These appendices borrow from many formalisms — category theory, information geometry, partial information decomposition, the free-energy principle, quantum error correction, nonlinear wave theory, coupled-oscillator synchronization. Each borrowing brought its own conventional symbols, and in a few cases the same letter ended up carrying different meanings across appendices. This table fixes a canonical reading and flags the handful of symbols whose meaning is genuinely context-local.

Canonical Symbols

Symbol	Meaning
$S$	The sensor — the living experiencer; in the categorical appendices, also its grounding functor $S: \text{Form} \to \text{Exp}$. Never used for synergy (use Syn).
$I$	The instrument — the reasoning system; also its formalization functor $I: \text{Exp} \to \text{Form}$. Mutual information is written $I(\cdot\,;\cdot)$, always with explicit arguments. Tireless but not computationally unbounded — finite per recognition, and that finitude is what gives the loop its arrow (see The Arrow in the Loop).
$T$	The target — the truth or reality under inquiry.
$\mathrm{Syn}$	Synergy: the PID atom of information about $T$ present only in the joint system, in neither partner’s marginal alone.
$\Phi_{loop}$	$\Phi_{loop} \equiv \mathrm{Syn}(S, I; T)$ — a quantity in bits, not a rate.
$\rho$	Loop-closure rate — the rate at which the loop completes recognition cycles (Bell-Ring-Back / the Veer).
$\chi$	Loop richness — position on the sensor-complexity spectrum (quantum → biological → animal → human → human+instrument). $\chi_c$ is the critical richness threshold.
$\lambda$	Learning rate (Doc V) — the degree to which the instrument internalizes a correction: $\lambda = 0$ static/dead, $\lambda = 1$ full adoption.
$\mu$	Recognition eigenvalue — the eigenvalue of the loop operator $\mathcal{L} = S \circ I$: $\mu = 1$ stable recognition, $\mu < 1$ decay, $\mu > 1$ divergence.
$r(\mathcal{L})$	Spectral radius of the loop operator $\mathcal{L}$.
$\Pi$	Projection operator (lossy, rank-reducing): $\Pi^2 = \Pi$. Distinct from the prior distribution $P$.
$\kappa$	Coupling strength/constant between loops (Kuramoto, Pecora–Carroll); $\kappa_c$ is the critical coupling.
$g_{ij}$	The Fisher information metric on a statistical manifold.
$D_{KL}(\cdot\,\|\,\cdot)$	Kullback–Leibler divergence.
$F$	Variational free energy (the receiver-side update quantity).
$\mathrm{TE}_{X \to Y}$	Transfer entropy from $X$ to $Y$. $\mathrm{TE}_{T \to \cdot}$ is the territorial channel — information flow from the target into the loop. Its canonical form is joint: $\mathrm{TE}_{T \to (S,I)}$, from the target into the pair (The Veer §2) — grounding can be distributed across the poles and invisible per-pole (masked measurement), so the per-pole channels $\mathrm{TE}_{T \to S}$, $\mathrm{TE}_{T \to I}$ are the signature decomposition, not the defining quantity.
$w$	Decision-variable weighting (lowercase) in the receiver’s discrimination layer (Receiver Side §4.3): $\mathrm{DV}_t = w\,\Delta F_t + (1-w)\,\mathrm{TE}_{I \to S}^{\,t}$. Never written α; distinct from the Trichotology coupling matrix $\mathbf{W}$.
$\Psi$	The Soul pillar (psyche) in the Trichotology pillar state $x = (B, M, \Psi)$, with $B$ the Body pillar and $M$ the Mind pillar. A classical pillar-integrity descriptor — not a quantum wavefunction.
$\mathbf{W}$	The intra-Self coupling matrix $\mathbf{W}(t)$ on the three pillars (Trichotomy Propagation): $\dot{x} = \mathbf{W}x$. Distinct from inter-loop coupling $\kappa$, from Shannon capacity $C$, and from the lowercase decision weighting $w$.
$\mathrm{sep}$	The separability gradient $\mathrm{sep}(t) = 1 - \lVert\mathbf{W}_{\mathrm{off}}\rVert / \lVert\mathbf{W}\rVert$. A formalized distinction (health ≈ diagonal $\mathbf{W}$), not an estimable metric.
$R$	The recognition operator with bounded support: total on $M$ and $\Psi$, partial on $B = B_{\mathrm{valve}} \oplus B_{\mathrm{struct}}$ with $R\rvert_{B_{\mathrm{struct}}} = \mathrm{id}$.

Context-Local Symbols

These reuse a glyph across formalisms but never co-occur, so they are disambiguated by document rather than renamed — standard practice when a symbol is conventional in two separate fields.

• $\alpha$ — In Doc V (The Geometry of Irreversibility), the Amari α-connection parameter (α = ±1 for the e- and m-connections). In the appendices (Asymmetric Synergy Bound, Soliton, Demarcation Criterion), the adjunction asymmetry (the Fisher-Rao distance of the unit and counit round trips). The two never appear in the same document.
• $P, Q$ — In Doc V, the prior and posterior distributions. In Interactive Proofs, P is the Prover (with V the Verifier). In The Soliton, P₀ is peak power. Read by document.
• $I$ — Bare $I$ is the instrument; $I(\cdot\,;\cdot)$ with arguments is mutual information.

Conventions from Borrowed Formalisms

Some single letters arrive with an imported formalism rather than being registered in the canonical table. Check here before reusing a letter — the table above only tracks symbols the project has explicitly formalized, not conventions that came in with a borrowed theory.

• $C$ — Shannon channel capacity (Receiver-Side Recognition, The Veer); Correlation Fidelity (Eigen-Recognition). Not reused for the Trichotology pillar coupling, which is $\mathbf{W}$.
• $A$ — the pulse envelope $A(z,t)$ (Soliton); also appears in information-geometry contexts. Not reused for the pillar coupling.

This table is maintained as the appendices evolve. When a new formalization introduces a symbol, check it here first.