Pith. sign in

REVIEW 3 major objections 6 minor 2 cited by

The paper establishes that every symmetric monoidal structured or decorated cospan double category admits a canonical exoskeleton functor and makes every object a special symmetric Frobenius pseudomonoid.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

T0 review · deepseek-v4-flash

2026-08-04 14:48 UTC pith:WBGHEQ66

load-bearing objection A useful survey that packages known cospan results into new exoskeleton/Frobenius theorems, but the headline theorem silently assumes pushouts in the interface category; fixable, not fatal, but needs a correction. the 3 major comments →

arxiv 2509.22584 v3 pith:WBGHEQ66 submitted 2025-09-26 math.CT

Double Categories of Open Systems: the Cospan Approach

classification math.CT MSC 18N1018M05
keywords open systemsdouble categoriescospansstructured cospansdecorated cospansFrobenius pseudomonoidsvariable sharing paradigmPetri nets
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper argues that open systems in the variable sharing paradigm—composing systems by identifying shared variables—are uniformly organized by double categories built from cospans. It presents two constructions, structured and decorated cospans, and shows that despite their differences they share a common core: any symmetric monoidal such double category has an 'exoskeleton' formed by cospans of interfaces and an 'outer shell' formed by formal finite colimits, together with canonical symmetric monoidal double functors into the category. The central structural conclusion is that every object in such a double category carries a special symmetric Frobenius pseudomonoid structure, which formalizes the operations of joining, splitting, capping, and bending wires. The paper works through open Petri nets, open dynamical systems, and open Petri nets with rates, including a black-boxing functor for dynamical systems.

Core claim

The central claim is that the cospan approach to open systems, in both its structured and decorated forms, has a universal algebraic skeleton. For any double category D constructed from structured or decorated cospans, there is a double functor ι:Csp(D0)→D, unique up to isomorphism, restricting to the identity on the tight category D0; when D is symmetric monoidal, ι can be made symmetric monoidal. Because ordinary cospans already make every object a special symmetric Frobenius pseudomonoid—a categorified version of the classical Frobenius algebra laws for joining, splitting, capping, and bending wires—applying ι transfers this structure to every object of D. Thus the variable sharing paradi

What carries the argument

The load-bearing object is a cospan, a diagram X→M←Y in which M is the system and X and Y are its interfaces. Structured cospans are built from a functor L:A→X that turns interfaces into systems, while decorated cospans are built from a lax monoidal pseudofunctor F:A→Cat that equips the apex with decorations. The argument's central mechanism is the exoskeleton double category Csp(D0), whose loose morphisms are interface cospans together with the canonical functor ι:Csp(D0)→D; a further outer shell Csp(˜D0) is built from formal finite colimits on the objects of D. These functors carry the universal Frobenius structure of ordinary cospans into every structured or decorated cospan double catego

Load-bearing premise

The construction needs the interface category to have pushouts so that the exoskeleton Csp(D0) exists; for structured cospans the paper's standing hypotheses guarantee pushouts in the system category, not necessarily in the interface category.

What would settle it

Find a structured cospan double category satisfying the paper's standing hypotheses whose interface category lacks a pushout; then Csp(D0) cannot be formed, the exoskeleton functor ι is undefined, and Theorem 7.4's conclusion cannot be established by the given proof.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Any symmetric monoidal structured/decorated cospan double category has a canonical, essentially unique double functor from its exoskeleton, so cospans of interfaces are automatically degenerate open systems.
  • Every object in such a double category is a special symmetric Frobenius pseudomonoid; wire-joining, wire-splitting, capping, and bending are always coherent, giving a common algebraic core to variable-sharing models.
  • The exoskeleton/outer-shell maps give a concrete route from purely formal finite-colimit wiring patterns into any structured/decorated cospan double category.
  • Open Petri nets, open dynamical systems, and open Petri nets with rates all fit this pattern, and black-boxing open dynamical systems is a symmetric monoidal double functor to relations, making steady-state behavior compositional.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: because symmetric monoidal double functors preserve Frobenius pseudomonoid structure, any functorial semantics—token semantics, mass action, black-boxing—will automatically respect the wiring operations; the paper does not spell this out.
  • Editorial inference: the uniqueness of ι suggests that Csp(D0) may be characterized by a free universal property among fibrant double categories; the paper leaves this as an open challenge, and completing it would likely make the Frobenius conclusion fully formal.
  • Editorial inference: the outer-shell construction implies that only finite-colimit structure is needed to generate wiring patterns; software implementations of compositional modeling could in principle derive their glue operations from formal finite colimits rather than hand-coded combinators.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. This paper is an expository survey of double-categorical formalisms for open systems in the variable-sharing paradigm. It develops structured cospans (Sections 3–4) and decorated cospans (Sections 3 and 5–6), with running examples of open Petri nets, open dynamical systems, and open Petri nets with rates. The main new theoretical claims appear in Section 7: every symmetric monoidal structured/decorated cospan double category D admits a symmetric monoidal double functor ι : Csp(D₀) → D from its 'exoskeleton', a further map from an 'outer shell' Csp(∼D₀), and every object of such a D carries a special symmetric Frobenius pseudomonoid structure. The paper also proposes two tentative definitions of 'hypergraph double category' and poses several challenges. The exposition is clear and the examples are worked out in detail, but the statements of Theorems 7.4, 7.13, and 7.14 require stronger hypotheses on the interface category than are currently stated.

