{"id":"c29dd4fe-95de-4b66-b612-6d22cd001b11","arxiv_id":"1908.04801","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A p-brane Newton-Cartan structure admits a torsion-free compatible affine connection exactly when it is Augustinian, and all such connections are classified by longitudinal-frame field strengths.","lead":"This paper defines p-brane Newton-Cartan geometry, a framework for describing gravity around non-relativistic p-branes. It proves when such a geometry admits a torsion-free compatible connection and describes the full space of those connections.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 3.7 derives (97) only as a necessary condition; it never verifies that (97) satisfies ∇τ=0 and ∇h=0, so the existence half of Theorem 3.1 is unproven.","rationale":"The paper is self-contained and clearly states its compatibility definition (57), so the reader's weakest assumption about alternative compatibility notions is not the central issue. The more load-bearing concern is the missing existence check in Proposition 3.7: the proof derives the connection components (97) as a consequence of assuming a compatible torsion-free connection exists, but it never verifies that the proposed formula actually satisfies ∇τ=0 and ∇h=0. Since τ and h are degenerate, the Koszul equations do not automatically guarantee compatibility, and the proof does not close the loop. If formula (97) fails for some genuine Augustinian structure, the sufficiency half of the necessary-and-sufficient characterization collapses, and Theorem 3.1 loses its origin. The proposed symbolic check would settle this directly. The result is plausible and the paper has independent support from the p=0 limit, but a rigorous acceptance should require the missing verification. Hence CONDITIONAL rather than outright ACCEPT.","tokens_in":23253,"tokens_out":26920,"duration_ms":250080,"concrete_test":"Perform a symbolic computation in a coordinate chart: insert (97) into ∇_μ τ_{νρ}=∂_μ τ_{νρ}−Γ^λ_{μν}τ_{λρ}−Γ^λ_{μρ}τ_{νλ} and ∇_μ h^{νρ}=∂_μ h^{νρ}+Γ^ν_{μλ}h^{λρ}+Γ^ρ_{μλ}h^{νλ}, using the Augustinian identities (74)–(75) and completeness (52). Verify both vanish identically and that Γ^λ_{μν}=Γ^λ_{νμ}. If any residual term survives, the theorem is false; if all cancel, the existence gap is closed and only a presentation fix remains.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that every Augustinian pNC structure admits a torsion-free compatible connection, and that for each longitudinal frame the connection τ∇ defined by (97) lies in D0(M,τ,h). In the proof of Proposition 3.7, the author assumes τ∇ exists and derives equations (100)–(101) from compatibility and torsion-freeness. Solving these yields the component expression (97). This establishes only a necessary condition: any such connection must have those components. It does not establish that the right-hand side of (97) actually satisfies the compatibility equations ∇_μ τ_{νρ}=0 and ∇_μ h^{νρ}=0, nor that it is torsion-free. Because τ and h are degenerate, the Koszul equations (100)–(101) do not uniquely determine a connection and the cancellation of cross-terms is non-trivial. The sufficiency of the Augustinian conditions, and the vector-space origin in Theorem 3.1, rest on this existence. The same gap affects the local-to-global patching: (97) is written using a longitudinal frame, which may exist only locally, and no argument shows the local expressions glue to a global connection.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper provides a formal, coordinate-free definition of p-brane Newton--Cartan (pNC) geometry and studies the existence and classification of torsion-free affine connections compatible with the pNC structure. The author constructs the relevant symmetry group Gp and the associated tensors (τ, h), defines Aristotelian and Augustinian pNC structures, and claims that the Augustinian conditions are necessary and sufficient for the existence of torsion-free compatible connections. The main theorem (Theorem 3.1) asserts that, given a longitudinal frame, the space D0 of such connections is a vector space whose origin is an explicitly constructed 'torsion-free special connection', and that D0 is isomorphic to a space of gravitational field strengths. The paper also discusses the relation of its results to earlier work on stringy Newton--Cartan geometry and shows that