Pith. sign in

REVIEW 4 major objections 4 minor 40 references

Ultracategories via Kan extensions of relative monads

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The left oplax Kan extension of a relative 2-monad is a pseudomonad with the same colax algebras.

desk verdict A plausible and useful unrelativisation theorem for relative 2-monads, with a genuinely new application to weak ultracategories; the coherence proofs are too compressed for the central claim to be accepted as is. read the letter →

arxiv 2506.09788 v1 pith:SZSNFO6K submitted 2025-06-11 math.CT

Umberto Tarantino , Joshua Wrigley This is my paper · ORCID
classification math.CT MSC 18N1518C1518C2003G3003C20
keywords relative2-monadsleftoplaxKanextensionspseudomonadscolaxalgebrasultracategoriesultracompletionultrafiltermonadprimecategories
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

Many structures in mathematics are naturally presented as algebras for a relative monad, where the underlying functor is not an endofunctor, and it is not obvious whether they are algebras for an ordinary monad. This paper proves a general 'unrelativisation' theorem: under three hypotheses on the root 2-functor $J$, the left oplax Kan extension of a $J$-relative 2-monad $T$ along $J$ is a pseudomonad on $\mathsf{CAT}$, and its colax algebras are exactly the colax algebras of $T$. Applying the theorem to the relative ultrafilter 2-monad on the inclusion of sets as discrete categories produces an 'ultracompletion' pseudomonad whose colax algebras are the weak ultracategories, a weakening of Lurie's ultracategories that keeps the same morphisms and therefore the same logical reconstruction theorems. If the theorem is right, weak ultracategories are not an ad hoc axiomatisation but the universal algebraic consequence of the ultrafilter monad, and the same construction yields prime categories from the prime upper filter monad on posets.

What carries the argument

The central object is the left oplax Kan extension $\widetilde{T}:\mathsf{CAT}\to\mathsf{CAT}$ of a $\mathsf{CAT}$-valued 2-functor $T$ along a 2-functor $J:\mathcal{B}\to\mathsf{CAT}$, defined by a universal property phrased through lax ends: maps out of $\widetilde{T}C$ correspond to lax transformations $[J-,C]\Rightarrow[T-,C]$. Concretely, $\widetilde{T}C$ is the category of triples $(b,h:Jb\to C,\nu\in Tb)$, with arrows given by equivalence classes of pairs of a 1-cell and a 2-cell. This object carries the argument because it converts a relative monad, whose 'multiplication' cannot be iterated, into an endofunctor whose structure can be iterated. The pseudomonad structure is assembled on the skew-monoidal category $\mathrm{Oplax}[\mathcal{B},\mathsf{CAT}]$ whose composition is $\widetilde{F}\circ G$; the critical structural map $s_{F,G}:\widetilde{F}\circ\widetilde{G}\Rightarrow\widehat{\widetilde{F}\circ G}$ is defined by taking the oplax colimits that the hypotheses require $J$ to preserve. The unit and multiplication of $\widetilde{T}$ are then induced by these universal properties, and the coherence modifications come from the same oplax colimits.

What would settle it

Exhibit a $J$-relative 2-monad satisfying hypotheses (1) and (3) of Section 4 for which $J$ fails to preserve a single oplax colimit of shape $(Jb)^{\mathrm{op}}$ and show that the resulting left oplax Kan extension does not admit a pseudomonad structure with colax algebras isomorphic to those of $T$. In the concrete case of $\beta$, the decisive check is the free weak ultracategory $\widetilde{\beta}(1)$ of Example 2.8: the paper's proof of Corollary 6.1 stands or falls on the associativity and unit modifications $p,q,x$ being invertible there.