Significance. If the missing-hypothesis issues are repaired, this paper would be a valuable unifying overview of the cospan approach to open systems. Its explicit description of the exoskeleton and outer shell functors, and its translation of Fong–Spivak hypergraph categories into double-categorical language, are useful contributions, as are the concrete computations of black-boxing and the mass-action semantics. The paper is honest about its reliance on prior work and clearly separates proved statements from conjectures and challenges. It does not provide machine-checked proofs, but the proofs it does give are direct and checkable. The main risk is that the central theorems, as stated, overreach: they apply to structured cospan double categories whose interface category need not have pushouts or finite colimits, even though the proofs require exactly these properties. Since all examples in the paper use A = FinSet, the intended applications are safe, but the formal statements need correction.

major comments (3)
  1. [§7.1, Lemma 7.5 and Theorem 7.4] Lemma 7.5 claims that for any functor L:A→X with X having pushouts, there is a double functor ι:Csp(A)→LCsp. But Csp(A) is defined (Lemma 7.3) only when A has pushouts. The proof of Lemma 7.5 invokes Theorems 4.1 and 4.2 with X′ = A; Theorem 4.1 explicitly requires X and X′ to have pushouts. Thus the proof silently uses that A has pushouts, which is not among the hypotheses of Theorem 3.1 or Definition 7.2(1). Consequently Theorem 7.4 is overbroad for structured cospans: if A lacks pushouts, Csp(D₀) does not exist and ι is undefined. The fix is to add 'A has pushouts' (and, for the symmetric monoidal version, 'A has finite colimits') to Definition 7.2(1) and to the statement of Theorem 7.4. This is not fatal for the paper's examples, all of which use A = FinSet, but the theorem as written is false in the stated generality.
  2. [§7.4, Theorem 7.14] The proof of Theorem 7.14 applies Theorem 7.13 with A = D₀. But Theorem 7.13 requires A to have finite colimits, and for a structured cospan double category D = LCsp the tight category D₀ is A, which under the hypotheses of Theorem 3.2 is only assumed to have finite coproducts (with X having finite colimits). Thus even if one repairs Lemma 7.5 by adding pushouts in A, the Frobenius pseudomonoid structure is not established for all symmetric monoidal structured cospan double categories: D₀ may still lack pushouts and hence finite colimits. The theorem needs the explicit hypothesis that A has finite colimits (or, equivalently, that D₀ is finitely cocomplete). With that added, the proof via ι:Csp(D₀)→D works.
  3. [§7.1, proof of Theorem 7.4 (uniqueness)] The uniqueness part of Theorem 7.4 is dispatched via Lemma 7.7, which assumes that D is a fibrant double category. For structured cospans, fibrancy is established in Theorem 3.1. For decorated cospans, however, Theorem 3.3 does not state fibrancy, and no reference is given in the proof of Theorem 7.4 asserting that FCsp is fibrant. The paper should either add a cited theorem (e.g., from [9] or [74]) showing that decorated cospan double categories are fibrant, or prove fibrancy directly. Without this, the claimed uniqueness up to isomorphism of ι is not fully supported for the decorated case.
minor comments (6)
  1. [§7.3, before Tentative Definition 7.11] Typo: 'wll be' should be 'will be' in the sentence 'every symmetric monoidal structured or decorated cospan double category wll be a hypergraph double category'.
  2. [§3.1, historical remarks] The name 'Fiaideiro' appears in the text but the reference is to 'Fiadeiro'; please standardize the spelling.
  3. [§3.4, Theorem 3.4 diagram] In the tensor product of 2-cells, the line 'τ_{α⊗β} : F(h+h′)(φ_{m₁,m₁′}(d₁,d₁′)) → φ_{m₂,m₂′}(d₂,d₂′) in F(d₂,d′₂)' appears to contain a typo: the target category should be F(m₂+m′₂), not F(d₂,d′₂).
  4. [§4.2, square diagram] The square used to define Open(F) is drawn with '1' as the bottom arrow; this is presumably the identity functor on FinSet, but the label is ambiguous. Please clarify.
  5. [§7.2, paragraph after Lemma 7.9] The notation 'Csp(D)' appears where the text means 'Csp(D₀)' (the exoskeleton). This occurs in the displayed sequence 'Csp(∼D₀) → Csp(D) → D' and in the following sentence. Please fix to avoid confusion between the double category D and its tight category.
  6. [§7.4, proof of Theorem 7.12] The proof is labeled 'Proof Sketch' but the theorem is a central ingredient. Consider either expanding the proof or explicitly citing the result in [36] that makes the coherence laws follow from the universal property of pushouts.