p = 0 and p = d − 1 recover known Leibnizian and Lorentzian results.","tokens_in":23462,"tokens_out":10430,"duration_ms":108061,"significance":"If the main theorem is correct, this is a valuable foundational contribution to non-relativistic geometry. The paper gives a precise classification of when torsion-free compatible connections exist, determines the space of such connections, and identifies a larger ambiguity in the gravitational field strengths than was previously recognized in the gauge-theoretic literature. The work is self-contained, derives its conditions from the definitions rather than assuming them, and explicitly shows how the general framework reduces to known Newton--Cartan and pseudo-Riemannian cases. The presentation is generally clear and the algebraic setup is sound in outline. However, as detailed below, the proof of the existence half of the main theorem is incomplete, and the global patching of the locally constructed connections is not addressed; these gaps must be fixed before the central claims are fully supported.","major_comments":[{"comment":"The proof of Proposition 3.7 does not establish the existence of the torsion-free special connection τ∇. The argument assumes τ∇ ∈ D0(M, τ, h) with vanishing field strength, then derives equations (100)–(101) and consequently the component expression (97). This is only a necessary condition: it shows that any connection satisfying the stated properties must have those components, but it never verifies that the right-hand side of (97) actually satisfies ∇μτνρ = 0, ∇μhνρ = 0, and vanishing torsion. The alternative uniqueness argument via the bijection τΘ of Lemma 3.1 presupposes Proposition 3.3, which asserts that D0(M, τ, h) is an affine space but does not prove that D0 is non-empty. Since the Augustinian conditions are claimed to be sufficient for existence, the proof must be completed either by directly verifying that (97) defines a torsion-free compatible connection, or by providing an independent existence argument.","section":"Section 3.2, Proposition 3.7 and Theorem 3.1"},{"comment":"The paper does not address local-to-global gluing of the connections defined via longitudinal frames. The formula (97) and the parametrization (105) require a longitudinal frame τA, but the paper itself notes (after Definition 2.5) that the bundle of longitudinal frames Lfp(M, τ) need not admit global smooth sections. Consequently, Theorem 3.1 as stated only applies when a global longitudinal frame exists, and the paper gives no argument that the locally defined special connections patch to a global connection on the whole manifold. The purported canonical bijection (117) uses an orbit space F(M, τ, h) constructed from global longitudinal frames and global LGp-valued transformations; without a patching or partition-of-unity argument, its validity for general Augustinian pNC structures is not established. The authors should either restrict the theorem to manifolds with a global longitudinal frame or supply the missing global argument.","section":"Section 3.2, equations (97) and (105), and Definition 3.5"},{"comment":"The affine-space structure of D0(M, τ, h) is stated without proof. The conditions defining V(M, τ, h) — S^a_[bc] = 0, S^d_{a(b}τ_{c)d} = 0, and h_{d(c}S^b_{a)d} = 0 — are plausible, but the proof that any two torsion-free compatible connections differ by exactly such an S, and that every such S can be added to a compatible connection to obtain another compatible connection, is omitted. Since Proposition 3.3 underlies the application of Lemma 3.1 and hence both the uniqueness and the claimed vector-space structure, this proof should be included or supplied from a cited reference.","section":"Section 3.2, Proposition 3.3"}],"minor_comments":[{"comment":"The notation τϕ^{-1}(∇ − ∇')Aab is confusing: the argument of τϕ^{-1} is a tensor, not a pair of connections. Consider writing τϕ^{-1}(∇ − ∇') explicitly as τϕ^{-1}(S) for S = ∇ − ∇', with the understanding that S is the difference tensor.","section":"Section 3.2, equation (94)"},{"comment":"The word 'strenghts' in the title of Proposition 3.6 is a typo and should be 'strengths'.","section":"Section 3.2, Proposition 3.6 title"},{"comment":"References [21] and [26] are both listed as 'Work in progress' by the same group. If these works are not publicly available, the claims about alternative compatibility