REVIEW 6 minor 32 references
An equivalence theorem for algebraic and functorial QFT
T0 review · 0 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash
Pith's one-line read This paper proves that algebraic and functorial quantum field theories on Lorentzian spacetimes are equivalent categories once time-slice and additivity are imposed.
desk verdict Real equivalence theorem between additive time-slice AQFTs and FQFTs over a new Lorentzian bordism pseudo-operad; the Remark 4.2 morphism caveat is a boundary limitation, not a flaw in the proof. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is the globally hyperbolic Lorentzian bordism pseudo-operad $\mathrm{LBop}_m$. Its operations are $n$-to-$1$ bordisms: a globally hyperbolic Lorentzian manifold $N$ with a tuple of causally disjoint partial Cauchy surfaces in its past and one full Cauchy surface in its future, each equipped with a collar region inside a larger spacetime. Composition is defined by gluing two bordisms along the intersection of the collar regions around the shared intermediate Cauchy surfaces, after trimming overhanging collar parts so that the pushout is again a globally hyperbolic Lorentzian manifold. A fibrancy result for this pseudo-operad lets the authors replace it by an ordinary operad $\tau(\mathrm{LBop}_m)$, turning functorial QFTs into ordinary algebra-valued multifunctors; this truncation is the technical bridge on which the equivalence theorem runs.
What would settle it
Compute the pushout from the gluing lemma for two flat two-dimensional slabs glued along a narrow collar region that satisfies the chronological-past and causal-past conditions, and check whether the resulting Lorentzian manifold is globally hyperbolic by testing whether every inextendible timelike curve meets the candidate Cauchy surface. A concrete failure—a closed timelike curve, a missing Cauchy surface, or a boundary point in the image of the collar map—would invalidate the gluing lemma and therefore the equivalence theorem.
Extended reading notes
Core claim
The central claim is Theorem 5.7: for every spacetime dimension, the categories $\mathrm{AQFT}^{W,\mathrm{add}}_m$ and $\mathrm{FQFT}^{W,\mathrm{add}}_m$ are equivalent. The forward functor sends an algebraic QFT $A$ to a functorial QFT whose value on a Cauchy surface $(M,\Sigma)$ is the algebra $A(M)$, with the time-slice axiom used to turn collar comparisons into isomorphisms. The inverse functor sends a functorial QFT $F$ to an algebraic QFT whose value on a spacetime $M$ is the filtered colimit of $F$ over the category of Cauchy surfaces of $M$. Additivity is used to assemble these colimits into operations on arbitrary causally disjoint tuples of regions, where the images of several partial Cauchy surfaces may not extend to a single later Cauchy surface. The paper emphasizes that both hypotheses are needed, and notes that the same pair of hypotheses appears in the earlier equivalence between algebraic QFTs and prefactorization algebras.
Load-bearing premise
The theorem rests on the gluing lemma: when two globally hyperbolic bordisms are composed, the pushout along their common collar must again be a globally hyperbolic Lorentzian manifold; if this gluing ever produced a spacetime with bad causal behavior, the bordism operad and hence the equivalence would lose its domain.
Editorial extensions
If this is right
- Any additive, time-slice algebraic QFT gives an explicit functorial QFT on the bordism operad, assigning the same observable algebra to each Cauchy surface and using the time-slice axiom to define evolution maps.
- Any additive, time-slice functorial QFT gives an explicit algebraic QFT, with the algebra on a spacetime recovered as a filtered colimit of functorial QFT values over its Cauchy surfaces; additivity then extends this to arbitrary causally disjoint tuples of regions.
- The equivalence is an equivalence of categories, not merely a bijection on objects, so natural transformations on one side correspond exactly to natural transformations on the other.
- Because the algebraic side automatically enforces Einstein causality, the equivalent functorial side inherits the same causality condition in this framework.
- The construction strictly generalizes the earlier bordism pseudo-category approach by allowing partial Cauchy surfaces, so spatially local data are captured by functorial QFTs rather than only global topology.
- The equivalence relies on both the time-slice axiom and additivity; the paper states explicitly that it does not expect either hypothesis to be removable.