Watch

Extended reading notes

Core claim

The paper's central claim is that left oplax Kan extensions 'unrelativise' relative 2-monads while preserving colax algebras. Concretely, for a 2-functor $J:\mathcal{B}\to\mathsf{CAT}$ and a $J$-relative 2-monad $T$, the left oplax Kan extension $\widetilde{T}:\mathsf{CAT}\to\mathsf{CAT}$ is a 2-endofunctor, and under the hypotheses that $\mathcal{B}$ has a terminal object preserved by $J$, that $J$ is 2-fully faithful, and that $\mathcal{B}$ admits oplax colimits of all diagrams of shape $(Jb)^{\mathrm{op}}$ preserved by $J$, $\widetilde{T}$ carries a pseudomonad structure. Theorem 5.4 states that the 2-categories of colax algebras of $T$ and of $\widetilde{T}$ are isomorphic for each flavour of morphisms, colax, lax, pseudo, or strict. In the motivating example, $T$ is the relative ultrafilter 2-monad $J\beta$ with root $J:\mathsf{Set}\to\mathsf{CAT}$, and the resulting pseudomonad $\widetilde{\beta}$ sends a category $C$ to the category of formal ultraproducts, so weak ultracategories are exactly its colax algebras (Corollary 6.1). The same template gives prime categories as colax algebras for the relative prime upper filter 2-monad.

Load-bearing premise

The whole construction depends on the root 2-functor $J$ preserving every oplax colimit of diagrams of the form $(Jb)^{\mathrm{op}}$ for $b\in \mathcal{B}$; if even one such colimit is not preserved, the multiplication of the pseudomonad $\widetilde{T}$ cannot be built and the algebra comparison in Theorem 5.4 is not guaranteed.

Editorial extensions

If this is right

  • Any relative 2-monad meeting the three hypotheses of Section 4 has an ordinary pseudomonad with the same colax algebras, so the distinction between 'relative' and 'ordinary' monadic structure collapses in these cases.
  • Weak ultracategories are the colax algebras of a pseudomonad on $\mathsf{CAT}$ constructed purely from the ultrafilter monad by Kan extensions, giving a universal explanation of the ultracategory axioms.
  • Because ultrafunctors and transformations are exactly the colax and pseudo morphisms of this pseudomonad, the reconstruction results for Lurie ultracategories transfer to the weaker setting.
  • The category of ultrasets is isomorphic to the category of compact Hausdorff spaces, recovering the known ultracategory analogue of Manes' theorem.
  • Prime categories, defined from the prime upper filter monad on posets, are likewise colax algebras for a pseudomonad, opening an ordered or Priestley-style version of the theory.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: the same unrelativisation recipe applies to any codensity monad on a 2-fully faithful inclusion into $\mathsf{CAT}$ whose root preserves enough oplax colimits, so the construction is a general machine for producing 'completion' pseudomonads from relative monads.
  • Beyond the paper: the comparison in Remark 6.2 suggests a precise test of whether Lurie's ultracategories are exactly the normal colax algebras of $\widetilde{\beta}$; if true, the heavy axioms of Definition 2.4 would be a normality condition rather than extra algebra.
  • Beyond the paper: the paper establishes isomorphism of algebra 2-categories, but not bi-monadicity of the forgetful 2-functor; a natural next step is to ask whether $\mathsf{WUlt}\to\mathsf{CAT}$ is monadic in the strong 2-categorical sense.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper develops a framework for 'unrelativising' relative 2-monads by taking left oplax Kan extensions. Under assumptions on a root 2-functor J:B->CAT (terminal object, preservation of oplax colimits of a certain shape, and 2-full faithfulness), the authors construct a pseudomonad structure on the left oplax Kan extension eT of a J-relative 2-monad T, and prove an isomorphism between the 2-categories of colax algebras for T and for eT (Theorems 4.13 and 5.4). They apply this to the inclusion Set->CAT and the ultrafilter monad to obtain that weak ultracategories are colax algebras for a pseudomonad eβ on CAT whose underlying 2-endofunctor is obtained by combining right Kan extensions and left oplax Kan extensions (Corollary 6.1). They also sketch an analogous application to prime categories. The paper contains explicit constructions for CAT-valued left oplax Kan extensions (Definition 3.7, Proposition 3.10), a skew-monoidal structure on Oplax[B,CAT] (Section 4.2), and detailed but incomplete proofs in the appendices.