Circularity Check

0 steps flagged

No significant circularity: Section 7 theorems are derived from cited external theorems and explicit pushout constructions; the unstated pushout hypothesis in A is a correctness gap, not a circular step.

full rationale

The derivation chain in Section 7 does not reduce to its own inputs. Theorem 7.4 asserts existence and uniqueness of the double functor ι: Csp(D0) → D; existence is proved in Lemma 7.5 by instantiating the earlier map-of-structured-cospans theorem (Theorem 4.1, itself [7, Thm. 4.2]) with the identity on A and L, and in Lemma 7.6 by applying Theorem 6.2 with empty decorations. These are genuine constructions, not restatements of the conclusion. Uniqueness is imported from Dawson–Paré–Pronk's freeness theorem (Lemma 7.7), an external result, not from the author's own prior work. Theorem 7.14 then applies the explicitly constructed ι to the special symmetric Frobenius pseudomonoid built in Csp(A) in Theorem 7.13 from the universal property of pushouts; that Frobenius structure is not assumed as a hypothesis. The only substantive issue found is a correctness gap: Lemma 7.5 and Theorem 7.4 silently require pushouts in the interface category A to form Csp(A), while Definition 7.2(1) only assumes pushouts in X. This makes the statements overbroad as written, but it is a missing hypothesis, not circularity: the proof would go through unchanged if 'A has pushouts' were added. Self-citations such as [7,9,19,20] are used for background constructions and are published, independent mathematical results; no fitted parameters, no definitional equivalences, and no renaming of known results are present. Accordingly, the circularity score is low.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 1 invented entities

No data-fitting parameters are present. The central claims rest on standard categorical background plus an unstated pushout hypothesis on the interface category for structured cospans.

axioms (5)
  • domain assumption D_0 (=A) has pushouts so Csp(D_0) is defined
    Lemma 7.5 constructs ι:Csp(A)→LCsp, but Theorem 3.1 only assumes X has pushouts; this pushout assumption on A is unflagged.
  • standard math Universal property of cospan double categories: Csp(A) is the free fibrant double category on A (Dawson-Paré-Pronk [38, Thm. 3.15])
    Used in Lemma 7.7 to prove uniqueness of ι in Theorem 7.4.
  • standard math Free category with finite colimits on a set X is FinSet↓X (Kelly [56, Thm. 5.35])
    Used in Lemma 7.9 to define the outer shell Csp(FinSet↓Ob(D_0)).
  • standard math Fibrant symmetric monoidal double categories give symmetric monoidal bicategories (Hansen-Shulman [53], Shulman [83])
    Used in Section 7.4 to pass from double categories to loose bicategories and then decategorify.
  • standard math Associator, unitors, and symmetry for cospans can be chosen via the universal property of pushouts to satisfy pentagon, triangle, and hexagon identities
    Proof sketch of Theorem 7.12 relies on this standard coherence argument without writing out all checks.
invented entities (1)
  • Hypergraph double category (Tentative Definition 7.11) no independent evidence
    purpose: Proposed unifying concept for symmetric monoidal structured/decorated cospan double categories in the variable sharing paradigm.
    Definition is explicitly tentative; the paper states coherence issues are unresolved (Challenge 7.15) and does not prove structured/decorated cospan categories satisfy it.

pith-pipeline@v1.3.0-alltime-deepseek · 37416 in / 13652 out tokens · 118968 ms · 2026-08-04T14:48:20.859501+00:00 · methodology

