Recognition: 5 theorem links
· Lean TheoremThermodynamics of analogue black holes in a non-Hermitian tight-binding model
Pith reviewed 2026-05-06 14:20 UTC · model claude-opus-4-7
The pith
A non-Hermitian lattice with gain, loss, and chiral hopping reproduces a Schwarzschild horizon and emits at a frequency fixed by the gain/loss alone.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
A one-dimensional lattice with balanced gain and loss on its two sublattices, plus directional next-nearest-neighbor hopping that differs on either side of a smooth interface, can be made to reproduce the Painlevé-Gullstrand form of the Schwarzschild metric near that interface. The interface then plays the role of an event horizon, and a semiclassical tunneling calculation across it yields an emission rate Γ ∝ exp(−8πM|γ|) of Parikh-Wilczek form. Matching the Bekenstein-Hawking entropy change to the thermodynamic entropy fixes the leading emitted frequency to ω₊ = 8|γ|/3, set entirely by the on-site gain/loss strength.
What carries the argument
Mapping the tilted cone of exceptional points of a PT-symmetric non-Hermitian Bloch Hamiltonian onto a tetrad, so that the non-reciprocal NNN hopping κ plays the role of the Painlevé-Gullstrand infall velocity √(2M/r), and a smooth κ(r) profile across an interface becomes a Schwarzschild horizon. The Parikh-Wilczek tunneling integral, evaluated on the exceptional cone, then converts gain/loss γ into a Hawking emission channel.
If this is right
- <parameter name="0">The on-site gain/loss parameter becomes a single experimental knob that tunes the analogue Hawking frequency
- so a sweep of γ should trace out the predicted ω = 8|γ|/3 line.
Load-bearing premise
The mapping works only after dropping a commutator term that is non-zero precisely because the hopping changes across the horizon, and after matching the metrics only to linear order near the interface; if that dropped term is comparable to γ where the tunneling integral lives, the identification of γ as the Hawking channel falls apart.
What would settle it
A direct numerical simulation of the two coupled non-Hermitian chains — computing the local density of states, transmission/reflection across the interface, or two-point correlations between interior and exterior — should show emission peaked at ω = 8|γ|/3 with weight exp(−8πM|γ|) as γ and the kink width l are varied. If the spectral response instead depends strongly on (κ₁−κ₂)/2l (the term neglected in Eq. 9) and shows no clean ω ∝ |γ| line, the central identification fails.
read the original abstract
We present a non-Hermitian model with gain/loss and non-reciprocal next-nearest-neighbor hopping that emulates black-hole physics. The model describes a one-dimensional lattice with a smooth connection between regions with distinct hopping parameters. By mapping the system to an effective Schwarzschild metric in the Painlev\'e-Gullstrand coordinates, we find that the interface is analogue to a black-hole event horizon. We obtain emission rates for particles and antiparticles, the Hawking temperature, the Bekenstein-Hawking entropy, and the mass of the analogue black hole as a function of the interface sharpness and the system parameters. An experimental realization of the theoretical model is proposed, thus opening the way to the detection of elusive black-hole features.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript presents a one-dimensional PT-symmetric non-Hermitian tight-binding chain with on-site gain/loss γ and non-reciprocal next-nearest-neighbour hopping κ. By expanding the Bloch Hamiltonian around k=π, the authors identify the resulting Dirac-like form with a 1+1D effective metric whose tetrad reproduces, after spatially varying κ→f(r) across a smooth interface, the Painlevé-Gullstrand form of the Schwarzschild line element to linear order in (r−r₁). On the exceptional cone they apply the Parikh-Wilczek tunnelling prescription, contour-integrate around the pole at r=r₁, and obtain a Hawking-like emission rate Γ ∝ exp(−8πM_o|γ|), interpret the exponent as a Bekenstein-Hawking entropy change, and (using dM = T_H dS) extract a leading-order emission frequency ω_+ = 8|γ|/3 controlled solely by the gain/loss parameter. An experimental implementation in photonic, topoelectric, or robotic-metamaterial platforms is proposed.
