Pith. sign in

REVIEW 2 major objections 4 minor 40 references

Any static decomposition theory lifts, under categorical hypotheses, to a temporal theory that recovers natural time-varying analogues of tree-width.

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 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-14 11:41 UTC pith:EHPA3MP2

load-bearing objection Clean categorical lift of static decompositions to persistent narratives; theorems hold under explicit axioms, recovered widths are max-over-snapshots, main external hypothesis is sheafification. the 2 major comments →

arxiv 2607.10442 v1 pith:EHPA3MP2 submitted 2026-07-11 math.CT math.CO

Decomposing time-varying data into simple pieces: structured decompositions of narratives

Benjamin Merlin Bumpus , Jana K. Nickel This is my paper
classification math.CT math.CO MSC 18A3018F2005C7568R10
keywords structured decompositionspersistent narrativesspined sd-categoriestemporal tree-widthtime-varying graphssheafificationadhesive categories
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 asks whether there is a systematic way to decompose objects that change with time, starting from ordinary static decompositions such as tree-decompositions. It combines two existing categorical tools: structured decompositions (functors out of the barycentric subdivision of a graph that send edges to monomorphisms) and persistent narratives (presheaves on a discrete time category that turn certain squares into pullbacks). The central result is that, once a few pullback, colimit and sheafification conditions hold, any spined structured-decomposition category on a static base lifts to a spined structured-decomposition category on the category of persistent narratives. Instantiating the lift on graphs recovers temporal analogues of ordinary tree-width, complemented tree-width and the tree-independence number, in each case by taking a maximum of the corresponding static sizes over a chosen sub-semilattice of time intervals. The construction therefore supplies a uniform recipe rather than a single ad-hoc definition of temporal width.

Core claim

Under four categorical hypotheses (spine inclusions preserve and reflect pullbacks, the base is finitely cocomplete, a sheafification left adjoint that is a retraction exists, and pushouts along monomorphisms are also pullbacks with monomorphisms stable), the category of persistent narratives valued in a spined sd-category is itself a spined sd-category. Consequently every static width measure induced by a spine lifts to a temporal width measure on time-varying data.

What carries the argument

Temporalization of a spined sd-category: the subcategories pΩ_n of those persistent narratives that land in the n-th spine layer on a chosen sub-semilattice of time intervals; Theorems 3.2 and 3.4 prove that these form a spine once the four hypotheses hold.

Load-bearing premise

The sheafification left adjoint that turns ordinary diagrams into persistent narratives must exist and be a retraction, so that colimits of decompositions remain inside the subcategory of narratives rather than merely in the larger presheaf category.

What would settle it

Exhibit a concrete spined sd-category that satisfies the other three hypotheses but admits no sheafification left adjoint that is a retraction, and check whether structured decompositions of its persistent narratives still possess colimits inside the narrative subcategory.

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

If this is right

  • Any static width measure already captured by a spined sd-category (tree-width, path-width, independence number, …) automatically yields a family of temporal width measures indexed by sub-semilattices of the time category.
  • Varying the sub-semilattice S of time intervals gives a tunable parameter: longer intervals force the temporal width to ignore short-lived fluctuations.
  • The same lifting recipe applies verbatim to other adhesive or quasitopos bases once their sheafification functors are known.
  • The recovered temporal tree-width, complemented tree-width and tree-independence number become well-defined invariants of persistent narratives of graphs.

Where Pith is reading between the lines

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

  • The dual cumulative-narrative construction sketched in the conclusion is likely to succeed under the dual hypotheses, producing a second, co-presheaf-based family of temporal widths.
  • If the conjectured embedding of structured decompositions of narratives into narratives of structured decompositions holds, one obtains a single 2-categorical setting in which static and temporal decompositions can be compared directly.
  • The framework supplies a candidate language for transferring algorithmic meta-theorems (Courcelle-type results) from static graphs to temporal graphs once the temporal width is bounded.

