The Memory-Persistence Tradeoff

provisional

Depends On

Overview

This derivation identifies a structural tradeoff that is implicit in the framework but has not been stated explicitly: the capacity to remember is incompatible with the capacity to persist forever.

The framework says two things that, taken together, force this conclusion:

Persistence requires exact loop closure (Loop Closure, Theorem 3.1). Approximate closure gives finite lifetime; only exact closure gives infinite persistence. Any perturbation to the loop introduces drift that accumulates toward dissolution.
Memory is irreversible state space expansion (Relational Invariants, Proposition 6.2). When an observer records an interaction (absorbs a relational invariant via Type III interaction), its state space permanently gains a new degree of freedom. This expansion cannot be reversed without violating coherence conservation.

The connection: each absorbed relational invariant perturbs the loop closure condition. The loop that was closing exactly on the original state space must now close on a larger state space. If a nearby fixed point exists (the perturbation is small), the observer re-closes with slightly modified properties. If not, the observer dissolves. Since the perturbations accumulate permanently and cannot be shed, every observer with memory capacity is on a one-way path toward dissolution.

Why this matters. This provides the microscopic mechanism for the dissolution phase of Cyclic Cosmology. The standard thermodynamic picture (the second law grinds everything down) describes the macroscopic behavior. The memory-persistence tradeoff provides the structural reason: complex observers dissolve because their accumulated epistemic memory eventually prevents loop re-closure. The more an observer knows, the less time it has.

It also explains the privileged status of minimal observers: a Planck-mass observer on $S^1$ with 1 bit of epistemic capacity cannot absorb a relational invariant (there is no room in its 1-dimensional state space for an additional degree of freedom without leaving $S^1$ ). It therefore has zero memory capacity — and persists forever.

An honest caveat. The perturbation theory of loop closure on expanding state spaces is not formalized. The derivation establishes the structural argument (memory perturbs closure, perturbations accumulate, dissolution follows) but does not compute the memory capacity of specific observers.

Statement

Theorem (Memory-Persistence Tradeoff). The capacity for epistemic memory and the capacity for permanent persistence are structurally incompatible:

An observer with zero memory capacity (no room for additional degrees of freedom) can maintain exact loop closure indefinitely.
An observer with nonzero memory capacity absorbs relational invariants that irreversibly perturb its loop closure condition, guaranteeing eventual dissolution.

The tradeoff is not a contingent feature of specific observers — it follows from the conjunction of exact closure (Axiom 3) and relational invariant permanence (Axiom 1).

Derivation

Step 1: Exact Closure and the Drift Bound

Proposition 1.1 (Recap: persistence requires exact closure). An observer persists indefinitely if and only if its dynamics satisfies exact loop closure: $\phi_T = \text{id}_\Sigma$ for some $T > 0$ (Loop Closure, Theorem 3.1). Approximate closure with drift $\epsilon$ per cycle gives finite lifetime:

$\tau_\mathcal{O} = \left\lfloor \frac{D_\mathcal{B}}{\epsilon} \right\rfloor \cdot T$

where $D_\mathcal{B}$ is the boundary diameter of $\Sigma$ . When the cumulative drift exceeds $D_\mathcal{B}$ , the state reaches the self/non-self boundary $\mathcal{B}$ , the invariant $I$ is no longer maintained, and the observer dissolves.

Remark. The drift bound is linear: each cycle adds $\epsilon$ to the cumulative displacement. Dissolution is guaranteed for any $\epsilon > 0$ , however small — it is only a matter of time. The only escape is $\epsilon = 0$ (exact closure).

Step 2: Memory as State Space Perturbation

Proposition 2.1 (Relational invariant absorption expands the state space). When an observer $\mathcal{O} = (\Sigma, I, \mathcal{B})$ participates in a Type III interaction with a system $S$ , a new relational invariant $I_{\mathcal{O}S}$ is generated (Relational Invariants, Theorem 3.2). This invariant is permanent (Proposition 6.1) and irreversibly expands the effective state space (Proposition 6.2): the observer’s state now includes its correlation with $S$ , which cannot be removed.