Significance. If the central derivation holds, the paper provides a notably compact bridge between non-Hermitian band physics and black-hole thermodynamics: a closed-form lattice Hawking temperature, an entropy-difference interpretation of the WKB exponent, and a parameter-light prediction ω_+ = 8|γ|/3 that is in principle directly testable in tabletop platforms with gain/loss and non-reciprocity. The specific identification of γ — an experimentally tunable on-site quantity — as the sole control of the leading emission frequency is the kind of falsifiable, platform-agnostic statement that gives the manuscript its appeal. The use of Painlevé-Gullstrand coordinates avoids horizon coordinate singularities in the analogue setting, and the supplemental material lays out the contour deformation explicitly. The work would be a useful contribution to the analogue-gravity / non-Hermitian topology literature, conditional on tightening the regime-of-validity argument for the central tunnelling integral.
major comments (4)
- [Eqs. (9)-(10), (S15)-(S21)] The derivation drops [k̃, h(k̃)] = k̃ ∂_r f(r) (Eq. 9) under the smallness condition |(κ₂−κ₁)/(2l) k̃| ≪ |γ| (Eq. 10). However, Eq. (13) fixes (κ₁−κ₂)/(2l) = 1/(2r₁), and on the exceptional cone (Eq. S15) the radial momentum diverges near the horizon: after the Laurent expansion in §S-II one has k̃_r ~ 2r₁γ/(r−r₁). Substituting this into Eq. (10) gives a left-hand side ~|γ|/|r−r₁|, which is NOT small compared to |γ| — it is parametrically larger — precisely in the integration window (r₁, r₁−2ω) where the residue (Eqs. S19-S21) supplies the entire imaginary part of the action. The dropped term has the same simple-pole structure as the kept term, so its inclusion can renormalise the residue by an O(1) factor and thereby modify the coefficient in Γ ∝ exp(−8πM|γ|), the surface-gravity identification, and ω_+ = 8|γ|/3. The authors should either (i) carry the commutator term through the contou
- [Eq. (10) and matching window] Eq. (10) is presented as the regime in which the model 'mimics the interior and exterior of a BH', but the regime in which Eq. (10) holds and the regime in which the integral (S16) picks up its residue appear to be disjoint. Inserting (κ₁−κ₂)/(2l)=1/(2r₁) into Eq. (10) yields |k̃|≪2r₁|γ|, while the integrand becomes singular precisely where |k̃| diverges. A self-consistency analysis is needed: state explicitly the joint conditions on γ, r₁, l, and ω under which (a) the linearisation around k=π is valid, (b) Eq. (10) holds throughout the integration path, and (c) the contour deformation in Fig. S1 captures the leading contribution. As written, the reader cannot verify that any non-empty parameter window satisfies all three.
- [Eq. (4) and tetrad construction] The promotion γ → ê_α^μ k̃_μ σ^α (Eq. 4) is the geometric heart of the mapping but is presented in a single line. γ in the microscopic Hamiltonian (Eq. 1) is a real on-site gain/loss; it is unclear in what sense it is being identified with an eigenvalue of a Dirac operator, and how the resulting metric (Eq. 5) inherits no dependence on γ even though γ is the parameter that ultimately controls the Hawking spectrum (Eq. 19). A more explicit derivation — including the precise relationship between the imaginary structure of h(k̃) and the Lorentzian signature of g_μν — would substantially strengthen the paper, since all subsequent thermodynamic identifications rest on this step.
- [Eq. (12)-(13), and Eq. (S29)-(S30)] The metric matching is performed only to linear order in (r−r₁) (Eq. 12). This is sufficient to fix the surface gravity, but Eq. (S28)-(S30) integrates the thermal relation 8πM dM over a finite range using a global expression M(r) = r f²(r)/2 derived from the full metric. It should be stated explicitly that the thermal-entropy calculation only uses the near-horizon behaviour, and that higher-order discrepancies between f²(r) and 1−2M_o/r away from r₁ do not contaminate the ω-dependent terms in Eq. (S30) that ultimately determine ω_+ = 8|γ|/3.
minor comments (6)
- [Eq. (1) and Fig. 1] Sign conventions on the non-reciprocal NNN hopping are easy to misread: please indicate explicitly which direction carries +κ and which −κ on the figure, and confirm consistency with the Bloch Hamiltonian used to obtain Eq. (2).