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

2 major / 4 minor

Summary. The paper develops a categorical procedure that lifts any static spined structured-decomposition category (C,G,Ω) to a temporal counterpart on the category Pe(T,C) of C-valued persistent narratives over a finite discrete time category T. Under hypotheses (T1)–(T4)—pullback-preserving full spine inclusions, finite cocompleteness of C, existence of a sheafification left adjoint that is a retraction of the inclusion Pe(T,C)↪[T^op,C], and pushouts along monomorphisms that remain pullbacks with monomorphisms stable—the pair (Pe(T,C),G) is shown to be an sd-category (Theorem 3.2) and the image subcategories pΩ_n form a spine (Theorem 3.4). Three case studies instantiate the construction for ordinary tree-width (via reflexive graphs), complemented tree-width, and the tree-independence number, recovering the expected “maximum over snapshots in S” formulae for the resulting temporal widths.

Significance. The work supplies a uniform, axiomatic route from static width theories to temporal ones, rather than ad-hoc combinatorial definitions. The proofs of the two main theorems are written out in full detail (pointwise colimits followed by sheafification; explicit pushout/pullback constructions and pasting for the spine axioms), and the case studies correctly recover the natural max-snapshot analogues once the ambient category is adjusted. This is a genuine contribution to the programmatic question of “the right way” to decompose objects of an arbitrary category, and it places several recent temporal-graph parameters inside a single categorical framework. The external character of (T3) for general C is a limitation, but it is already flagged by the authors and is verified for the graph examples they treat.

major comments (2)
  1. Axiom (T3) is load-bearing for Theorem 3.2: without a sheafification left adjoint that is a retraction, the pointwise colimits constructed in the presheaf category need not land back inside Pe(T,C). The paper verifies the axiom for Grph_refl via ordinary set-valued sheafification, but leaves it as an external hypothesis for general C. A short additional paragraph (or appendix remark) listing further classes of C for which (T3) is known, or stating a sufficient condition, would strengthen the claim that the lift is “systematic” beyond graphs.
  2. In Case Study 4.1 the ambient category is changed from ordinary simple graphs to reflexive graphs in order to obtain finite cocompleteness and to make the identity-on-vertices maps into morphisms. The resulting pΓ-width is therefore a temporal width for reflexive graphs, not for ordinary simple graphs. The manuscript should state explicitly whether (and how) the ordinary simple-graph tree-width is recovered by restriction or by a forgetful functor, so that the claim “natural temporal analogues of ordinary tree-width” is fully justified.
minor comments (4)
  1. Several typographical slips appear: “althoughthetheory” (p. 2), “time-respectingnotion” (p. 12), “spinedsd-categories” (p. 22), and missing spaces after periods in the abstract and introduction.
  2. Figure 2.1 and the schematic diagrams in §3 would benefit from larger fonts and consistent arrow styles; the current rendering makes the pullback squares hard to parse at a glance.
  3. The notation pΩ_n is introduced as the image of π_n, yet later used interchangeably with the full subcategory of narratives whose values on S land in Ω_n (Lemma 3.1). A single clarifying sentence after the definition would avoid momentary confusion.
  4. References [9] and [10] are cited heavily; a brief parenthetical reminder of the precise statements being lifted would help readers who have not memorised those preprints.

Circularity Check

1 steps flagged

No significant circularity: lift theorems proved from explicit axioms (T1)–(T4); self-citations supply input frameworks only.

specific steps
  1. self citation load bearing [Section 2 (Defs 2.3, 2.10–2.16) and Section 4 (case studies citing [10, Props 3.1.1, 3.3.4, 3.4.5])]
    "We apply structured decompositions to categories of persistent narratives. ... the static spined sd-categories of [10] that recover the time-independent notions of ordinary tree-width, complemented tree-width and the tree-independence number and to apply our theorem to those contexts."

    The input notions of structured decomposition, spine, and the three static width parameters are taken from prior work by overlapping authors. This is ordinary framework citation rather than a load-bearing circular step: the lift theorems and the explicit max-over-S formulae are proved independently from the axioms and do not reduce to the cited results by construction.