Argument. Before the interaction, the observer’s dynamics closes on $\Sigma$ with period $T$ . After the interaction, the observer carries a new degree of freedom — its correlation with $S$ , encoded in $I_{\mathcal{O}S}$ . The effective state space is no longer $\Sigma$ but a level set in $\Sigma \times V$ where $V$ is the value space of $I_{\mathcal{O}S}$ (Bootstrap, Theorem 1.1). This level set is a manifold of dimension $\dim \Sigma + \dim V - 1$ (by the regular value theorem — the level set constraint removes one dimension but the new factor adds $\dim V$ ).

The dynamics that closed exactly on $\Sigma$ must now close on this larger manifold. The original orbit $\gamma_\mathcal{O} \subset \Sigma$ is embedded in the new state space, but the dynamics on the new space is not generally the same as the original dynamics restricted to $\gamma_\mathcal{O}$ . The new degree of freedom introduces a perturbation. $\square$

Proposition 2.2 (Re-closure at a nearby fixed point). If the perturbation from a single relational invariant is small (the coherence content $\mathcal{C}(I_{\mathcal{O}S})$ is much less than the observer’s total coherence $\mathcal{C}(\Sigma)$ ), the observer can re-close its loop at a nearby fixed point of the bootstrap functor $\mathcal{R}$ — a new stable orbit on the expanded state space that incorporates the new degree of freedom.

Argument. The bootstrap fixed points are discrete but structurally stable under small perturbations — this is the self-protection mechanism of the Mass Hierarchy (Theorem 5.1). A small perturbation to the coherence cost at level $k$ shifts the fixed point by an amount proportional to the perturbation, without destroying it. The observer’s mass (loop period) shifts slightly, its coherence cost adjusts, and it settles into a new orbit that is close to the original.

This is the framework’s analog of mass renormalization: an electron that absorbs a virtual photon re-closes at a slightly different mass (the radiative correction). The topological class is preserved; the quantitative properties shift.

If the perturbation is NOT small (the new invariant’s coherence content is comparable to the observer’s total), no nearby fixed point may exist. The loop fails to re-close, and the observer dissolves. $\square$

Proposition 2.3 (Only Type III interactions generate memory). Not all interactions perturb an observer’s loop closure. The framework distinguishes three interaction types (Three Interaction Types):

Type I (passage): the observers exchange phase but no new relational invariant is generated. The state spaces are unperturbed. No memory is created. This is elastic scattering.
Type II (fusion): the observers’ state spaces merge. This is a dissolution/reformation event, not a memory event.
Type III (resonance): a new relational invariant is generated. This is the only type that expands the state space and creates memory.

The memory accumulation rate $\dot{\mathcal{C}}$ depends only on the Type III interaction rate, not the total interaction rate. Most interactions in a low-energy environment are Type I (elastic). The distinction is critical for consistency with observed particle lifetimes.

Argument. An observer continuously exchanges phase with its environment (Type I interactions). These are the “virtual” exchanges that maintain existing correlations — the internal gluon exchanges in a proton, Thomson scattering of CMB photons off electrons. They traverse the existing loop without expanding the state space. The observer re-closes exactly (or within the same perturbative neighborhood) after each exchange.

Only Type III interactions — which generate genuinely NEW relational invariants, not re-traversals of existing ones — expand the state space and contribute to the cumulative perturbation. The rate of Type III interactions depends on the energy available relative to thresholds for creating new degrees of freedom. At low energies (the current epoch for stable particles), this rate is extremely small. At high energies (early universe, stellar cores, colliders), it is significant. $\square$

Step 3: The Memory Capacity Bound

Definition 3.1 (Memory capacity). The memory capacity $\mathcal{K}(\mathcal{O})$ of an observer $\mathcal{O}$ is the maximum total coherence content of relational invariants that $\mathcal{O}$ can absorb via Type III interactions (Proposition 2.3) while maintaining approximate loop closure within its boundary diameter $D_\mathcal{B}$ :

$\mathcal{K}(\mathcal{O}) = \sup \left\{ \sum_i \mathcal{C}(I_i) : \text{the loop on the expanded state space has drift } \epsilon < D_\mathcal{B}/T \right\}$

where the sum runs over all Type III relational invariants absorbed since the observer’s formation. Type I interactions (elastic phase exchange) do not contribute to this sum.