Reading between the lines
- Because the proof gives explicit quasi-inverse functors, it should transport examples: any known additive time-slice algebraic QFT supplies a functorial QFT on the bordism operad, and conversely any such functorial QFT supplies a local net. The paper does not spell out named examples.
- The equivalence suggests that the time-slice and additivity conditions mark exactly the overlap between the two axiomatizations; dropping either should break one of the two composite functors from being the identity, which could be checked on concrete examples.
- The partial $n$-to-$1$ Cauchy surface bordisms mimic pair-of-pants products, so one could use them to formulate operator-product-like multiplicative structures directly from Lorentzian bordisms without passing through an algebraic QFT.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper defines a new pseudo-operad LBop_m of globally hyperbolic Lorentzian bordisms whose operations go from tuples of causally disjoint partial Cauchy surfaces to a later full Cauchy surface, and uses it to introduce a notion of functorial QFT (FQFT) as an algebra over the truncated operad τ(LBop_m) with values in unital associative algebras. On the AQFT side, the paper uses the known operadic formulation of locally covariant AQFTs. The main result, Theorem 5.7, constructs explicit quasi-inverse functors F and A and establishes an equivalence of categories AQFT^{W,add}_m ≃ FQFT^{W,add}_m for theories satisfying the time-slice axiom and the additivity property. The proof is supported by two appendices, one on pseudo-operads internal to groupoids and one on Lorentzian geometric gluing details.
Significance. If the result stands, it is a substantial conceptual bridge: it shows that, under the standard hypotheses of time-slice and additivity, the observables-based and bordism-based axiomatizations of Lorentzian QFT carry exactly the same information. The introduction of partial Cauchy surfaces as sources of bordisms is a natural and well-motivated resolution of the obstruction to topology change in globally hyperbolic bordisms, and the explicit quasi-inverse functors make the equivalence concrete rather than merely formal. The paper is also commendably transparent about its limitations, especially Remark 4.2, which acknowledges that non-invertible FQFT morphisms are added by hand and do not currently have a pseudo-operadic interpretation.
minor comments (6)
- [Abstract and Definition 4.1 / Remark 4.2] The abstract and introduction state that FQFTs are pseudo-multifunctors, but Definition 4.1 actually defines the category of FQFTs using ordinary multifunctors on the truncation τ(LBop_m), with non-invertible multinatural transformations added by hand as in Remark 4.2. Please qualify the main claim throughout: Theorem 5.7 is an equivalence for the category defined in Definition 4.1, and the pseudo-operadic interpretation currently covers only the invertible morphisms. This is not a flaw in the theorem, but the current wording overstates the scope of the categorical equivalence.
- [Proposition B.5, Eq. (3.16)] The proof that the pushout N^-_0 ⊔_{V01∩V10} N^+_1 is again an object of Loc_m is very terse. In particular, the assertion that every inextensible future-pointing timelike curve in the pushout meets ι+ι11(Σ2) exactly once should be justified in detail, including curves that cross the gluing interface, and the boundary-point argument with [ST11, Lemma 2.23] should be expanded. Since the well-definedness of the operadic composition, and hence of the entire FQFT category, depends on this gluing lemma, a more complete proof would strengthen the paper.
- [Lemma 5.2, proof of additivity of FA] The finality of the forgetful functor RC(M,Σ) → RC_{I^-_M(Σ)} in (5.6) is asserted as 'easily checked'. This finality is essential for the additivity transfer from A to FA, so please provide a proof or at least a precise reference to a lemma; a short argument using Proposition B.6 would suffice.
- [Lemma 5.6, proof of additivity of AF] In the proof of Lemma 5.6, the finality of the functor ∫_{RC_M} Q → Q_M is stated without proof. Since this finality is used to identify the colimit in the additivity condition for AF, please supply the argument or include it as an explicit lemma in Appendix B.
- [Theorem 5.7, proof, Eq. (5.25b)] The diagram in (5.25b) is dense and is the key step showing that the composite F∘A reproduces the original FQFT action. The inversion of the bottom-row isomorphisms via [BMS25, Lemma 3.4] should be spelled out more explicitly, because the reader needs to see precisely why the composite of the bottom row is exactly F([N,ι0,ι1]) and not merely an isomorphic map.