full rationale

The central claims (Theorems 3.2 and 3.4) construct a temporal spined sd-category (Pe(T,C),G,pΩ) by verifying the sd-category and spine axioms (S1)–(S5) under hypotheses (T1)–(T4). Proofs proceed by pointwise colimits in the presheaf category (via (T2)), preservation under the sheafification left adjoint S that is a retraction ((T3)), monomorphism stability and pasting of pullbacks/pushouts ((T4)), and fullness of the Ω_n layers (Lemma 3.1). Case studies then compute the induced sizes/widths explicitly (Propositions 4.3, 4.6, 4.7), obtaining max_s s_Γ(X(s)) and max_s w_Γ(d_s); these equalities follow by direct unwinding of the general definitions, not by fitting or tautological redefinition. Self-citations to [9] (persistent narratives) and [10] (structured decompositions/spined sd-categories) introduce the static ingredients being lifted; they are not used to justify the lift itself or to import uniqueness/ansatz that forces the temporal widths. No parameter is fitted and then re-presented as a prediction; no known empirical pattern is merely renamed. The derivation is therefore self-contained given the stated categorical hypotheses.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 2 invented entities

The central claims rest on four additional categorical hypotheses (T1)–(T4) beyond the already-published definitions of spined sd-categories and persistent narratives. No numerical free parameters are fitted. The only invented entities are the technical pullback categories Pn and the image subcategories pΩ_n that realize the temporal spine; they are defined explicitly and have no independent empirical content.

axioms (5)
  • domain assumption (T1) C and each Ω_n admit all pullbacks; inclusions ι_n preserve (and, being full, reflect) pullbacks.
    Required so that Pe(T,Ω_n) is well-defined and the change-of-base functors exist; invoked throughout §3.1–3.3.
  • domain assumption (T2) C is finitely cocomplete.
    Used in the proof of Theorem 3.2 to obtain pointwise colimits of structured decompositions.
  • domain assumption (T3) The inclusion Pe(T,C) ↪ [T^op,C] admits a left-adjoint sheafification S that is a retraction.
    Load-bearing for existence of colimits inside Pe(T,C); assumed rather than constructed for general C.
  • domain assumption (T4) C admits pushouts along monomorphisms; those squares are also pullbacks; monomorphisms are stable under pushouts.
    Used heavily in the verification of spine axiom (S4) to glue monomorphisms across time intervals of length one.
  • domain assumption T is a finite discrete time category (finite join-semilattice of discrete intervals).
    Finiteness is used to take maxima of sizes and to ensure the zigzag of length-one intervals is finite.
invented entities (2)
  • pΩ_n (image of the pullback category Pn under projection to Pe(T,C)) no independent evidence
    purpose: Serves as the n-th layer of the temporal spine on persistent narratives.
    Defined by a categorical pullback construction; no external evidence claimed or needed.
  • Temporalized spined sd-category (Pe(T,C),G,pΩ) no independent evidence
    purpose: The object whose existence is the main theorem; carries the lifted width measure.
    Constructed from the static data plus (T1)–(T4); the entity is the theorem’s output.

pith-pipeline@v1.1.0-grok45 · 31282 in / 3000 out tokens · 35061 ms · 2026-07-14T11:41:45.895447+00:00 · methodology

0 comments
read the original abstract

Graphs that change over time arise throughout applications, but there is no single standard way to decompose them into smaller pieces. In this paper, we propose a systematic categorical method for doing so. The main idea is to combine structured decompositions, which generalize graph decompositions, such as tree-decompositions, with persistent narratives, which model time-varying data as diagrams. We prove that, under suitable categorical hypotheses, any static theory of decompositions can be lifted to a corresponding temporal theory. As case studies, we apply this construction to time-varying graphs and recover natural temporal analogues of ordinary tree-width, complemented tree-width, and the tree-independence number.

Figures

Figures reproduced from arXiv: 2607.10442 by Benjamin Merlin Bumpus, Jana K. Nickel.

Figure 2.1
Figure 2.1. Figure 2.1: A schematic visualization of an example of a finite discrete time category. [PITH_FULL_IMAGE:figures/full_fig_p005_2_1.png] view at source ↗
Figure 2.2
Figure 2.2. Figure 2.2: A tree-decomposition of a graph (on the left-hand side) and its decomposition tree (on [PITH_FULL_IMAGE:figures/full_fig_p006_2_2.png] view at source ↗
Figure 2.3
Figure 2.3. Figure 2.3: The cycle C 5 , the category R C 5 and a C 5 -shaped structured decomposition of graphs. Replacing T by an arbitrary graph leads to the more general notion of a graph-decomposition, studied for example in [12] and [16]. The concept of graph decompositions moreover gives rise to a great variety of combinatorial width parameters measuring the structural resemblance of a graph with respect to a specific gra… view at source ↗

discussion (0)