0 comments
read the original abstract

This is an overview of double categories of "open systems": systems that can interact with their environment. We focus on the variable sharing paradigm, where we compose open systems by identifying variables. This paradigm is often implemented using structured or decorated cospans. We explain this approach using three main examples: open Petri nets, open dynamical systems, and open Petri nets with rates. We compare the virtues of structured and decorated cospan double categories, and study their common features. We show that any symmetric monoidal structured or decorated cospan double category comes with maps from two simpler double categories: its "exoskeleton" and its "outer shell". Finally, we study the concept of "hypergraph double category", a kind of double category that should subsume structured and decorated cospans in a common framework for studying open systems in the variable sharing paradigm.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Enhanced $2$-categories of models of sketches as enhanced $2$-categories of algebras over monads

    math.CT 2026-05 unverdicted novelty 7.0

    Models of enhanced limit 2-sketches are equivalent to algebras over enhanced 2-monads, including lax morphisms, and inherit w-rigged limits.

  2. Enhanced $2$-categories of models of sketches as enhanced $2$-categories of algebras over monads

    math.CT 2026-05 unverdicted novelty 7.0

    Proves equivalence of models of enhanced 2-sketches with algebras over enhanced 2-monads in locally presentable enhanced 2-categories, characterizing w-rigged limits and generalizing enriched orthogonality and monadic...

Reference graph

Works this paper leans on

93 extracted references · 46 linked inside Pith · cited by 1 Pith paper

  1. [1]

    Aduddell, J

    R. Aduddell, J. Fairbanks, A. Kumar, P. S. Ocal, E. Patterson and B. T. Shapiro, A compositional account of motifs, mechanisms, and dynamics in biochemical regulatory networks,Compositionality6(2024). Also available as arXiv:2301.01445. (Referred to on page 2, 30.)

  2. [2]

    Aleiferi,Cartesian Double Categories with an Emphasis on Characterizing Spans, Ph.D

    E. Aleiferi,Cartesian Double Categories with an Emphasis on Characterizing Spans, Ph.D. Thesis, De- partment of Mathematics, Dalhousie University. Available as arXiv:1809.06940. (Referred to on page 8.)

  3. [3]

    (Referred to on page 35.)

    AlgebraicDynamics.jl, available at https://algebraicjulia.github.io/AlgebraicDynamics.jl/. (Referred to on page 35.)

  4. [4]

    (Referred to on page 20, 35.)

    AlgebraicPetri.jl, available at https://algebraicjulia.github.io/AlgebraicPetri.jl/. (Referred to on page 20, 35.)

  5. [5]

    A. Baas, J. Fairbanks, M. Halter, S. Libkind and E. Patterson, An algebraic framework for structured epidemic modeling,Phil. Trans. Roy. Soc. A380(2022), 20210309. Also available as arXiv:2203.16345. (Referred to on page 19, 20, 30.)

  6. [6]

    J. C. Baez and A. Chaudhuri, Graphs with polarities. Available as arXiv:2506.23375. (Referred to on page 2, 9.)

  7. [7]

    J. C. Baez and K. Courser, Structured cospans,Theory Appl. Categ.35(2020), 1771–1822. Also available as arXiv:1911.04630. (Referred to on page 2, 8, 9, 18.)

  8. [8]

    J. C. Baez and K. Courser, Coarse-graining open Markov processes,Theory Appl. Categ.33(2018), 1223–

  9. [9]

    J. C. Baez, K. Courser and C. Vasilakopoulou, Structured versus decorated cospans,Compositionality43 (2022). Also available as arXiv:2101.09363. (Referred to on page 9, 10, 11, 12, 29, 34.) 52 DOUBLE CATEGORIES OF OPEN SYSTEMS: THE COSPAN APPROACH

  10. [10]

    J. C. Baez, B. Coya and F. Rebro, Props in circuit theory,Theory Appl. Categ.33(2018), 727–783. Available as arXiv:1707.08321. (Referred to on page 2.)

  11. [11]

    J. C. Baez and J. Erbele, Categories in control,Theory Appl. Categ.30(2015), 836–881. Also available as arXiv:1405.6881. (Referred to on page 2, 27.)

  12. [12]

    J. C. Baez and B. Fong, A compositional framework for passive linear networks,Theory Appl. Categ.33 (2018), 1158–1222. Also available as arXiv:1504.05625. (Referred to on page 2, 37.)