Significance. If the main theorems are correct, the paper gives a genuinely useful mechanism for constructing pseudomonads from relative 2-monads and provides a clean conceptual explanation of weak ultracategories as algebras for a universally induced pseudomonad. The explicit description of eT and the connection to the codensity monad of the ultrafilter monad are attractive and likely to be of interest to category theorists working on monads and on ultracategories. The paper also makes a good-faith effort to supply explicit constructions and several detailed calculations, and the application to weak ultracategories is concrete and falsifiable. However, the central proofs of pseudomonad coherence and of the algebra comparison are not carried out in the manuscript to the standard required for publication; several load-bearing verifications are either asserted via universal properties, deferred to an unpublished preprint, or relegated to 'similar arguments'.

major comments (4)
  1. [§4.3, Theorem 4.13] The proof of Theorem 4.13 consists of the sentence: 'The two coherence conditions spelled out in [Lac00] are satisfied since all arrows describing them are uniquely induced by the universal properties of oplax colimits.' This does not establish the pentagon and triangle axioms for the modifications p, q, and x defined in Definition 4.12. In particular, the associator x is defined by a composite pasting of five modifications, and one must verify that the two pasted composites in [Lac00] are equal at the level of 2-cells. The universal property of oplax colimits gives uniqueness of 1-cells and 2-cells induced from a cocone, but it does not by itself compare the specific pastings that define the two sides of the coherence axioms. This is a load-bearing gap: without a verified coherence theorem, the pseudomonad structure on eT is not established.
  2. [§4.2(5) and Proposition 4.9(c)] The proof of Proposition 4.9(c) relies on the assertion 'Since oplax colimits commute with oplax colimits (cf. [Bon24])' and on the identification of an iterated oplax colimit with a single oplax colimit over (Jℓ)^op. This commutation result is not proved or even stated precisely in the paper, and [Bon24] is an unpublished preprint. Moreover, the identification uses that J preserves the relevant oplax colimits and that the diagram shapes match; this is not an immediate consequence of assumption (2) in Section 4. Since Proposition 4.9(c) provides the invertible modification that underlies the associativity of the skew-monoidal structure and is later used in the definition of the pseudomonad associator x, the gap propagates to Theorem 4.13. The authors should either prove the needed commutation lemma in the appendix or cite a published source with a precise statement.
  3. [§4.2(5), definition of s_{F,G}] The definition of the strict transformation s_{F,G} maps equivalence classes of triples (f,α,φ) in eF(eGC), but the paper omits the verification that the assignment is well-defined on the equivalence relation of Definition 3.7, as well as the verification of naturality in F and G. This is not a routine check: one must show that the induced 1-cell \bar f and the 2-cells \bar α and θ are independent of the chosen representative, and that the families used in the universal property of the oplax colimit satisfy the required cocone conditions. The paper states only 'We omit the routine verification that the above functors are well-defined on equivalence classes and define a strict transformation...'. Since s_{T,T} is used in Definition 4.12(2) to define the multiplication μ^♯, any failure of well-definedness would invalidate the pseudomonad structure and the algebra comparison in Theorem 5.4.
  4. [Appendix B and Theorem 5.4] Lemmas B.1–B.4 each verify only one axiom (usually the first) and state that the remaining axioms 'can be shown with similar arguments.' Given the complexity of Definitions 5.1 and 1.4—especially axiom (b) for colax morphisms, which involves pastings involving u, m, and eT^2u—the omitted verifications are substantial and not purely formal. Furthermore, the proof of Theorem 5.4 claims an isomorphism of 2-categories, but Corollaries 5.8 and 5.10 assert that the composites Θ∘Σ and Σ∘Θ are identities on morphisms 'immediately' without checking that the induced 2-cells υ' and u' coincide with the original ones. These points need to be proved in detail, or the statement of Theorem 5.4 should be weakened to an equivalence. Since the application to weak ultracategories is the central motivation, the algebra comparison must be on solid ground.
minor comments (4)
  1. [§4.2(2)] The notation ι_{F,Jb} is used in the definition of the right unit r_F, but no ι is introduced in Section 3; the reader is presumably meant to use the universal oplax transformation ζ of Remark 3.11. Please define the notation or replace it with ζ.
  2. [§4.2(5)] In the definition of the action of s_{F,G} on arrows, the natural transformation θ is described by a component formula that refers to χ_x and w_x defined in the preceding paragraph, but no pasting diagram is provided. A diagram would considerably help the reader verify that θ is indeed a natural transformation.
  3. [Appendix A, Definition A.3] There is a typo: 'folllowing' should be 'following', and in item (1) the unique 1-cell u is said to live 'in B', but it should be 'in A' since it is a 1-cell from the oplax coend to the vertex A.
  4. [§5.1.1] The definition of the natural transformation m says 'We omit the routine proof of the naturality of m.' Given that m is one of the two structure 2-cells of a colax algebra for eT, this proof should at least be sketched; the current omission makes the verification of the algebra axioms in Lemma B.1 harder to follow.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the algebra-comparison theorem is a substantive proof, not a definitional restatement.