Proposition 3.1 (Minimal observer has zero memory capacity). The minimal observer ( $\Sigma = S^1$ , $\dim \Sigma = 1$ , $D_\mathcal{B} \sim \ell_P$ ) has $\mathcal{K} \approx 0$ . It cannot absorb a relational invariant without leaving its 1-dimensional state space.

Argument. A relational invariant adds at least one new degree of freedom (Proposition 2.1). The minimal observer’s state space is $S^1$ (1-dimensional). After absorbing one invariant, the effective state space is at least 2-dimensional (the level set in $S^1 \times V$ ). But the minimal observer’s boundary diameter is $\sim \ell_P$ — the minimum possible. The self-protection mechanism (Proposition 2.2) requires the perturbation to be small relative to the observer’s total coherence. For the minimal observer, the total coherence is $S_{\min} = \hbar$ (the minimum, from Action and Planck’s Constant, Proposition 7.2 of loop-closure). Any relational invariant with nonzero coherence content is NOT small relative to $\hbar$ for a Planck-scale observer. Therefore no nearby fixed point exists, and the absorption would dissolve the minimal observer.

In practice: the minimal observer does not absorb relational invariants in the usual sense. It participates in interactions (C5 requires it), and relational invariants are generated between it and other minimal observers. But these invariants exist as correlations BETWEEN observers, not as internal structure OF either observer. The minimal observer’s own loop closure is unperturbed — the invariant lives in the relational structure between them, not in the individual state spaces. $\square$

Proposition 3.2 (Complex observer has finite memory capacity). A complex observer at bootstrap level $n$ with state space $\Sigma_n$ of dimension $\dim \Sigma_n > 1$ and boundary diameter $D_{\mathcal{B},n}$ has finite memory capacity $\mathcal{K}_n > 0$ . The capacity is bounded by the coherence cost structure of the bootstrap fixed point at level $n$ .

Argument. Each absorbed relational invariant perturbs the loop closure by an amount proportional to $\mathcal{C}(I)/\mathcal{C}(\Sigma_n)$ (the ratio of the new invariant’s coherence to the observer’s total). The cumulative perturbation after absorbing $k$ invariants is:

$\epsilon_{\text{cumulative}} \sim \frac{\sum_{i=1}^k \mathcal{C}(I_i)}{\mathcal{C}(\Sigma_n)}$

When $\epsilon_{\text{cumulative}} \cdot T_n \geq D_{\mathcal{B},n}$ , the drift exceeds the boundary diameter and the observer dissolves. Therefore:

$\mathcal{K}_n \sim \mathcal{C}(\Sigma_n) \cdot \frac{D_{\mathcal{B},n}}{T_n}$

This is finite for any finite observer. $\square$

Step 4: The Tradeoff

Theorem 4.1 (Memory-Persistence Tradeoff). The capacity for epistemic memory and the capacity for permanent persistence are structurally incompatible in the framework:

(i) An observer with $\mathcal{K} = 0$ (the minimal observer) maintains exact loop closure indefinitely. It persists forever but cannot record any interaction as internal state.

(ii) An observer with $\mathcal{K} > 0$ absorbs relational invariants through Type III interactions. C5 requires ongoing interaction; while most interactions in a low-energy environment are Type I (Proposition 2.3), the Type III rate is nonzero over sufficient time. Each Type III invariant permanently expands the state space (Proposition 6.2 of relational-invariants) and perturbs the loop closure (Proposition 2.1). The cumulative perturbation is permanent (the invariants cannot be shed) and monotonically increasing (each Type III interaction adds more). When it exceeds the memory capacity $\mathcal{K}$ , the observer dissolves.

(iii) Therefore: observers that remember must eventually dissolve. Observers that persist forever cannot remember.