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

Reference graph

Works this paper leans on

40 extracted references · 4 canonical work pages

  1. [1]

    Time-Varying Graphs and Dynamic Networks

    A. Casteigts, P. Flocchini, W. Quattrociocchi, and N. Santoro. “Time-Varying Graphs and Dynamic Networks”. In:Ad-hoc, Mobile, and Wireless Networks. Ed. by H. Frey, X. Li, and S. Ruehrup. Vol. 6811. Lecture Notes in Computer Science. Springer Berlin Heidelberg, 2011, pp. 346–359

    2011

  2. [2]

    Compo- sitional Modeling with Stock and Flow Diagrams

    John Baez, Xiaoyan Li, Sophie Libkind, Nathaniel D. Osgood, and Evan Patterson. “Compo- sitional Modeling with Stock and Flow Diagrams”. In:Electronic Proceedings in Theoretical Computer Science380 (2023).url:http://dx.doi.org/10.4204/EPTCS.380.5

    doi:10.4204/eptcs.380.5 2023

  3. [3]

    url:https://arxiv.org/abs/1911.04630v3

    JohnC.BaezandKennyCourser.Structured Cospans.2020.arXiv:1911.04630v3 [math.CT]. url:https://arxiv.org/abs/1911.04630v3

    Pith/arXiv arXiv 2020

  4. [4]

    Structured versus Deco- rated Cospans

    John C. Baez, Kenny Courser, and Christina Vasilakopoulou. “Structured versus Deco- rated Cospans”. In:Compositionality4 (2022).url:http : / / dx . doi . org / 10 . 32408 / compositionality-4-3

    2022

  5. [5]

    Open Petri nets

    John C. Baez and Jade Master. “Open Petri nets”. In:Mathematical Structures in Computer Science30 (2020), pp. 314–341.url:https://doi.org/10.1017/S0960129520000043

    work page doi:10.1017/s0960129520000043 2020

  6. [6]

    A compositional framework for reaction networks

    John C. Baez and Blake S. Pollard. “A compositional framework for reaction networks”. In:Reviews in Mathematical Physics29.09 (2017).url:https : / / doi . org / 10 . 1142 / S0129055X17500283

    2017

  7. [7]

    Basic Category Theory

    Francis Borceux.Handbook of Categorical Algebra 1. Basic Category Theory. Cambridge Uni- versity Press, 1994

    1994

  8. [8]

    Binh-Minh Bui-Xuan, Florent Krasnopol, Bruno Monasson, and Nathalie Sznajder.Model checking with temporal graphs and their derivative. 2026. arXiv:2602.12446v3 [cs.DS].url: https://arxiv.org/abs/2602.12446v3. 32

    Pith/arXiv arXiv 2026

  9. [9]

    Benjamin Merlin Bumpus, James Fairbanks, Martti Karvonen, Wilmer Leal, and Frédéric Simard.Towards a Unified Theory of Time-Varying Data. 2025. arXiv:2402 . 00206v3 [math.CT].url:https://arxiv.org/abs/2402.00206v3

    Pith/arXiv arXiv 2025

  10. [10]

    Kocsis, Jade Edenstar Master, and Emilio Minichiello

    Benjamin Merlin Bumpus, Zoltan A. Kocsis, Jade Edenstar Master, and Emilio Minichiello. Structured Decompositions: Structural and Algorithmic Compositionality. 2025. arXiv:2207. 06091v7 [math.CT].url:https://arxiv.org/abs/2207.06091v7

    Pith/arXiv arXiv 2025

  11. [11]

    Edge Exploration of Temporal Graphs

    Benjamin Merlin Bumpus and Kitty Meeks. “Edge Exploration of Temporal Graphs”. In: 85 (2023), pp. 688–716.url:https://link.springer.com/article/10.1007/s00453-022- 01018-7#citeas

    work page doi:10.1007/s00453-022- 2023

  12. [12]

    Jacobs, Paul Knappe, and Jan Kurkofka.Canonical graph decompositions and local separations: From infinite coverings to a finite combinatorial theory

    Johannes Carmesin, Raphael W. Jacobs, Paul Knappe, and Jan Kurkofka.Canonical graph decompositions and local separations: From infinite coverings to a finite combinatorial theory