conditions and covariant expansions should either be stated more explicitly or cited as private communications, so that the reader can assess the dependence of the main results on the chosen compatibility condition.","section":"References [21] and [26]"},{"comment":"The comparison with the results of [5] would be easier to follow if the equations from [5] were cited by number in the text, particularly where the author states that the conventional curvature constraints imply the Augustinian conditions and that the field strengths are fixed as τF Aμν = 2D[μ mν]A.","section":"Section 4, discussion of [5]"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a serious foundational contribution, but the proof of the existence half of the main theorem is incomplete: the explicit connection formula (97) is derived only as a necessary condition, and the affine-space argument presupposes what it is meant to prove. I believe this gap is fixable, either by a direct verification of (97) or by a local-existence argument followed by patching, and I do not see evidence of a fundamental error in the mathematical framework. The unpublished self-references [21,26] are used to motivate alternative compatibility conditions; the editor may wish to ask the authors to clarify their status, since the main theorem's domain depends on the chosen definition of compatibility. If the existence proof is completed and the patching issue is addressed, the paper would be acceptable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this paper gives the first formal treatment of the equivalence problem for p-brane Newton-Cartan geometry, and I think the main theorem is true. But the written proof does not quite deliver the existence half, and a referee needs to push on that.\n\nWhat's new: the paper defines pNC structures in the Bekaert-Morand style, derives necessary conditions for the existence of torsion-free compatible connections (the Augustinian conditions), and determines the moduli space of such connections as an affine space. It recovers the Leibnizian and Lorentzian cases at p=0 and p=d-1. It also catches an ambiguity in the gravitational field strength that [5] missed: the kernel of the parametrization includes mixed longitudinal-transverse components, not just longitudinal ones. That is a real correction.\n\nWhere it works: the construction of bundles, torsors, projectors, and inverse metrics is careful and self-contained. The necessary conditions in Prop 3.2 are clean. The algebra in Prop 3.8 is long but the logic is clear.\n\nThe soft spot: the sufficiency proof. Prop 3.3 asserts that D0 is an affine space modelled on V. That assertion already carries existence: an affine space has to have at least one point. Then Lemma 3.1 is used to define tau-nabla as the zero of an affine map, which again presupposes D0 is non-empty. So the proof that every Augustinian structure admits a torsion-free compatible connection is circular as written. Prop 3.7 only derives the coordinate formula (97) under the assumption that tau-nabla exists. It never checks that the right-hand side of (97) actually satisfies ∇τ=0 and ∇h=0 and is torsion-free. Because τ and h are degenerate, that is not a formality. The same gap hits the gluing between charts: (97) uses a longitudinal frame, which generally exists only locally, and no argument shows the local expressions patch.\n\nI want to be clear: this looks fixable. The Augustinian conditions seem exactly crafted to make (97) work, and I expect a direct verification will go through. But as submitted, the central theorem is not proven. This is a load-bearing gap, not a typo.\n\nWho should read this: anyone working on non-relativistic gravity, SNC geometry, or Newton-Cartan foundations. The conceptual contribution is solid.