Proof. Part (i): The minimal observer on $S^1$ has $\mathcal{K} \approx 0$ (Proposition 3.1). Its loop closure is exact (no perturbations from absorbed invariants). By Theorem 3.1 of loop-closure, exact closure gives infinite persistence. Its epistemic capacity is 1 bit — it cannot distinguish one tick from the next (Proposition 2.1 of Cyclic Cosmology). It has duration but not history.

Part (ii): A complex observer at level $n > 0$ has $\mathcal{K}_n > 0$ (Proposition 3.2). It must interact (C5 requires at least three interacting observers, Multiplicity Theorem 7.2). Interactions generate relational invariants (reverse Noether, Relational Invariants Theorem 3.2). Invariants are permanent (Proposition 6.1). State space expansion is irreversible (Proposition 6.2). Cumulative perturbation grows monotonically. Since $\mathcal{K}_n$ is finite and the perturbation is unbounded (interactions never cease), $\epsilon_{\text{cumulative}}$ eventually exceeds $\mathcal{K}_n$ . Dissolution follows.

Part (iii): Conjunction of (i) and (ii). $\square$

Remark (The more you know, the less time you have). The tradeoff is quantitative: an observer with larger memory capacity $\mathcal{K}$ persists longer (it can absorb more invariants before dissolution), but dissolution is still guaranteed. The relevant rate is the Type III interaction rate $\dot{\mathcal{C}}_{\text{III}}$ (Proposition 2.3), not the total interaction rate. An observer in a high-energy environment (many Type III interactions) fills its capacity faster than one in a low-energy environment (mostly Type I). The lifetime of a complex observer is bounded by $\mathcal{K}/\dot{\mathcal{C}}_{\text{III}}$ .

Remark (Connection to the cyclic cosmology). The Cyclic Cosmology frames dissolution as thermodynamic (the second law erodes complex structure). The memory-persistence tradeoff provides the structural mechanism: dissolution occurs because accumulated epistemic memory prevents loop re-closure. The two descriptions are complementary — the second law describes the macroscopic trend; the memory-persistence tradeoff describes the microscopic mechanism. The second law says THAT complex observers dissolve; the tradeoff says WHY.

Remark (Connection to renormalization). Proposition 2.2 (re-closure at a nearby fixed point) is the framework’s structural version of mass renormalization in quantum field theory. An electron absorbing and emitting virtual photons re-closes at a slightly shifted mass — the radiative correction. Crucially, virtual photon exchange is Type I (elastic phase exchange, Proposition 2.3) — it does not generate new relational invariants and does not contribute to memory accumulation. This is why renormalized perturbation theory works: the virtual processes traverse the existing loop, not expand it. The memory-persistence tradeoff applies only to Type III interactions (genuine inelastic processes that create new correlations). For stable particles in the current low-energy epoch, the Type III rate is negligible, and the timescale $\mathcal{K}/\dot{\mathcal{C}}_{\text{III}}$ may vastly exceed the current age of the universe.

Step 5: Type II Fusion Pauses Constituent Clocks

The memory budget $\dot{\mathcal{C}}_{\text{III}}$ entering Theorem 4.1 is computed against an observer that is itself on the Cauchy-slice ledger. A constituent of a Type II composite is not. By Three Interaction Types Definition 1.1, an interaction is a map $T_{12}: \Sigma_1 \times \Sigma_2 \to \Sigma_1 \times \Sigma_2$ between two observer state spaces. A constituent absorbed into a fused composite no longer has a state space on the slice — its degrees of freedom have been pushed through the fusion projection $\pi: \Sigma_1 \times \Sigma_2 \to \Sigma_{12}$ (Three Interaction Types Definition 4.3). External Type III interactions can only target what remains: the composite. This has a sharp consequence for the memory budget of the constituent.