  13. [13]

    arXiv:2501.16170v1 [math.CO].url:https://arxiv.org/abs/2501.16170v1

    arXiv

  14. [14]

    Rewriting Structured Cospans: A Syntax for Open Systems

    Daniel Cicala. “Rewriting Structured Cospans: A Syntax for Open Systems”. PhD Thesis. University of California, Riverside, 2019.url:https : / / www . proquest . com / docview / 2308216336?pq-origsite=gscholar&fromopenview=true

    2019

  15. [15]

    The monadic second-order logic of graphs. I. Recognizable sets of finite graphs

    Bruno Courcelle. “The monadic second-order logic of graphs. I. Recognizable sets of finite graphs”. In:Information and Computation85.1 (1990), pp. 12–75.url:https : / / www . sciencedirect.com/science/article/pii/089054019090043H

    arXiv 1990

  16. [16]

    Open Systems: A Double Categorical Perspective

    Kenny Allen Courser. “Open Systems: A Double Categorical Perspective”. PhD Thesis. University of California, Riverside, 2020.url:https : / / www . proquest . com / docview / 2404393265?pq-origsite=gscholar&fromopenview=true

    2020

  17. [17]

    Jacobs, Paul Knappe, and Jan Kurkofka.Canonical graph decompositions via coverings

    Reinhard Diestel, Raphael W. Jacobs, Paul Knappe, and Jan Kurkofka.Canonical graph decompositions via coverings. 2025. arXiv:2207.04855v8 [math.CO].url:https://arxiv. org/abs/2207.04855v8

    Pith/arXiv arXiv 2025

  18. [18]

    Michelle Döring, Jessica Enright, Laura Larios-Jones, and George Skretas.FO and MSO Model Checking on Temporal Graphs. 2026. arXiv:2602 . 14592v1 [cs.DM].url:https : //arxiv.org/abs/2602.14592v1

    arXiv 2026

  19. [19]

    Hand, Laura Larios-Jones, and Kitty Meeks.Families of tractable problems with respect to vertex-interval-membership width and its generalisations

    Jessica Enright, Samuel D. Hand, Laura Larios-Jones, and Kitty Meeks.Families of tractable problems with respect to vertex-interval-membership width and its generalisations. 2026. arXiv: 2505.15699v5 [cs.DM].url:https://arxiv.org/abs/2505.15699v5

    Pith/arXiv arXiv 2026

  20. [20]

    Structural Parame- ters for Dense Temporal Graphs

    Jessica Enright, Samuel D. Hand, Laura Larios-Jones, and Kitty Meeks. “Structural Parame- ters for Dense Temporal Graphs”. In: vol. 306. Schloss Dagstuhl – Leibniz-Zentrum für Infor- matik, 2024, 52:1–52:15.url:https://drops.dagstuhl.de/entities/document/10.4230/ LIPIcs.MFCS.2024.52

    2024

  21. [21]

    Dynamic graph models

    F. Harary and G. Gupta. “Dynamic graph models”. In:Mathematical and Computer Modelling 25.7 (1997), pp. 79–87. 33

    1997

  22. [22]

    As Time Goes By: Reflections on Treewidth for Temporal Graphs

    Till Fluschnik, Hendrik Molter, Rolf Niedermeier, Malte Renken, and Philipp Zschoche. “As Time Goes By: Reflections on Treewidth for Temporal Graphs”. In:Treewidth, Kernels, and Algorithms. Springer International Publishing, 2020, pp. 49–77.url:http://dx.doi.org/ 10.1007/978-3-030-42071-0_6

    doi:10.1007/978-3-030-42071-0_6 2020

  23. [23]

    Brendan Fong.The Algebra of Open and Interconnected Systems. 2016. arXiv:1609.05382 [math.CT].url:https://arxiv.org/abs/1609.05382

    Pith/arXiv arXiv 2016

  24. [24]

    S-functions for graphs

    Rudolf Halin. “S-functions for graphs”. In:Journal of Geometry8 (1976), pp. 171–186.url: https://api.semanticscholar.org/CorpusID:120256194

    1976

  25. [25]

    Modern temporal network theory: a colloquium

    Petter Holme. “Modern temporal network theory: a colloquium”. In:The European Physical Journal B88.234 (2015)

    2015

  26. [26]