- [Eq. (16)] The step Γ ∝ (A_+ + A_−)² with the cross term vanishing because A_+ and A_− 'do not belong to the same domain' (Eq. S26 discussion) deserves one or two more sentences. Naively |A_+ + A_−|² always contains an interference term; the argument that γ>0 vs γ<0 are disjoint sectors should be made operational (i.e., what physical observable corresponds to incoherent addition?).
- [Eq. (19)] It would help to state ω_− as well, comment on its sign and physical interpretation (or lack thereof), and note that (κ₁−κ₂)/(2l) ≪ 1, given Eq. (13), is equivalent to r₁ ≫ 1 in lattice units (large-BH limit). The reader should not have to re-derive this connection.
- [Discussion] The statement that the 'particle/antiparticle' channels in a tabletop realisation correspond to correlated mode-conversion excitations rather than literal antimatter is welcome. Consider moving this clarification earlier (around Eq. 14), where the labels are introduced, since it changes how the reader interprets the entire tunnelling section.
- [References] Refs. [8] and [10] appear to refer to the same Living Reviews article. Please consolidate. Also, ref. [26] is presented as a footnote-style remark inside the bibliography; convert to a proper footnote.
- [General] Several typos: 'sufficeintly' (after Eq. 14); 'thermodynamicalfeatures' missing space (intro); 'Panlevé' should be 'Painlevé' in ref. [18]; equation labels in supplemental are referenced as 'Eq. (S33)' but the body text cites them as 'Eq. 19, S33'. Please proofread.
Simulated Author's Rebuttal
We thank the referee for a careful and constructive reading. The recommendation of major revision is conditioned on tightening the regime-of-validity of the central tunnelling integral, and we believe each of the four major points can be addressed in the revised manuscript. The most substantive concern — whether the dropped commutator term [k̃, h(k̃)] can renormalise the residue that fixes the surface gravity and ω₊ = 8|γ|/3 — is a legitimate technical question that we will treat explicitly in a new appendix in the Supplemental Material. The remaining points (joint regime of validity of Eqs. (10) and (S16); a more pedagogical derivation of the tetrad promotion γ → ê^μ_α k̃_μ σ^α; and clarification that only the near-horizon behaviour of f(r) feeds the entropy-balance argument) are clarifications that strengthen the manuscript without altering its conclusions. Below we respond point by point and indicate the corresponding revisions.
read point-by-point responses
-
Referee: [Eqs. (9)-(10), (S15)-(S21)] The dropped commutator [k̃,h(k̃)] = k̃ ∂_r f(r), under (κ₁−κ₂)/(2l) = 1/(2r₁) and k̃_r ~ 2r₁γ/(r−r₁), gives a contribution ~|γ|/|r−r₁|, not small compared to |γ|, in the very integration window where the residue is picked up. Its inclusion can renormalise the exponent in Γ ∝ exp(−8πM|γ|).
Authors: The referee is correct that Eq. (10), which is sufficient for the linearised eigenvalue equation in the bulk of each chain, becomes marginal on the exceptional cone in the immediate neighbourhood of r₁, since k̃_r develops a simple pole. We will add a dedicated subsection in the Supplemental Material in which the commutator term is retained from the outset and the eigenvalue problem is solved with the full operator. Two observations are central. (i) The commutator k̃ ∂_r f(r) shifts h(k̃) by a term proportional to σ⁰ that is real and has the same simple-pole structure as the kept term; it therefore modifies the *real* part of k̃_r near r₁ but contributes to the imaginary part only through a change in the residue of the same pole at r̃=0. (ii) Carrying out the contour deformation of Fig. S1 with the corrected k̃_r, the residue is shifted by a factor that vanishes in the limit (κ₁−κ₂)/(2l) → 1/(2r₁) imposed by the matching condition (13), because ∂_r f(r₁) is precisely cancelled by the regular part of γf/(1−f²) at the horizon. We will present this calculation in detail; the leading exponent 8πM_o|γ| and hence ω₊ = 8|γ|/3 are preserved, while subleading O(ω/M_o) corrections are modified. We will state this explicitly and quantify them. revision: yes
-
Referee: [Eq. (10) and matching window] The regime in which Eq. (10) holds and the regime in which (S16) picks up its residue appear disjoint. The reader cannot verify that any non-empty parameter window jointly satisfies (a) linearisation around k=π, (b) Eq. (10) on the integration path, and (c) the contour deformation of Fig. S1.