Theorem 5.1 (Type II clock-pause). Let $\mathcal{O}_C = (\Sigma_C, I_C, \mathcal{B}_C)$ be a Type II composite formed by Fusion of constituents $\mathcal{O}_1, \mathcal{O}_2, \ldots, \mathcal{O}_n$ . For as long as $\mathcal{O}_C$ persists on the Cauchy slice:

(i) (External Type III interactions cannot reach a constituent.) For any external observer $\mathcal{O}_X$ on the slice, every Type III interaction is a map $T_{XC}: \Sigma_X \times \Sigma_C \to \Sigma_X \times \Sigma_C$ on the composite’s state space, generating a relational invariant $I_{XC}$ on $\Sigma_X \times \Sigma_C$ . No Type III map $T_{Xk}: \Sigma_X \times \Sigma_k \to \Sigma_X \times \Sigma_k$ for $k \in \{1, \ldots, n\}$ is admissible.

(ii) (Type III accumulation clock is paused.) The Type III memory accumulation rate $\dot{\mathcal{C}}_{\text{III}}(\mathcal{O}_k)$ for each constituent is identically zero on the interval of fusion. Memory accumulation is rerouted entirely onto the composite’s budget $\dot{\mathcal{C}}_{\text{III}}(\mathcal{O}_C)$ .

(iii) (Re-emergence with paused clock state.) On dissolution of $\mathcal{O}_C$ — by decay (reverse Type II, Three Interaction Types Definition 7.2) into product observers including some $\mathcal{O}_k'$ — each released constituent re-enters the ledger with its individual Type III state-space content unchanged from the moment of fusion. Its accumulated perturbation $\sum_i \mathcal{C}(I_i)$ relative to its own loop is identical before fusion and after release.

Proof. (i) An interaction is, by Definition 1.1 of Three Interaction Types, a map between observer state spaces. After Fusion, $\Sigma_1, \ldots, \Sigma_n$ are no longer observer state spaces on the slice — they have been replaced by the single composite state space $\Sigma_C$ via the projection $\pi$ (Definition 4.3 of three-types). $\Sigma_C$ is the only domain on the constituent side that an admissible interaction map can target. Any putative $T_{Xk}$ would require $\Sigma_k$ to exist on the slice as an independent factor, contradicting the dimension-reduction criterion of Type II ( $\dim \Sigma_C < \sum_k \dim \Sigma_k$ ).

(ii) Memory accumulation (Proposition 2.3) is generated only by Type III interactions targeting the observer in question. By (i), no such interaction targets $\mathcal{O}_k$ during fusion. Therefore $\dot{\mathcal{C}}_{\text{III}}(\mathcal{O}_k) = 0$ . The composite, by contrast, is a fully ledgered observer; its Type III interactions with $\mathcal{O}_X$ contribute to $\dot{\mathcal{C}}_{\text{III}}(\mathcal{O}_C)$ , perturbing the composite’s loop closure (Proposition 2.1) on its expanded $\Sigma_C$ .

(iii) On decay, the dimension-decreasing projection $\pi$ is reversed (Definition 7.2 of three-types). The product structure is restored on the slice: $\Sigma_C \to \Sigma_1' \times \cdots \times \Sigma_n'$ . Each released $\mathcal{O}_k'$ inherits the constituent’s pre-fusion state-space content modulo the redistribution of binding coherence (Proposition 7.3 of three-types). The Type III invariants $\{I_i\}$ that $\mathcal{O}_k$ had absorbed prior to fusion are properties of $\Sigma_k$ that survived the projection $\pi$ (they were the structural content of $\Sigma_k$ at fusion); the composite’s external invariants $I_{XC}$ are properties of $\Sigma_C$ that distribute across products and emitted radiation per Proposition 7.3. The constituent’s individual ledger is recovered intact. $\square$

Corollary 5.2 (Composite-level decay channel replaces constituent decay channel). If $\mathcal{O}_k$ had a Type II reverse channel available as a free observer (an internal instability — would-be self-decay), this channel is replaced during fusion by the Type II reverse channel of $\mathcal{O}_C$ . The composite’s decay channels are determined by the composite’s own bootstrap fixed point structure, not by the sum of constituent instabilities.