- [General presentation] The paper relies on several 'easily verified' or 'straightforward' checks, for example in Constructions 5.1 and 5.4 and in the proof of Theorem A.7. These are acceptable in a research paper, but given the complexity of the pseudo-operadic coherence data, it would be helpful to indicate in each case which axiom of Definition A.1 or A.3 is being verified.
Circularity Check
No circular derivation: Theorem 5.7 is proved by explicit quasi-inverse functors, and its technical self-citations are independent prior results.
full rationale
The claimed equivalence AQFTW,add_m ≃ FQFTW,add_m is not a restatement of its inputs. The functors F and A are constructed explicitly (Constructions 5.1 and 5.4): on objects, FA(M,Σ)=A(M), while AF(M)=colim over the Cauchy-surface category Σ_M of F(M,Σ). The proof that A∘F equals the identity reduces to the observation that FA| is the constant functor with value A(M), and the proof that F∘A is naturally isomorphic to the identity uses the time-slice axiom to identify the canonical inclusion F(M,Σ)→colim(F|); neither step assumes the theorem. The time-slice and additivity hypotheses are parallel assumptions on both sides, not fitted parameters, and no data or normalized quantities are fitted to a subset. The paper cites [BMS25] for technical results, most notably Proposition 3.3 (fibrancy of the bordism pseudo-operad) and Lemma 3.4 (used in the proof of Theorem 5.7), but these are previously established statements about the Lorentzian bordism pseudo-category and do not assert the present equivalence; they are independent support rather than a self-citation chain that forces the result. Remark 4.2 explicitly limits the pseudo-operadic interpretation of FQFT morphisms to the invertible case, acknowledging that non-invertible morphisms are added by hand; this is a scope caveat about the category compared, not a circular reduction of the equivalence. The central derivation is therefore self-contained in the sense of not presupposing its conclusion.
Assumptions & free parameters
assumptions (6)
- domain assumption Bernal-Sanchez metric splitting: every globally hyperbolic Lorentzian manifold is diffeomorphic to R × Σ with Cauchy surfaces Σ.
- domain assumption Every achronal compact subset of a globally hyperbolic spacetime extends to a Cauchy surface (BS06, Proposition 3.6).
- standard math The categories RC_M and RC(M,Σ) are filtered and the relevant forgetful functors are final (Lemma B.3, Proposition B.6, Lemma B.7).
- standard math The 2-adjunction tau: PsOp_fib ⇄ Op(2,1): iota from Theorem A.7 is valid, including the companion construction for fibrancy.
- domain assumption T is a cocomplete closed symmetric monoidal category and AlguAs(T) inherits filtered colimits; the monoidal product preserves colimits.
- standard math The pushout of two bordisms along matched collar regions exists in Loc_m and is globally hyperbolic (Proposition B.5).
Cite this review
Pith. "Pith review of An equivalence theorem for algebraic and functorial QFT." pith.science (2026). https://pith.science/paper/DA46D44B
@misc{pith2026250415759,
author = {Pith},
title = {Pith review of: An equivalence theorem for algebraic and functorial QFT},
year = {2026},
howpublished = {\url{https://pith.science/paper/DA46D44B}},
note = {Machine review of arXiv:2504.15759}
}
read the original abstract
This paper develops a novel approach to functorial quantum field theories (FQFTs) in the context of Lorentzian geometry. The key challenge is that globally hyperbolic Lorentzian bordisms between two Cauchy surfaces cannot change the topology of the Cauchy surface. This is addressed and solved by introducing a more flexible concept of bordisms which provide morphisms from tuples of causally disjoint partial Cauchy surfaces to a later-in-time full Cauchy surface. They assemble into a globally hyperbolic Lorentzian bordism pseudo-operad, generalizing the geometric bordism pseudo-categories of Stolz and Teichner. The associated FQFTs are defined as pseudo-multifunctors into a symmetric monoidal category of unital associative algebras. The main result of this paper is an equivalence theorem between such globally hyperbolic Lorentzian FQFTs and algebraic quantum field theories (AQFTs), both subject to the time-slice axiom and a mild descent condition called additivity.