\n\nRecommendation: send to peer review, but insist that the referee ask for a proof that (97) defines a compatible torsion-free connection, plus a global gluing argument. If the author supplies that, the paper is very acceptable.","headline":"The main theorem is probably right, but the proof assumes what it needs to show.","tokens_in":23976,"tokens_out":6263,"would_cite":true,"duration_ms":62489,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that a p-brane Newton–Cartan structure admits torsion-free affine connections compatible with its metric data exactly when it is Augustinian, and that the space of such connections is a vector space whose origin is a…","keywords":["p-brane Newton-Cartan geometry","non-relativistic gravity","torsion-free compatible connections","equivalence problem","Augustinian structures","Aristotelian structures","gravitational field strengths","longitudinal frames"],"falsifier":"Take a pNC spacetime whose transverse directions are non-integrable, so that $\\mathrm{d}\\tau^A(V,W)\\neq 0$ for some transverse vector fields $V,W$; the theorem says no torsion-free connection with $\\nabla\\tau=0$ and $\\nabla h=0$ exists. Producing one such connection would falsify the central claim.","tokens_in":23047,"feed_emoji":"🌌","tokens_out":8575,"duration_ms":77632,"temperature":0.7,"pith_summary":"The paper sets up a formal foundation for p-brane Newton–Cartan (pNC) geometry, the non-relativistic geometry that describes gravity coupled to p-branes. Its central claim is that a pNC structure admits torsion-free affine connections compatible with both of its metric tensors if and only if the structure is Augustinian, meaning the exterior derivative of the longitudinal coframe vanishes on transverse vectors and a related mixed condition holds. For such structures, the paper determines the entire space of torsion-free compatible connections: it is a vector space, and each choice of longitudinal frame selects a unique torsion-free special connection that serves as the origin. This matters because it settles the equivalence problem for pNC geometry, giving precise control over the degrees of freedom in the connection and generalizing the classical results for Newton–Cartan and Lorentzian geometry.","feed_headline":"Torsion-free connections exist only for Augustinian p-brane spacetimes","feed_subtitle":"The theorem turns the space of compatible connections into a vector space and generalizes Newton-Cartan gravity.","key_machinery":"The load-bearing object is the longitudinal frame $\\tau_A$, a local section of the bundle of Lorentz-orthonormal frames inside the longitudinal part of the pNC structure, together with the projector it defines onto the transverse distribution. From these the paper builds the inverse metric pieces ${}^\\tau\\tau^{ab}$ and ${}^\\tau h^{ab}$, and then the gravitational field strength ${}^\\tau F_A$, a 2-form valued in a $(p+1)$-dimensional space, defined up to a gauge ambiguity ${}^\\tau K_A$. The central isomorphism sends a field strength $[F_A]$ to the tensor $\\tau^A_{(b}F_{|A|c)d}h^{ad}$, which parametrizes the difference between two compatible connections; the torsion-free special connection is the kernel of the affine map ${}^\\tau\\Theta(\\nabla)=[2{}^\\tau h_{c[b}\\nabla_{a]}\\tau^c_A]$. This map turns the a priori affine space of connections into a vector space with a canonical origin relative to the chosen frame.","core_discovery":"Choosing a longitudinal frame $\\tau_A$ on an Augustinian pNC structure $(M,\\tau,h)$, the paper proves (Theorem 3.1) that the space $D_0(M,\\tau,h)$ of torsion-free connections compatible with $\\tau$ and $h$ is a vector space whose origin is the torsion-free special connection ${}^\\tau\\nabla$, and that $D_0$ is naturally isomorphic to the space ${}^\\tau\\mathcal{F}$ of gravitational field strengths relative to $\\tau_A$. Every such connection has the explicit form $\\Gamma^\\lambda_{\\mu\\nu}={}^\\tau\\Gamma^\\lambda_{\\mu\\nu}+\\tau^A_{(\\mu}\\,{}^\\tau F_{|A|\\nu)\\rho}h^{\\lambda\\rho}$, where ${}^\\tau\\Gamma$ is given by a Christoffel-like formula built from the inverse longitudinal and transverse metrics. The Augustinian conditions, $\\mathrm{d}\\tau^A(V,W)=0$ and $\\eta_{A(B}\\,\\mathrm{d}\\tau^A(\\tau_C),V)=0$ for transverse $V,W$, are shown to be necessary and sufficient for any torsion-free compatible connection to exist.","pith_inferences":["If the dependence of the special connection on the longitudinal frame is pure gauge, then the canonical object is the quotient space $\\mathcal{F}(M,\\tau,h)$ of gravitational field strengths, not any individual frame-dependent connection.","The enlarged ambiguity found in the field strengths suggests that the conventional curvature constraints used to build stringy Newton–Cartan geometry could be relaxed without losing torsion-free compatible connections; counting the degrees of freedom after relaxing them would test this.","A natural next step, left open by the paper, is to characterize