full rationale

After walking the claimed derivation chain, I find no step in which a prediction or first-principles result is equivalent by construction to its inputs. The central universal property (Definition 3.3) is proved independently for CAT-valued 2-functors (Propositions 3.10, A.5, A.6), and the pseudomonad structure is built explicitly via the transformations d, s_{F,G}, l, r, a in Sections 4.2–4.3. The algebra-comparison Theorem 5.4 is not a definitional restatement: although the bijection between extension operators and functors eTC→C is part of the left-oplax-Kan-extension universal property, the verification that the colax algebra axioms correspond is carried out in Lemmas 5.7–5.9 and B.1–B.4. No constants are fitted, no parameter is renamed as a prediction, and the paper does not rely on a self-citation chain. The citations to [ACU15], [Ham25], [Lur18], and [Mak87] are external anchors; [Bon24] is cited for commutation of oplax colimits, which is a nontrivial external dependency but not a circular one. The proof of Theorem 4.13 is compressed to a one-sentence universal-property argument and several appendix lemmas end with 'similar arguments'; these are rigor gaps, not circular reductions.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

Pure mathematics: the paper introduces no fitted constants or physical entities. The load-bearing axioms are the standing hypotheses on the root 2-functor J and the external 2-categorical machinery (lax end calculus, commutation of oplax colimits) cited from preprints. The 'weak ultracategory' and 'left oplax Kan extension' are new definitions, not unexplained postulates.

assumptions (5)
  • standard math Standard 2-category theory (2-categories, 2-functors, 2-natural transformations, modifications, CAT).
    Used throughout, e.g., Sections 1.1, 3, 4.
  • standard math The ultrafilter monad β on Set is a monad and the codensity monad of FinSet -> Set.
    Invoked in Corollary 6.1; cited to [KG71], [Man76], [Lei13].
  • standard math Lax end and oplax coend calculus as developed in [Hir22], including the Fubini-style identities used in Proposition A.4.
    Foundation of left oplax Kan extensions and their universal property; cited in Section 3.1 and Appendix A. This is a preprint dependency.
  • domain assumption Hypotheses (1)-(3) of Section 4 hold for the root 2-functor J: B -> CAT, namely: B has a terminal object preserved by J; B admits and J preserves oplax colimits of shape (Jb)^op; J is 2-fully faithful.
    These are the standing assumptions of Theorem 4.13 and Theorem 5.4; they hold for Set -> CAT but fail for dense roots used in ACU15.
  • standard math Oplax colimits commute with oplax colimits, as cited to [Bon24].
    Used in Proposition 4.9(c) to compare two iterated oplax colimits; correctness depends on this preprint.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Ultracategories via Kan extensions of relative monads." pith.science (2026). https://pith.science/paper/SZSNFO6K

