Source
Baez and Fong, “A Compositional Framework for Passive Linear Networks”, LIT-1644 (arXiv 1504.05625v6), Thm 1.1 (p. 7), Thm 7.7 (p. 47), Section 8 (pp. 48–51); read in NOTE-1223.
What was actually shown
The categories. Circ has finite sets as objects and open circuits as morphisms: a cospan X → N ← Y of finite sets with a passive linear circuit on N, composed by pushout, which glues the outputs of one circuit to the inputs of the next (Def. 5.3, Cor. 5.4, p. 30). LagRel has finite sets as objects and, as morphisms X → Y, Lagrangian subspaces of F^X ⊕ (F^X)* ⊕ F^Y ⊕ (F^Y)*, composed as relations (Def. 5.15, p. 35). Both are hypergraph categories.
The theorem. “There exists a hypergraph functor, the black box functor ■ : Circ → LagRel, mapping a finite set X to the symplectic vector space F^X ⊕ (F^X)* it generates, and an open circuit (N, E, s, t, r, i, o) to the Lagrangian relation ⋃_{v∈Graph(dQ)} S^t i(v) × So(v) ⊆ F^X ⊕ (F^X)* ⊕ F^Y ⊕ (F^Y)*, where Q is the circuit’s power functional” (Thm 1.1, p. 7; the black square and conjugation overlines are lost in extraction). Thm 7.7 builds it as a composite of four hypergraph functors: extended power functional, graph of its differential, restriction to the terminals by Kirchhoff’s laws, and a sign flip on input currents. Thm 8.4 and the computation on p. 51 show the result is Thm 1.1’s formula, in which the interior has been eliminated by the principle of minimum power.
What it means, in the authors’ words. “The Lagrangian relation does not shed any light on the internal workings of a circuit … Fortunately, the Lagrangian relation of a circuit is enough to completely characterize its external behavior, including how it interacts when connected with other circuits. Put more precisely, the black boxing process is functorial: we can compute the black boxed version of a circuit made of parts by computing the black boxed versions of the parts and then composing them” (p. 7).
What could have come out otherwise. The obvious semantics, Dirichlet forms composed by minimizing over the shared terminals, is not a category: it has no identity, because an ideal wire has no power functional (Prop. 3.6, p. 23). The theorem needs the move to Lagrangian relations, where the ideal wire is a relation (p. 5). Composition in the target is not automatic either. Gluing must impose equal potentials and opposite currents on the shared terminals (Thm 7.4, p. 46), and that the black box of a glued circuit equals the composite of black boxes is the content of the functor’s existence.
The identification. The functor is many-to-one: a 2-ohm resistor “has the same Lagrangian relation as two 1-ohm resistors in series” (p. 7), and two parallel resistors give the same image as one (p. 32). By functoriality, two circuits with one black box give one black box in any composite and any tensor product.
What this does not say
- Nothing outside passive, linear, two-terminal networks. Sources, nonlinear components and transformers are excluded (p. 3). The analogy with mechanical, hydraulic and chemical networks is a table and citations (p. 3), not a result.
- Nothing in the time domain. Inductors and capacitors enter as impedances in ℝ(s) after a Laplace transform with zero initial conditions (p. 16). The black box of an RLC circuit is an exact frequency-dependent relation; the paper does not turn it into a time-domain statement, and it says nothing about a circuit with a nonzero initial interior state.
- Not that the composite’s boundary relation is “emergent” beyond its parts. It is new relative to each part, as in CLAIM-480’s “neither part has”, but functoriality means it is computed from the parts’ black boxes and the gluing. No information about the interiors beyond their black boxes enters.
- Not that the external behaviour is a function. It is a relation: an ideal wire, or a circuit with a terminal used twice, gives no map from potentials to currents (p. 23), which is why the target is LagRel.
- Not that every Lagrangian relation is some circuit’s behaviour. The surjectivity proved is onto Dirichlet forms for real resistor circuits (Prop. 2.12, by citation to the star-mesh transform); the image of ■ in LagRel is not characterized here.
- Not a statement about constraints or apparatus. Those readings are the record’s (CLAIM-480), not the paper’s.