    Temporal networks

    Petter Holme and Jari Saramäki. “Temporal networks”. In:Physics Reports519.3 (2012), pp. 97–125

    2012

  27. [27]

    Compositionality

    Theo MV Janssen and Barbara H. Partee. “Compositionality”. In:Handbook of logic and language. Elsevier, 1997, pp. 417–473

    1997

  28. [28]

    A note on discrete Conduché fibrations

    Peter Johstone. “A note on discrete Conduché fibrations”. In:Theory and Applications of Categories5.1 (1999), pp. 1–11

    1999

  29. [29]

    Connectivity and Inference Problems for Temporal Networks

    David Kempe, Jon Kleinberg, and Amit Kumar. “Connectivity and Inference Problems for Temporal Networks”. In:Journal of Computer and System Sciences64.4 (2002), pp. 820–842

    2002

  30. [30]

    Adhesive Categories

    Stephen Lack and Paweł Sobociński. “Adhesive Categories”. In:Foundations of Software Sci- ence and Computation Structures. Ed. by Igor Walukiewicz. Springer Berlin Heidelberg, 2004, pp. 273–288

    2004

  31. [31]

    An algebraic framework for structured epidemic modelling

    Sophie Libkind, Andrew Baas, Micah Halter, Evan Patterson, and James P. Fairbanks. “An algebraic framework for structured epidemic modelling”. In:Philosophical Transac- tions of the Royal Society A: Mathematical, Physical and Engineering Sciences380.2233 (2022), p. 20210309. eprint:https : / / royalsocietypublishing . org / rsta / article - pdf / doi / 10...

    work page doi:10.1098/rsta.2021.0309 2022

  32. [32]

    Composing Behaviors of Networks

    Jade E. Master. “Composing Behaviors of Networks”. PhD Thesis. University of California, Riverside, 2021.url:https://www.proquest.com/docview/2565195807?pq- origsite= gscholar&fromopenview=true

    arXiv 2021

  33. [33]

    An Introduction to Temporal Graphs: An Algorithmic Perspective

    Othon Michail. “An Introduction to Temporal Graphs: An Algorithmic Perspective”. In:In- ternet Mathematics12 (2015)

    2015

  34. [34]

    Structured and Decorated Cospans from the Viewpoint of Double Category Theory

    Evan Patterson. “Structured and Decorated Cospans from the Viewpoint of Double Category Theory”. In:Electronic Proceedings in Theoretical Computer Science397 (2023).url:http: //dx.doi.org/10.4204/EPTCS.397.13. 34

    work page doi:10.4204/eptcs.397.13 2023

  35. [35]

    Open Markov Processes and Reaction Networks

    Blake Stephen Swistock Pollard. “Open Markov Processes and Reaction Networks”. PhD Thesis. University of California, Riverside, 2017.url:https://www.proquest.com/docview/ 1972046592?pq-origsite=gscholar&fromopenview=true

    2017

  36. [36]

    Graph minors. XVII. Taming a Vortex

    Neil Robertson and P.D. Seymour. “Graph minors. XVII. Taming a Vortex”. In:Journal of Combinatorial Theory, Series B77.1 (1999), pp. 162–210

    1999

  37. [37]

    Graph minors. III. Planar Tree-Width

    Neil Robertson and P.D Seymour. “Graph minors. III. Planar Tree-Width”. In:Journal of Combinatorial Theory, Series B36.1 (1984), pp. 49–64.url:https://www.sciencedirect. com/science/article/pii/0095895684900133

    arXiv 1984

  38. [38]

    Spivak, and Christina Vasilakopoulou.Dynamical Systems and Sheaves

    Patrick Schultz, David I. Spivak, and Christina Vasilakopoulou.Dynamical Systems and Sheaves. 2019. arXiv:1609 . 08086v4 [math.CT].url:https : / / arxiv . org / abs / 1609 . 08086v4

    2019

  39. [39]

    Compositionality

    Zoltán Gendler Szabó and Richmond H. Thomason. “Compositionality”. In:Philosophy of Language. Cambridge Textbooks in Linguistics. Cambridge University Press, 2018, pp. 41–63

    2018

  40. [40]

    Oxford University Press, 2012

    Markus Werning, Wolfram Hinzen, and Edouard Machery, eds.The Oxford handbook of com- positionality. Oxford University Press, 2012. 35

    2012