Authors: We agree this should be made explicit. In the revision we will add a paragraph stating the joint window: (a) linearisation around k=π requires |k̃| ≪ 1 in lattice units, equivalently ω, |γ| ≪ τ; (b) the smallness condition (10) is to be understood as a condition on the *bulk* eigenvalue problem away from the horizon, controlling the spatial profile of f(r) on scales l ≫ 1/|γ|; (c) inside the residue window |r−r₁| ≲ 2ω, the relevant smallness parameter is not (10) but the ratio ω/r₁ = ω/(2M_o), which is the standard Parikh–Wilczek small parameter and is independent of l. The contour of Fig. S1 captures the leading contribution provided ω ≪ M_o and l|γ| ≫ 1 (so the domain wall is wide on the scale of the horizon). We will give a parameter table making the non-empty intersection explicit, e.g. l|γ| ~ 10–10², ω/|γ| ~ O(1), κ₁−κ₂ ~ O(1). revision: yes
-
Referee: [Eq. (4) and tetrad construction] The promotion γ → ê^μ_α k̃_μ σ^α is the geometric heart of the mapping but is given in one line. It is unclear in what sense the real on-site γ is identified with an eigenvalue of a Dirac operator, and why g_μν inherits no γ-dependence even though γ controls the Hawking spectrum.
Authors: We will expand this step substantially, following and extending the construction of Ref. [16]. The logic is: on the exceptional cone, the *defining equation* γ = −κk̃ ± τ|k̃| has the algebraic form of a generalised dispersion relation γ² = (k̃ τ)² + (k̃ κ)² − 2γ(κk̃) — i.e. γ plays the role of a mass-shell eigenvalue of the Dirac-like operator −κk̃ σ⁰ + τ|k̃|σ¹ obtained by squaring h(k̃). Promoting this scalar relation to an operator equation γ̂ = ê^μ_α k̃_μ σ^α is then a kinematic identification that reads off the tetrad components from the coefficients of σ⁰ and σ¹. The metric g_μν = ê^μ_α ê^ν_β η^{αβ} therefore depends only on (τ, κ), not on γ; γ enters subsequently as the on-shell value, i.e. as the energy scale at which the cone is probed, and it is precisely this on-shell condition that selects the exceptional cone — the analogue of the light cone — and propagates into the surface gravity. We will add an explicit derivation, including the sign/structure of the imaginary parts of h(k̃) and how they map onto the Lorentzian signature η = diag(−1,1). revision: yes
-
Referee: [Eq. (12)-(13), (S29)-(S30)] The metric matching is only to linear order in (r−r₁), but the thermal entropy in (S28) integrates the global expression M(r)=rf²(r)/2 over a finite range. It should be stated that only the near-horizon behaviour enters ω₊.
Authors: We agree, and we will state this explicitly in the revised text. The point is that the integral in Eq. (S28), after Taylor expansion to leading order in (r−r₁) carried out in Eq. (S30), retains only the linear coefficient of f(r) at r₁ — i.e. precisely the same data fixed by the matching condition (12)-(13). Higher-order discrepancies between f²(r) and 1−2M_o/r away from r₁ contribute only to terms of order (ω/r₁)³ and higher in ∆S, which do not affect the coefficient of ω in (S30) nor, consequently, the leading result ω₊ = 8|γ|/3. We will add a sentence after Eq. (S29) and a footnote in the main text emphasising that the thermodynamic identification uses the metric only in its near-horizon expansion, in line with standard analogue-gravity practice. revision: yes
- The detailed claim in our response to point 1 — that the residue is preserved exactly under the matching condition (13) once the commutator term is retained — is what we expect on structural grounds, but the explicit calculation has not yet been written up in the present version. Should the full computation reveal an O(1) renormalisation of the exponent (and hence of ω₊), the numerical coefficient 8/3 would need to be updated; the qualitative claim that the leading emission frequency is set solely by γ would, however, be unaffected.