  13. [13]

    J. C. Baez, B. Fong and B. S. Pollard, A compositional framework for Markov processes,Jour. Math. Phys. 57(2016), 033301. Also available as arXiv:1508.06448. (Referred to on page 2.)

  14. [14]

    J. C. Baez, F. Genovese, J. Master and M. Shulman, Categories of nets, in36th Annual ACM/IEEE Sym- posium on Logic in Computer Science (LICS), IEEE, Rome, Italy, 2021, pp. 1–13. Also available as arXiv:2101.04238. (Referred to on page 13, 20, 21.)

  15. [15]

    J. C. Baez and A. Lauda, A prehistory ofn-categorical physics, inDeep Beauty: Mathematical Innovation and the Search for an Underlying Intelligibility of the Quantum World, ed. Hans Halvorson, Cambridge U. Press, Cambridge, 2011, pp. 13-128. Also available as arXiv:0908.2469. (Referred to on page 2, 3.)

  16. [16]

    J. C. Baez, X. Li, S. Libkind, N. D. Osgood and E. Redekopp, A categorical framework for modeling with stock and flow diagrams, inMathematics of Public Health: Mathematical Modelling from the Next Generation, eds. J. David and J. Wu, Springer, 2003, pp. 175–207. Also available as arXiv:2211.01290. (Referred to on page 2, 36.)

  17. [17]

    J. C. Baez, X. Li, S. Libkind, N. D. Osgood and E. Patterson, Compositional modeling with stock and flow diagrams,Electron. Proc. Theor. Comput. Sci.380(2023), 77–96. Also available as arXiv:2205.08373. (Referred to on page 2, 36.)

  18. [18]

    J. C. Baez, O. Lynch and J. Moeller, Compositional thermostatics,J. Math. Phys.64(2023), 023304. Also available as arXiv:2111.10315. (Referred to on page 2.)

  19. [19]

    J. C. Baez and J. Master, Open Petri nets,Math. Struct. Comput. Sci.30(2020), 314–341. Also available as arXiv:1808.05415. (Referred to on page 13, 16, 17.)

  20. [20]

    J. C. Baez and B. S. Pollard, A compositional framework for chemical reaction networks,Rev. Math. Phys. 29(2017), 1750028. Also available as arXiv:1704.02051. (Referred to on page 24, 29, 34, 35.)

  21. [21]

    J. C. Baez and M. Stay, Physics, topology, logic and computation: a Rosetta Stone, inNew Structures for Physics, ed. B. Coecke, Lecture Notes in Physics 813, Springer, Berlin, 2011, pp. 95–172. Also available as arXiv:0903.0340. (Referred to on page 2, 5.)

  22. [22]

    J. C. Baez, D. Weisbart and A. M. Yassine, Open systems in classical mechanics,Jour. Math. Phys.62 (2021), 042902. Also available as arXiv:1710.11392. (Referred to on page 2, 27.)

  23. [23]

    Bonchi, P

    F. Bonchi, P. Soboci´nski and F. Zanasi, A categorical semantics of signal flow graphs, inCONCUR 2014– Concurrency Theory, eds. P. Baldan and D. Gorla, Lecture Notes in Computer Science8704, Springer, Berlin, 2014, pp. 435–450. Also available at http://users.ecs.soton.ac.uk/ps/papers/sfg.pdf. (Referred to on page 2.)

  24. [24]

    Carboni, Matrices, relations, and group representations,J

    A. Carboni, Matrices, relations, and group representations,J. Algebra136(1991), 497–529. (Referred to on page 38.)

  25. [25]

    Available at https://catcolab.org/help

    CatColab. Available at https://catcolab.org/help. (Referred to on page 36.)

  26. [26]

    Available at https://github.com/AlgebraicJulia/Catlab.jl

    Catlab.jl. Available at https://github.com/AlgebraicJulia/Catlab.jl. (Referred to on page 11, 35.)

  27. [27]

    Courser, A bicategory of decorated cospans,Theory Appl

    K. Courser, A bicategory of decorated cospans,Theory Appl. Categ.32(2017), 995–1027. Also available as arXiv:1605.08100. (Referred to on page 6.)

  28. [28]

    Courser,Open Systems: a Double Categorical Perspective, Ph.D

    K. Courser,Open Systems: a Double Categorical Perspective, Ph.D. thesis, Department of Mathematics, U. C. Riverside, 2020. Available as arXiv:2008.02394. (Referred to on page 7, 8, 9.)