@misc{pith2026250609788,
  author       = {Pith},
  title        = {Pith review of: Ultracategories via Kan extensions of relative monads},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SZSNFO6K}},
  note         = {Machine review of arXiv:2506.09788}
}
read the original abstract

Many structured categories of interest are most naturally described as algebras for a relative monad, but turn out nonetheless to be algebras for an ordinary monad. We show that, under suitable hypotheses, the left oplax Kan extension of a relative 2-monad on categories yields a pseudomonad having the same category of colax algebras. In particular, we apply this to the study of ultracategories to recover the 'ultracompletion' pseudomonad.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 36 canonical work pages

  1. [1]

    Altenkirch, J

    T. Altenkirch, J. Chapman, and T. Uustalu. Monads need not be endofunctors. In L. Ong, editor, Foundations of Software Science and Computational Structures , pages 297--311, Berlin, Heidelberg, 2010. Springer Berlin Heidelberg

    work page 2010

  2. [2]

    Altenkirch, J

    T. Altenkirch, J. Chapman, and T. Uustalu. Monads need not be endofunctors. Logical Methods in Computer Science , Volume 11, Issue 1, 2015

    work page 2015

  3. [3]

    Arkor and D

    N. Arkor and D. McDermott. The formal theory of relative monads. Journal of Pure and Applied Algebra , 228(9):107676, September 2024

    work page 2024

  4. [4]

    Adámek and L

    J. Adámek and L. Sousa. D-ultrafilters and their monads. Advances in Mathematics , 377:107486, 2021

    work page 2021

  5. [5]

    Arkor, P

    N. Arkor, P. Saville, and A. Slattery. Bicategories of algebras for relative pseudomonads, 2025. arXiv:2501.12510

    arXiv 2025

  6. [6]

    P. Bonart. Towards a general theory of dependent sums, 2024. arXiv:2404.08139

    work page Pith review arXiv 2024

  7. [7]

    Bozapalides

    S. Bozapalides. Les fins cartésiennes généralisées. Archivum Mathematicum , 13(2):75--87, 1977

    work page 1977

  8. [8]

    Bozapalides

    S. Bozapalides. Some remarks on lax-presheafs. Illinois J. Math , 24(4):676--680, 1980

    work page 1980

Show all 40 references
  1. [9]

    C. C. Chang and H. J. Keisler. Model Theory . North-Holland, 1973

    1973

  2. [10]

    Di Liberti

    I. Di Liberti . The geometry of coherent topoi and ultrastructures, 2022. arXiv:2211.03104

    2022 arXiv

  3. [11]

    Eilenberg and J

    S. Eilenberg and J. C. Moore. Adjoint functors and triples . Illinois Journal of Mathematics , 9(3):381 -- 398, 1965

    1965

  4. [12]

    Fiore, N

    M. Fiore, N. Gambino, M. Hyland, and G. Winskel. Relative pseudomonads, kleisli bicategories and substitution monoidal structures. Selecta Mathematica , 24:2791--2830, 2018

    2018

  5. [13]

    B. Flagg. Algebraic theories of compact pospaces. Topology and its Applications , 77(3):277--290, 1997

    1997

  6. [14]

    R. Garner. Ultrafilters, finite coproducts and locally connected classifying toposes. Annals of Pure and Applied Logic , 171(10):102831, December 2020

    2020

  7. [15]

    A. Hamad. Ultracategories as colax algebras for a pseudo-monad on CAT , 2025. arXiv:2502.20597

    2025 arXiv

  8. [16]

    K. Hirata. Notes on lax ends, 2022. arXiv:2210.01522

    2022 arXiv

  9. [17]