Reference graph
Works this paper leans on
-
[1]
M. Atiyah, ``Topological quantum field theories,'' Inst.\ Hautes \'E tudes Sci.\ Publ.\ Math.\ 68, 175--186 (1988)
work page 1988
-
[2]
C. B\"ar, N. Ginoux and F. Pf\"affle, Wave equations on Lorentzian manifolds and quantization , Eur.\ Math.\ Soc., Z\"urich (2007) [arXiv:0806.1036 [math.DG]]
arXiv 2007
- [3]
- [4]
-
[5]
Quantization of Lorentzian free BV theories: factorization algebra vs algebraic quantum field theory
M. Benini, G. Musante and A. Schenkel, ``Quantization of Lorentzian free BV theories: factorization algebra vs algebraic quantum field theory,'' Lett.\ Math.\ Phys.\ 114, no.\ 1, 36 (2024) [arXiv:2212.02546 [math-ph]]
work page Pith review arXiv 2024
-
[6]
M. Benini, M. Perin and A. Schenkel, ``Model-independent comparison between factorization algebras and algebraic quantum field theory on Lorentzian manifolds,'' Commun.\ Math.\ Phys.\ 377, 971--997 (2020) [arXiv:1903.03396 [math-ph]]
work page Pith review arXiv 2020
- [7]
-
[8]
Operads, homotopy theory and higher categories in algebraic quantum field theory
M. Benini and A. Schenkel, ``Operads, homotopy theory and higher categories in algebraic quantum field theory,'' in: R. Szabo and M. Bojowald (eds.), Encyclopedia of Mathematical Physics , Second Edition, Volume 5, 556--568 (2025) [arXiv:2305.03372 [math-ph]]
work page Pith review arXiv 2025
Show all 32 references
-
[9]
Benini, A
M. Benini, A. Schenkel and L. Woike, ``Operads for algebraic quantum field theory,'' Commun.\ Contemp.\ Math.\ 23, no.\ 2, 2050007 (2021) [arXiv:1709.08657 [math-ph]]
2021 arXiv
-
[10]
A. N. Bernal and M. Sanchez, ``Smoothness of time functions and the metric splitting of globally hyperbolic space-times,'' Commun.\ Math.\ Phys.\ 257, 43--50 (2005) [arXiv:gr-qc/0401112 [gr-qc]]
2005 arXiv
-
[11]
A. N. Bernal and M. Sanchez, ``Further results on the smoothability of Cauchy hypersurfaces and Cauchy time functions,'' Lett.\ Math.\ Phys.\ 77, 183--197 (2006) [arXiv:gr-qc/0512095 [gr-qc]]
2006 arXiv
-
[12]
Brunetti, K
R. Brunetti, K. Fredenhagen and R. Verch, ``The generally covariant locality principle: A new paradigm for local quantum field theory,'' Commun.\ Math.\ Phys.\ 237, 31--68 (2003) [arXiv:math-ph/0112041]
2003 arXiv
-
[13]
S. Bunk, J. MacManus and A. Schenkel, ``Lorentzian bordisms in algebraic quantum field theory,'' Lett.\ Math.\ Phys.\ 115, no.\ 1, 16 (2025) [arXiv:2308.01026 [math-ph]]
2025 arXiv
-
[14]
Costello and O
K. Costello and O. Gwilliam, Factorization algebras in quantum field theory: Volume 1 , New Mathematical Monographs 31, Cambridge University Press, Cambridge (2017)
2017
-
[15]
Costello and O
K. Costello and O. Gwilliam, Factorization algebras in quantum field theory: Volume 2 , New Mathematical Monographs 41, Cambridge University Press, Cambridge (2021)
2021
-
[16]
A. D. Elmendorf and M. A. Mandell, ``Permutative categories, multicategories and algebraic K -theory,'' Algebr.\ Geom.\ Topol.\ 9, no.\ 4, 2391--2441 (2009) [arXiv:0710.0082 [math.KT]]
2009 arXiv
-
[17]
C. J. Fewster and K. Rejzner, ``Algebraic quantum field theory -- an introduction,'' in: F. Finster, D. Giulini, J. Kleiner and J. Tolksdorf (eds.), Progress and visions in quantum theory in view of gravity , Birkh\"auser, Cham (2020) [arXiv:1904.04051 [hep-th]]