  29. [29]

    B. Coya, B. Fong, Corelations are the prop for extraspecial commutative Frobenius monoids,Theory Appl. Categ.32(2017), 380–395. Also available as arXiv:1601.02307. (Referred to on page .)

  30. [30]

    Coya, A compositional framework for bond graphs

    B. Coya, A compositional framework for bond graphs. Available as arXiv:1710.00098. (Referred to on page 2, 27.)

  31. [31]

    Coya,Circuits, Bond Graphs, and Signal-Flow Diagrams: A Categorical Perspective, Ph.D

    B. Coya,Circuits, Bond Graphs, and Signal-Flow Diagrams: A Categorical Perspective, Ph.D. thesis, Department of Mathematics, U. C. Riverside, 2018. Available as arXiv:1805.08290. (Referred to on page 2, 27.)

  32. [32]

    Craciun, Toric differential inclusions and a proof of the Global Attractor Conjecture

    G. Craciun, Toric differential inclusions and a proof of the Global Attractor Conjecture. Available as arXiv:1501.02860. (Referred to on page 36.)

  33. [33]

    Craciun, M

    G. Craciun, M. Mincheva, C. Pantea and P. Y . Yu, A graph-theoretic condition for delay stability of reaction systems. Available as arXiv:2105.07321. (Referred to on page 37.)

  34. [34]

    Craciun, F

    G. Craciun, F. Nazarov and C. Pantea, Persistence and permanence of mass-action and power-law dynam- ical systems,SIAM J. Appl. Math.73(2013), 305–329. (Referred to on page 36.) DOUBLE CATEGORIES OF OPEN SYSTEMS: THE COSPAN APPROACH 53

  35. [35]

    Craciun, Y

    G. Craciun, Y . Tang and M. Feinberg, Understanding bistability in complex enzyme-driven reaction net- works,PNAS103(2006), 8697–8702. (Referred to on page 30, 37.)

  36. [36]

    Day and R

    B. Day and R. Street, Monoidal bicategories and Hopf algebroids,Adv. Math.129(1997), 99–157. (Re- ferred to on page 48, 49.)

  37. [37]

    R. J. M. Dawson, B. Par ´e and D. A. Pronk, Universal properties of Span,Theory Appl. Categ.13(2004), 61–85. Available at http://www.tac.mta.ca/tac/volumes/13/4/13-04abs.html. (Referred to on page .)

  38. [38]

    R. J. M. Dawson, B. Par ´e and D. A. Pronk, The span construction,Theory Appl. Categ.24(2010), 302–

  39. [39]

    Feinberg, Chemical reaction network structure and the stability of complex isothermal reactors: I

    M. Feinberg, Chemical reaction network structure and the stability of complex isothermal reactors: I. The deficiency zero and deficiency one theorems,Chem. Eng. Sci.42(1987), 2229–2268. (Referred to on page 36, 37.)

  40. [40]

    Feinberg,Foundations of Chemical Reaction Network Theory, Springer, Berlin, 2019

    M. Feinberg,Foundations of Chemical Reaction Network Theory, Springer, Berlin, 2019. (Referred to on page 36, 37.)

  41. [41]

    J. L. Fiadeiro and V . Schmitt, Structured co-spans: an algebra of interaction protocols, inInternational Conference on Algebra and Coalgebra in Computer Science, eds. T. Mossakowski, U. Montanari and M. Haveraaen, Lecture Notes in Computer Science 4624, Springer, Berlin, 2007, pp. 194–208. (Referred to on page 9.)

  42. [42]

    Fong, Decorated cospans,Theory Appl

    B. Fong, Decorated cospans,Theory Appl. Categ.30(2015), 1096–1120. Also available as arXiv:1502.00872. (Referred to on page 9, 11, 37.)

  43. [43]

    Fong,The Algebra of Open and Interconnected Systems, Ph.D

    B. Fong,The Algebra of Open and Interconnected Systems, Ph.D. thesis, Department of Computer Sci- ence, University of Oxford, 2016. Also available as arXiv:1609.05382. (Referred to on page 7, 11, 37.)

  44. [44]

    B. Fong, P. Rapisarda and P. Sobocinski, A categorical approach to open and interconnected dynamical systems, inProceedings of the 31st Annual ACM/IEEE Symposium on Logic in Computer Science (LICS), IEEE, New York, 2016, pp. 1–10. Available as arXiv:1510.05076. (Referred to on page 2.)