Circularity Check
Spacetime-mapping ansatz imported via self-citation; "prediction" ω₊ = 8|γ|/3 is partly structurally forced once the residue ∝ r₁|γ| and r₁ = 2M_o matching are accepted, but the residue calculation itself supplies independent computational content.
specific steps
-
ansatz smuggled in via citation
[Eq. (4), 'Analogue spacetime' section]
"Noticing that γ takes the form of the eigenvalues of a Dirac operator, we promote it to an operator, γ̂ = -κ k̃ σ_0 + |τ k̃| σ_x → γ̂ = e_α^μ k̃_μ σ^α ... To relate the exceptional cone ... with the light cone of a Schwarzschild BH, we follow the procedure outlined in Ref. [16]."
The promotion of γ into a vielbein-dressed Dirac operator is the single step that converts a lattice dispersion into a spacetime metric (Eq. 5). It is justified only by 'we follow the procedure outlined in Ref. [16]', a paper co-authored by one of the present authors (Stålhammar). The choice of which Pauli structures get vielbein indices, and hence which metric emerges, is fixed by this self-cited procedure rather than by an external uniqueness argument. Without this ansatz, no Schwarzschild analogue follows.
-
self citation load bearing
[Refs. [16], [27]; used at Eq. (4) and Eqs. (14)-(16)]
"a relation between nH PT-symmetric models and BHs becomes evident [16]. ... Im(S_±) = Im[∫ dr k̃_r^±] = I_± [16, 19]. ... we obtain the semiclassical emission rate [16, 19] Γ ∝ (A_+ + A_-)^2 = e^{-8πM_o|γ|}"
The framework for treating the lattice exceptional cone as a tunneling problem and identifying particle/antiparticle channels via radial null geodesics is taken from Ref. [16] (overlapping author). Ref. [16] is invoked at every load-bearing structural step (mapping, channel identification, emission-rate formula), so the present paper's central claim that γ plays the role of the Hawking channel is inherited rather than independently established here.
-
fitted input called prediction
[Eq. (19) / Eq. (S33), 'leading order ω_+ = 8|γ|/3']
"Recalling Eq. (10), (κ_1 - κ_2)/2l ≪ 1, the leading order contribution to the frequency becomes ω_+ = 8|γ|/3. This remarkably simple relation indicates that the frequency of the emitted particles ... is controlled exclusively by the gain/loss parameter."
The residue (S19-S21) yields I_+ = 2πr_1|γ|, linear in |γ|, with r_1 fixed to 2M_o by the matching condition (Eq. 13). The thermal-entropy side ΔS = -3π(2Mω+ω²)/2 starts linearly in ω. Equating the two therefore forces ω ∝ |γ| at leading order by construction; only the numerical coefficient 8/3 is genuinely computed (and it depends on the chosen f(r) profile in Eq. 8, not on universal lattice physics). The proportionality 'ω is set entirely by γ' is therefore not a discovery but a consequence of how the variables enter the integrals.
-
renaming known result
[Eqs. (16)-(17), identification ΔS_BH = -8πM|γ|]
"Γ ∝ (A_+ + A_-)^2 = e^{-8πM_o|γ|} = e^{ΔS_{B-H}} ... ΔS_{B-H} = -8πM_o|γ| = -4π|γ| l/(κ_1-κ_2)."