2020 arXiv
-
[18]
C. J. Fewster and R. Verch, ``Algebraic quantum field theory in curved spacetimes,'' in: R. Brunetti, C. Dappiaggi, K. Fredenhagen and J. Yngvason (eds.), Advances in algebraic quantum field theory , Springer Verlag, Heidelberg (2015) [arXiv:1504.00586 [math-ph]]
2015 arXiv
-
[19]
Gwilliam and K
O. Gwilliam and K. Rejzner, ``Relating nets and factorization algebras of observables: Free field theories,'' Commun.\ Math.\ Phys.\ 373, no.\ 1, 107--174 (2020) [arXiv:1711.06674 [math-ph]]
2020 arXiv
-
[20]
Gwilliam and K
O. Gwilliam and K. Rejzner, ``The observables of a perturbative algebraic quantum field theory form a factorization algebra,'' arXiv:2212.08175 [math-ph]
-
[21]
Haag and D
R. Haag and D. Kastler, ``An algebraic approach to quantum field theory,'' J.\ Math.\ Phys.\ 5, 848 (1964)
1964
-
[22]
Johnson-Freyd, ``Heisenberg-picture quantum field theory,'' in: A
T. Johnson-Freyd, ``Heisenberg-picture quantum field theory,'' in: A. Alekseev, E. Frenkel, M. Rosso, B. Webster and M. Yakimov (eds.), Representation theory, mathematical physics, and integrable systems , Progress in Mathematics 340, Birkh\"auser, Cham (2021) [arXiv:1508.0590...
2021 arXiv
-
[23]
Johnson and D
N. Johnson and D. Yau, 2 -dimensional categories , Oxford University Press, Oxford (2021) [arXiv:2002.06055 [math.CT]]
2021 arXiv
-
[24]
Martins-Ferreira, ``Pseudo-categories,'' J.\ Homotopy Relat.\ Struct.\ 1, no.\ 1, 47--78 (2006) [arXiv:math/0604549 [math.CT]]
N. Martins-Ferreira, ``Pseudo-categories,'' J.\ Homotopy Relat.\ Struct.\ 1, no.\ 1, 47--78 (2006) [arXiv:math/0604549 [math.CT]]
2006 arXiv
-
[25]
Minguzzi, ``Lorentzian causality theory,'' Living Rev.\ Relativ.\ 22, 3 (2019)
E. Minguzzi, ``Lorentzian causality theory,'' Living Rev.\ Relativ.\ 22, 3 (2019)
2019
-
[26]
O'Neill, Semi-Riemannian geometry , Academic Press, New York (1983)
B. O'Neill, Semi-Riemannian geometry , Academic Press, New York (1983)
1983
-
[27]
C. I. Scheimbauer, Factorization homology as a fully extended topological field theory , PhD thesis, ETH Z\"urich (2014). https://doi.org/10.3929/ethz-a-010399715
2014 doi
-
[28]
Schreiber, ``AQFT from n -functorial QFT,'' Commun.\ Math.\ Phys.\ 291, 357--401 (2009) [arXiv:0806.1079 [math.CT]]
U. Schreiber, ``AQFT from n -functorial QFT,'' Commun.\ Math.\ Phys.\ 291, 357--401 (2009) [arXiv:0806.1079 [math.CT]]
2009 arXiv
-
[29]
G. B. Segal, ``The definition of conformal field theory,'' in: U. Tillmann (ed.), Topology, Geometry and Quantum Field Theory , London Math.\ Soc.\ Lecture Note Ser., Vol.\ 308, Cambridge University Press, Cambridge (2004)
2004
-
[30]
M. A. Shulman, ``Constructing symmetric monoidal bicategories,'' arXiv:1004.0993 [math.CT]
-
[31]
Stolz and P
S. Stolz and P. Teichner, ``Supersymmetric field theories and generalized cohomology,'' in: H. Sati and U. Schreiber (eds.), Mathematical foundations of quantum field theory and perturbative string theory , Proc.\ Sympos.\ Pure Math., Vol.\ 83, Amer.\ Math.\ Soc., Providence, ...
2011 arXiv
-
[32]
Witten, ``Topological quantum field theory,'' Comm.\ Math.\ Phys.\ 117, 353--386 (1988)
E. Witten, ``Topological quantum field theory,'' Comm.\ Math.\ Phys.\ 117, 353--386 (1988)
1988
Reviewed August 16, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.