the subset of connections whose field strengths are closed or exact; that would give the pNC analogue of the classical closure condition selecting Galilean connections in Newton–Cartan gravity."],"forward_implications":["For any Augustinian pNC structure, a longitudinal frame fixes one distinguished torsion-free compatible connection, and all other compatible torsion-free connections are obtained by adding a gravitational field strength.","The space of torsion-free compatible connections is a vector space isomorphic to the space of gravitational field strengths, so the connection's propagating degrees of freedom are exactly captured by those field strengths.","The Augustinian conditions are necessary and sufficient for the existence of any torsion-free compatible connection, giving a concrete criterion that can be checked from the metric data alone.","Setting $p=0$ recovers the Newton–Cartan/Leibnizian results, and setting $p=d-1$ makes the torsion-free special connection the Levi–Civita connection; the theorem therefore interpolates between the two classical cases.","In the stringy case $p=1$, the conventional curvature constraints previously used imply the Augustinian conditions, and the new analysis shows the field-strength ambiguity is larger than previously noticed: the mixed components obey the antisymmetry condition ${}^\\tau K_A(\\tau_B,V)=-{}^\\tau K_B(\\tau_A,V)$."],"supporting_citations":[{"why":"Establishes the geometric reformulation of Newtonian gravity that this paper extends to p-branes.","marker":"[1]"},{"why":"Supplies the Leibnizian-structure framework, the Aristotelian/Augustinian classification, and the parametrization to which the p=0 results reduce.","marker":"[2]"},{"why":"Introduces stringy Newton–Cartan geometry and the conventional curvature constraints that imply the Augustinian conditions; the explicit connection form (105) first appeared there.","marker":"[5]"},{"why":"Gives the definition of Leibnizian structures, recovered here when p=0.","marker":"[22]"},{"why":"Defines the torsion-free special connections of Newton–Cartan gravity that are generalized to arbitrary p.","marker":"[23, 24]"}],"fun_headline_variants":["Augustinian p-branes are the only ones with torsion-free connections","Torsion-free connections form a vector space for Augustinian p-branes","Only Augustinian p-brane spacetimes have torsion-free connections","pNC geometry: torsion-free connections require Augustinian frames","Generalizing Newton-Cartan: p-brane geometry with torsion-free connections"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole characterization rests on the paper's definition of compatibility as $\\nabla\\tau=0$ and $\\nabla h=0$; if a physical problem calls for a different compatibility notion, the Augustinian conditions and Theorem 3.1 need not describe the relevant connection space.","fun_headline_variants_meta":{"raw":{"variants":["Augustinian p-branes are the only ones with torsion-free connections","Torsion-free connections form a vector space for Augustinian p-branes","Only Augustinian p-brane spacetimes have torsion-free connections","pNC geometry: torsion-free connections require Augustinian frames","Generalizing Newton-Cartan: p-brane geometry with torsion-free connections"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001475,"raw_usage":{"total_tokens":5913,"prompt_tokens":916,"completion_tokens":4997,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":532,"completion_tokens_details":{"reasoning_tokens":4905}},"tokens_in":532,"tokens_out":4997,"duration_ms":40575,"temperature":1.0,"reasoning_tokens":4905,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:33:10.958745+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a pNC spacetime whose transverse directions are non-integrable, so that $\\mathrm{d}\\tau^A(V,W)\\neq 0$ for some transverse vector fields $V,W$; the theorem says no torsion-free connection with $\\nabla\\tau=0$ and $\\nabla h=0$ exists. Producing one such connection would falsify the central claim.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the geometric reformulation of Newtonian gravity that this paper extends to p-branes."},{"cited_title":"Leibnizian, Galilean and Newtonian structures of spacetime","cited_arxiv_id":"gr-qc/0211030","evidence_quote":"Gives the definition of Leibnizian structures, recovered here when p=0."}],"review_version":1}