  45. [45]

    Fong and D

    B. Fong and D. Spivak, Hypergraph categories,J. Pure Appl. Alg.223(2019), 4746–4777. Also available as https://arxiv.org/abs/1806.08304. (Referred to on page 38, 43, 46, 51.)

  46. [46]

    Gadducci and R

    F. Gadducci and R. Heckel, An inductive view of graph transformation, inRecent Trends in Algebraic Development Techniques (Tarquinia, 1997), Springer Lecture Notes in Computer Science, vol. 1376, Springer, Berlin, 1998, pp. 223–237. (Referred to on page 38.)

  47. [47]

    Girault and R

    C. Girault and R. Valk,Petri Nets for Systems Engineering: a Guide to Modeling, Verification, and Appli- cations, Springer, Berlin, 2013. (Referred to on page 13.)

  48. [48]

    Grandis and R

    M. Grandis and R. Par ´e, Limits in double categories,Cah. Top. G´eom. Diff.40(1999), 162–220. Available at https://www.numdam.org/item/CTGDC 1999 40 3 162 0/(Referred to on page 2.)

  49. [49]

    Grandis and R

    M. Grandis and R. Par ´e, Adjoints for double categories,Cah. Top. G ´eom. Diff.45(2004), 193–240. (Re- ferred to on page 2.)

  50. [50]

    Gordon, A

    R. Gordon, A. J. Power and R. Street, Coherence for tricategories,Mem. Amer. Math. Soc.558, 1995. (Referred to on page 6.)

  51. [51]

    P. J. Haas,Stochastic Petri Nets: Modelling, Stability, Simulation, Springer, Berlin, 2002. (Referred to on page 30.)

  52. [52]

    Halter and E

    M. Halter and E. Patterson, Compositional epidemiological modeling using structured cospans, 2020. Available at https://www.algebraicjulia.org/blog/post/2020/10/structured-cospans. (Referred to on page 11.)

  53. [53]

    L. W. Hansen and M. Shulman, Constructing symmetric monoidal bicategories functorially. Available as arXiv:1910.09240. (Referred to on page 47.)

  54. [54]

    Heunen and J

    C. Heunen and J. Vicary,Categories for Quantum Theory: an Introduction, Oxford U. Press, Oxford, 2019. (Referred to on page 5.)

  55. [55]

    Horn and R

    F. Horn and R. Jackson, General mass action kinetics,Arch. Ration. Mech. Anal.47(1972), 81–116. (Referred to on page 36.)

  56. [56]

    G. M. Kelly,Basic Concepts of Enriched Category Theory, Cambridge U. Press, Cambridge, 1982. Also available at http://www.tac.mta.ca/tac/reprints/articles/10/tr10abs.html. (Referred to on page 43.)

  57. [57]

    Kissinger, Finite matrices are complete for (dagger-)hypergraph categories

    A. Kissinger, Finite matrices are complete for (dagger-)hypergraph categories. Available as arXiv:1406.5942. (Referred to on page 38.)

  58. [58]

    Koch, Petri nets—a mathematical formalism to analyze chemical reaction networks,Mol

    I. Koch, Petri nets—a mathematical formalism to analyze chemical reaction networks,Mol. Inform.29 (2010), 838–843. (Referred to on page 30.)

  59. [59]

    Kock,Frobenius Algebras and 2D Topological Quantum Field Theories, Cambridge U

    J. Kock,Frobenius Algebras and 2D Topological Quantum Field Theories, Cambridge U. Press, Cam- bridge, 2003. Short version available at https://mat.uab.cat/˜kock/TQFT/FS.pdf. (Referred to on page 3.)

  60. [60]

    Kock, Whole-grain Petri nets and processes,J

    J. Kock, Whole-grain Petri nets and processes,J. ACM70(2022), 1–58. Also available as arXiv:2005.05108. (Referred to on page 2, 20.) 54 DOUBLE CATEGORIES OF OPEN SYSTEMS: THE COSPAN APPROACH

  61. [61]

    F. W. Lawvere,Functorial Semantics of Algebraic Theories, Ph.D. thesis, Columbia University, 1963. Reprinted inTheory Appl. Categ.5(2004), 1–121. Available at . (Referred to on page 2.)

  62. [62]

    X. Li, P. L. Mabry, N. D. Osgood and E. Patterson, Compositional system dynamics: the higher mathe- matics underlying system dynamics diagrams & practice. Available at arXiv:2509.18475. (Referred to on page 36.)

