing down: as the landscape flattens near the transition, the system recovers more sluggishly from small perturbations, and its fluctuations show rising autocorrelation and variance. Scheffer, van de Leemput, and colleagues have demonstrated these early-warning signals empirically in the onset of depression. The orbital model predicts that the same statistical signatures should precede each of the collision-collapses — not only depressive transitions but burnout, the disruptive eruption, the inflationary crash — because they are all bifurcations of the same formal type. This is a strong, unifying, falsifiable prediction, and it is developed in Appendix F.

D.5 Interventions as Principled Perturbations

The book's practices, intuitive there, here become specific operations on the equations — which is one test of whether the formalism is faithful, since a good model should reconstruct the interventions its informal version discovered.

"Shorten the barrel" (the compound action) is a reduction of the action threshold θ, or equivalently a reduction of the activation-energy barrier ΔV separating the deadlocked fixed point from the mobile one; in Kramers' rate terms the escape probability rises exponentially as the barrier falls, so shrinking the unit of action produces a disproportionate increase in the chance of initiation. "Upgrade the wire" is an increase in the leak ℓ of the disruption motif, converting a relaxation oscillator that must discharge into a system that dissipates continuously. "The third body" is the addition of a variable that changes the topology of the fixed-point set — introducing the cost-of-stasis term shifts a symmetric double well into an asymmetric one with a single global minimum, so the deadlock ceases to be a stable state at all. "Behavioral activation" (act first, motivation follows) is a small constant drive that tilts the landscape enough for output to begin, after which the dynamics themselves carry the system — formally, action changes the state, and the changed state changes the very drive that action was waiting on. That each informal intervention maps to a clean, principled perturbation is not proof the model is correct. It is evidence the model is faithful — that it captures the same structure the practices were blindly exploiting.

D.6 The Normative Layer: Why These Functions

A remaining question is why the self should be organized into these functions at all, rather than some other decomposition. The model's proposed answer draws on the free-energy / active-inference framework (Friston): an adaptive system persists by minimizing, over time, the discrepancy between its predictions and its sensory evidence — equivalently, by keeping itself within the states it expects to occupy. Under that imperative, a small set of sub-problems is generic to any embodied agent: maintain a boundary and a viable internal milieu (Earth/Saturn), allocate limited processing (Mercury/Sun), value and approach (Venus), mobilize against resistance and defend the boundary (Mars), coordinate multi-scale action over time (Jupiter), revise the generative model when prediction error is structural (Uranus), relax boundaries to integrate wide context (Neptune), reprocess stored high-precision errors (Pluto), and orient toward distal, not-yet-observed goals (Planet Nine). The claim — conjectural, and offered as such — is that the book's functional decomposition is not arbitrary but approximates the natural factorization of the control problem any persistent embodied agent must solve, which is why recognizable versions of these functions recur across very different minds and, per Appendix E, across very different scales of system.

D.7 Status, Limits, and How the Model Could Fail

To keep the register honest, the model's liabilities stated plainly. It is low-dimensional by design, and real nervous systems are not; the reduction to a handful of coupled functions is a modeling choice whose adequacy is itself an empirical question. The coupling matrix W is, at present, schematic — the model specifies its form (which functions inhibit which) far more confidently than its parameters, which would have to be estimated from data. The mapping from these abstract variables to measurable quantities (which physiological or behavioral signal indexes "constraint"?) is proposed, not established. And the free-energy grounding of D.6 is a live and contested research frame, borrowed as scaffolding, not as settled fact.

The model would be shown wrong, or badly incomplete, by any of the following: if the collision states carry no distinct dynamical signatures and are behaviorally indistinguishable from simple deficits; if the predicted critical-slowing-down markers fail to precede the non-depressive collapses; if the intervention-as-perturbation mappings do not produce the effects the equations imply (for instance, if shrinking the action quantum does not preferentially help high-barrier deadlock over low-arousal apathy); or if attempts to fit the coupling structure to real longitudinal data find no stable, person-specific W at all. These are not rhetorical concessions. They are the specific ways this model has agreed to be defeated, and a model that has named them has done the one thing a metaphor never can.