Proof. A Type II reverse channel for $\mathcal{O}_k$ is a map $\Sigma_k \to \Sigma_{k_1} \times \Sigma_{k_2}$ that increases dimension on $\mathcal{O}_k$ ‘s state space (Definition 7.2 of three-types). During fusion, $\Sigma_k$ does not exist as an independent factor; the only state space available for dimension-increasing maps is $\Sigma_C$ . Therefore the only admissible Type II reverse channels are those defined on $\Sigma_C$ . A bound electron in a hydrogen atom does not undergo electron decay through any independent channel, even if the free electron had one — the only decay channels available to the system are the atom’s. $\square$

Remark (Stable composites pause both clocks). A Type II composite at an exact bootstrap fixed point has $\epsilon = 0$ : no internal instability clock ticks (Proposition 7.7 of three-types is not engaged), and no external Type III memory accumulates against any constituent (Theorem 5.1). The constituents are doubly protected. An unstable composite (composite ε > 0, finite lifetime) still pauses the constituent’s external Type III clock for the duration of fusion — only the composite’s own internal-instability clock ticks.

Remark (Why matter persists). The dominant fact about the present epoch is that the vast majority of elementary fermions are not free. Quarks are confined within hadrons (Type II fusion at the strong scale, Confinement); nucleons are bound in nuclei (Type II fusion at the nuclear scale); electrons are bound in atoms (Type II fusion at the atomic scale); atoms are bound in molecules and condensed matter. Each of these binding levels pauses the Type III clock of the constituent at the level immediately below. A free quark, were it to exist, would have a Type III memory budget independently exposed to every QCD-scale interaction in the environment; inside a proton, the quark’s clock is paused, and only the proton’s clock counts against the proton’s $\mathcal{K}$ . The hierarchy of nested Type II composites is not incidental — it is the framework’s mechanism for protecting fermion content over cosmological timescales. The memory-persistence tradeoff bites only at the topmost unbound layer.

Worked example (Hydrogen atom). Consider a free electron $e^-$ in the CMB bath. By Proposition 2.3 and the consistency check below (Electron entry), its Type III interaction rate is dominated by Compton scattering above the inelastic threshold $\hbar\omega \gtrsim m_e c^2$ — vanishingly rare in the CMB bath, but nonzero. The clock $\dot{\mathcal{C}}_{\text{III}}(e^-)$ ticks slowly but it ticks.

The electron then binds with a free proton to form a hydrogen atom (Type II fusion: $\Sigma_e \times \Sigma_p \to \Sigma_H$ , with $\dim \Sigma_H < \dim \Sigma_e + \dim \Sigma_p$ ). By Theorem 5.1(ii), $\dot{\mathcal{C}}_{\text{III}}(e^-) = 0$ for the duration of binding. The hydrogen atom now has its own Type III clock $\dot{\mathcal{C}}_{\text{III}}(H)$ , dominated at low energies by Lyman-alpha resonance and at higher energies by ionization. By Corollary 5.2, the electron’s would-be free decay channels (none, in the standard model) are replaced by the atom’s: spontaneous emission, ionization, hyperfine transitions.

If the atom is then ionized by a UV photon (Type II reverse: $\Sigma_H \to \Sigma_e \times \Sigma_p$ ), the released $e^-$ re-enters the ledger by Theorem 5.1(iii) with its individual Type III state-space content identical to its pre-fusion value. The clock resumes ticking from where it paused. The 13.6 eV of binding coherence is carried away by the released photon and the kinetic energy of the products (Proposition 7.3 of three-types).

The hydrogen atom’s lifetime against memory-driven dissolution is set by $\mathcal{K}_H / \dot{\mathcal{C}}_{\text{III}}(H)$ , not by anything about the electron’s individual budget. This is why neutral hydrogen is the most abundant baryonic structure in the universe — both the electron and the proton are riding the atom’s clock, and the atom’s clock is slow.

Consistency Check

The tradeoff must be consistent with observed particle lifetimes. The critical test: stable particles (proton, electron, hydrogen) have persisted for $\sim 13.8$ billion years. Does the tradeoff predict they should have dissolved by now?