  63. [63]

    Libkind, An algebra of resource sharing machines

    S. Libkind, An algebra of resource sharing machines. Available as arxiv:2007.14442. (Referred to on page 2, 36.)

  64. [64]

    Libkind and D

    S. Libkind and D. J. Myers, Towards a double operadic theory of systems. Available as arXiv:2505.18329. (Referred to on page 2, 36, 39, 44, 46, 47, 51.)

  65. [65]

    Lynch,Relational Composition of Physical Systems: A Categorical Approach, M.Sc

    O. Lynch,Relational Composition of Physical Systems: A Categorical Approach, M.Sc. Thesis, Depart- ment of Physics, Universiteit Utrecht, 2022. Available at arXiv:2310.06088. (Referred to on page 2, 27.)

  66. [66]

    Master, Petri nets based on Lawvere theories,Math

    J. Master, Petri nets based on Lawvere theories,Math. Struct. Comp. Sci.30(2020), 833–864. Also available as arXiv:1904.09091. (Referred to on page 15.)

  67. [67]

    McCrudden, Balanced coalgebroids,Theory Appl

    P. McCrudden, Balanced coalgebroids,Theory Appl. Categ.7(2000), 71–147. Available at http://www.tac.mta.ca/tac/volumes/7/n6/7-06abs.html. (Referred to on page 6.)

  68. [68]

    Available at https://modelcollab.usask.ca/

    ModelCollab. Available at https://modelcollab.usask.ca/. (Referred to on page 36.)

  69. [69]

    Morton, Belief propagation in monoidal categories,Electron

    J. Morton, Belief propagation in monoidal categories,Electron. Proc. Theor. Comput. Sci.172(2014) 262–269. Also available as arXiv:1504.2618. (Referred to on page 38.)

  70. [70]

    D. J. Myers, Double categories of open dynamical systems (extended abstract),EPTCS333(2021), 154–

  71. [71]

    D. J. Myers,Categorical Systems Theory, draft as of September 3, 2023. Available at https://www.davidjaz.com/Papers/DynamicalBook.pdf. (Referred to on page 2, 36, 44.)

  72. [72]

    Niefield, Span, cospan, and other double categories,Theory Appl

    S. Niefield, Span, cospan, and other double categories,Theory Appl. Categ.26(2012), 729–742. Available as arXiv:1201.3789. (Referred to on page 40.)

  73. [73]

    Par ´e, Superspans, talk at the Octoberfest, Ottawa, 2015

    B. Par ´e, Superspans, talk at the Octoberfest, Ottawa, 2015. Available at https://www.mscs.dal.ca/pare/Superspans(Beamer).pdf. (Referred to on page 9.)

  74. [74]

    Patterson, Structured and decorated cospans from the viewpoint of double category theory,Electron

    E. Patterson, Structured and decorated cospans from the viewpoint of double category theory,Electron. Proc. Theor. Comput. Sci.397(2023). Also available as arXiv:2304.00447. (Referred to on page 8, 9.)

  75. [75]

    Patterson and M

    E. Patterson and M. Halter, Compositional epidemiological modeling using structured cospans,Alge- braicJulia Blog, 2020. Available at https://blog.algebraicjulia.org/post/2020/10/structured-cospans/. (Re- ferred to on page 35.)

  76. [76]

    J. L. Peterson,Petri Net Theory and the Modeling of Systems, Prentice-Hall, New Jersey, 1981. (Referred to on page 13.)

  77. [77]

    B. S. Pollard,Open Markov Processes and Reaction Networks, Ph.D. thesis, Department of Physics, U. C. Riverside, 2017. Available as arXiv:1709.09743. (Referred to on page 7.)

  78. [78]

    Pstra ¸gowski, On dualizable objects in monoidal bicategories

    P. Pstra ¸gowski, On dualizable objects in monoidal bicategories. Available as arXiv:1411.6691. (Referred to on page 50.)

  79. [79]

    Rosebrugh, N

    R. Rosebrugh, N. Sabadini, R. F. C. Walters, Generic commutative separable algebras and cospans of graphs,Theory Appl. Categ.15(2005), 164–177. Available at http://tac.mta.ca/tac/volumes/15/6/15- 06abs.html. (Referred to on page 38.)

  80. [80]

    Selinger, Autonomous categories in whichA A ∗ (extended abstract), talk at QPL 2010

    P. Selinger, Autonomous categories in whichA A ∗ (extended abstract), talk at QPL 2010. Available at https://ncatlab.org/nlab/files/SelingerSelfDual.pdf. (Referred to on page 5.)

Showing first 80 references.