    P. T. Johnstone. Sketches of an Elephant: A Topos theory Compendium . Oxford University Press, 2002

    2002

  10. [18]

    Johnson and D

    N. Johnson and D. Yau. 2-Dimensional Categories . Oxford University Press, 2020

    2020

  11. [19]

    G. M. Kelly. Basic concepts of enriched category theory. Reprints in Theory and Applications of Categories , (10):1--136, 2005

    2005

  12. [20]

    J. F. Kennison and D. Gildenhuys. Equational completion, model induced triples and pro-objects. Journal of Pure and Applied Algebra , 1(4):317--346, 1971

    1971

  13. [21]

    S. Lack. A coherent approach to pseudomonads. Advances in Mathematics , 152(2):179--202, 2000

    2000

  14. [22]

    D. Lascar. On the category of models of a complete theory. Journal of Symbolic Logic , 47(2):249–266, 1982

    1982

  15. [23]

    Leinster

    T. Leinster. Basic bicategories, 1998. arXiv:9810017

    1998

  16. [24]

    Leinster

    T. Leinster. Codensity and the ultrafilter monad. Theory and Applications of Categories , 28(13):332--370, 2013

    2013

  17. [25]

    Loregian

    F. Loregian. (Co)end Calculus . Cambridge University Press, June 2021

    2021

  18. [26]

    J. Lurie. Ultracategories, 2018. https://www.math.ias.edu/ lurie/papers/Conceptual.pdf

    2018

  19. [27]

    S. MacLane. Categories for the Working Mathematician . Springer-Verlag, New York, 1971. Graduate Texts in Mathematics, Vol. 5

    1971

  20. [28]

    M. Makkai. Stone duality for first order logic. Advances in Mathematics , 65(2):97--170, 1987

    1987

  21. [29]

    M. Makkai. Strong conceptual completeness for first-order logic. Annals of Pure and Applied Logic , 40(2):167--215, 1988

    1988

  22. [30]

    E. G. Manes. A triple theoretic construction of compact algebras. In B. Eckmann, editor, Seminar on Triples and Categorical Homology Theory , pages 91--118. Springer Berlin Heidelberg, 1969

    1969

  23. [31]

    E. G. Manes. Algebraic Theories . Springer, 1976

    1976

  24. [32]

    Marmolejo

    F. Marmolejo. Ultraproducts and continuous families of models . PhD thesis, Dalhousie University, 1995

    1995

  25. [33]

    Marmolejo and R

    F. Marmolejo and R. J. Wood. Monads as extension systems -- no iteration is necessary. Theory and Applications of Categories , 24(4):84--113, 2010

    2010

  26. [34]

    Marmolejo and R

    F. Marmolejo and R. J. Wood. No-iteration pseudomonads. Theory and Applications of Categories , 28(14):371--402, 2013

    2013

  27. [35]

    Moraschini, J

    T. Moraschini, J. J. Wannenburg, and K. Yamamoto. Elementary equivalence in positive logic via prime products. The Journal of Symbolic Logic , pages 1--18, 2023

    2023

  28. [36]

    Rosolini

    G. Rosolini. Ultracompletions, 2024. talk at Category Theory 2024, https://www.usc.gal/regaca/ct2024/files/Abstracts/Rosolini.pdf

    2024

  29. [37]

    R. Street. Two constructions on Lax functors. Cahiers de topologie et g\'eom\'etrie diff\'erentielle , 13(3):217--264, 1972

    1972

  30. [38]

    R. Street. Fibrations and Yoneda's lemma in a 2-category. In Category Seminar , pages 104--133. Springer Berlin Heidelberg, 1974

    1974

  31. [39]

    Szlachányi

    K. Szlachányi. Skew-monoidal categories and bialgebroids. Advances in Mathematics , 231(3–4):1694--1730, October 2012

    2012

  32. [40]

    M. Štěpán. Lax structures in 2-category theory . PhD thesis, Masaryk University, 2024

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.