Standard Parikh-Wilczek gives Γ ∝ exp(-8πMω(1-ω/2M)), with ω the emission frequency. Here |γ| occupies the slot of ω in the leading thermal exponent, and the result is then declared to equal ΔS_BH. This is a renaming step: the Schwarzschild entropy formula S_BH = 4πM² is borrowed wholesale and |γ| is substituted in for ω, before the actual emission frequency ω is solved for in the next step. The identification of the lattice quantity with the gravitational entropy is by stipulation, not derivation.
full rationale
The paper has genuine independent computational content (residue calculation in §S-II, thermal entropy expansion in §S-III), so it is not predominantly circular. However, two load-bearing pieces are weakened by importation: (a) The whole "vielbein promotion" ansatz (Eq. 4) — promoting γ to e_α^μ k̃_μ σ^α — is justified solely by self-citation to Stålhammar et al. [16]. This ansatz is what creates the Schwarzschild metric from the lattice; without it there is no analogue spacetime. Since one author (Stålhammar) overlaps, this is a self-citation and is load-bearing. (b) The advertised "prediction" ω₊ = 8|γ|/3 is structurally forced rather than independent: the residue gives I_+ = 2πr_1|γ| (linear in |γ|, with r_1 set by the matching condition r_1 = 2M_o), so the exponent of Γ is by construction 8πM|γ|. Equating to a thermal-entropy expansion that begins with 2Mω makes ω ∝ |γ| at leading order automatic; only the numerical coefficient 8/3 is genuinely computed (and that coefficient comes from the f⁴(r) factor specific to the chosen analogue f(r)). (c) The intermediate identification exp(-8πM|γ|) = exp(ΔS_BH) effectively occupies the slot where standard Parikh-Wilczek has exp(-8πMω), i.e., |γ| is being substituted for an emission frequency in a Schwarzschild formula that is then used to derive ω. This is more a renaming step than a derivation, though the subsequent matching against ΔS_thermal does add nontrivial algebra. The skeptic's flagged issue (dropped commutator k̃ ∂_r f(r) ~ γ/(r-r_1) being the same singular order as the kept term at the residue pole) is a correctness/regime concern, not a circularity per se, so it does not enter this score. Net: one load-bearing self-citation (the vielbein ansatz from [16]) and one "prediction" that is largely forced by construction. Score 4 — some self-citation; central calculation still has independent content.
Axiom & Free-Parameter Ledger
Lean theorems connected to this paper
-
IndisputableMonolith.Constants (octave := 8 * tick) and Foundation.DimensionForcing (eight_tick = 2^D with D=3)eight_tick_is_2_cubed unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
ω₊ = 8|γ|/3 ... the leading order contribution to the frequency becomes ω₊ = 8|γ|/3. This remarkably simple relation indicates that the frequency of the emitted particles ... is controlled exclusively by the gain/loss parameter.
-
Unification.YangMillsMassGap (mass gap as J(φ), strict spectral gap on φ-ladder)Jcost_phi_eq_massGap, spectral_gap unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
Γ ∝ exp(−8πM|γ|) = exp(ΔS_{B-H}), with ΔS_{B-H} = −8πM|γ|.
-
Unification.SpacetimeEmergence (Lorentzian signature (1,3), light cone, proper time forced from J-cost)spacetime_emergence_cert, lorentzian_signature, lightlike_iff_speed_c unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
The metric for this particular model is ds² = −(1−f²(r))dt² + 2f(r)dr dt + dr², matched order-by-order to Schwarzschild in Painlevé–Gullstrand coordinates by tuning κ₁,κ₂,l,r₁.
-
Foundation.DAlembert.Inevitability (forced functional form has no tunable smallness regime)bilinear_family_forced unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
We have neglected these contributions when solving the eigenvalue equation. ... |((κ₂−κ₁)/(2l)) k̃| ≪ |γ|, |(κ₂−κ₁)/(2l)| ≪ |τ|.
-
Cost.Jcost / Foundation.LawOfExistence (ratio-symmetric J(x) = ½(x+x⁻¹) − 1, unique calibrated cost)Jcost_symm, Jcost_eq_zero_iff unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
Bloch Hamiltonian h(k̃) with eigenvalues ε±(k̃) = ±√(τ²k̃² − (γ+κk̃)²); exceptional points form a tilted cone parameterized by κ.
What do these tags mean?
- matches
- The paper's claim is directly supported by a theorem in the formal canon.
- supports
- The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
- extends
- The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
- uses
- The paper appears to rely on the theorem as machinery.
- contradicts
- The paper's claim conflicts with a theorem or certificate in the canon.
- unclear
- Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.