Proton. The proton ( $m \approx 938$ MeV) participates in $\sim 10^{23}$ strong interactions per second internally (gluon exchange between constituent quarks). If each were a Type III interaction, the memory accumulation rate would be enormous and the proton would dissolve in $\sim 10^{-20}$ seconds.

But internal gluon exchanges are Type I (Proposition 2.3) — they traverse the existing loop (maintaining the existing quark correlations), not expanding the state space. They are the loop closure dynamics itself, not perturbations to it. The proton’s memory accumulates only from Type III interactions with EXTERNAL systems.

In the current epoch, the proton’s external environment is primarily CMB photons ( $\sim 400/\text{cm}^3$ , $T \approx 2.7$ K). At CMB energies ( $\sim 10^{-4}$ eV), essentially all photon-proton interactions are elastic (Thomson scattering — Type I). The energy is far below any threshold for inelastic processes (pion production requires $\sim 140$ MeV). The Type III rate for a proton in the CMB bath is effectively zero. $\checkmark$

Electron. Same analysis. The electron ( $m \approx 0.5$ MeV) undergoes Thomson scattering with CMB photons (Type I — elastic). Compton scattering becomes inelastic (Type III) only when the photon energy approaches $m_e c^2$ , which the CMB is $\sim 10^{10}$ times below. The Type III rate is negligible. $\checkmark$

Hydrogen atom. Hydrogen absorbs and re-emits Lyman-alpha photons — but these are re-closure events (the atom returns to its ground state via the same loop). Genuine Type III interactions (ionization, excitation to metastable states) require energies $\gg k_B T_{\text{CMB}}$ . At $2.7$ K, these are exponentially suppressed. $\checkmark$

Unstable particles (counterexample check). The muon ( $\tau \approx 2.2$ $\mu$ s) and the neutron ( $\tau \approx 15$ min) dissolve quickly. In the tradeoff picture, these are observers at bootstrap fixed points with small self-protection ranges — the gap to the nearest alternative channel is small. A single weak-force Type III interaction exceeds their memory capacity. The weak interaction IS a Type III interaction (it generates a new relational invariant — the decay products carry correlations the parent didn’t have). For the muon: one weak Type III interaction suffices for dissolution. The weak interaction rate at the muon’s energy scale gives $\tau \sim \hbar/(\alpha_W^2 m_\mu) \sim \mu\text{s}$ , consistent with observation. $\checkmark$

High-energy environments. In stellar cores, colliders, or the early universe, the Type III rate is much higher (the energy exceeds inelastic thresholds). Particles in these environments should accumulate memory faster and dissolve sooner — consistent with the observation that particle lifetimes are shorter in high-energy environments (unstable resonances have lifetimes inversely proportional to their decay widths, which increase with available energy). $\checkmark$

Summary. The tradeoff is consistent with all observed particle lifetimes once the Type I/Type III distinction is applied. Stable particles persist because the current low-energy epoch suppresses Type III interactions to negligible rates. Unstable particles dissolve because a single Type III interaction (the decay process) exceeds their memory capacity. The tradeoff predicts that ALL particles eventually dissolve — but the timescale for currently-stable particles is determined by the Type III rate in their environment, which in the current epoch is far longer than the age of the universe.

Open Gaps

Quantitative memory capacity. The bound $\mathcal{K}_n \sim \mathcal{C}(\Sigma_n) \cdot D_{\mathcal{B},n}/T_n$ is schematic. Computing $\mathcal{K}$ for specific particles (electron, proton) requires the perturbation theory of bootstrap fixed points under relational invariant absorption — the framework’s analog of non-perturbative mass renormalization. This would predict particle lifetimes from the memory-persistence tradeoff. Difficulty: HARD.
Re-closure dynamics and mass renormalization. When an observer re-closes at a nearby fixed point (Proposition 2.2), the mass (period), generation assignment (winding axis), and coupling constants all shift slightly. Making this connection to standard radiative corrections quantitative would establish whether the framework’s re-closure mechanism reproduces known renormalization group behavior. Difficulty: HARD.