Indistinguishability is not a brute property of particles.
HGUT proposes it is the structural identity of excitations of one medium —
a proposal whose test is whether the occupation algebra can be derived.
Objective
The Hong–Ou–Mandel (HOM) effect is one of the most precise and unambiguous demonstrations of quantum interference. Two identical photons are sent into a 50:50 beam splitter from opposite sides. Classically, one expects four equally likely outcomes: both photons exit through port A, both exit through port B, or one photon exits through each port (with two ways for the latter to occur). The classical particle prediction is therefore that coincidence detection occurs half the time. Quantum mechanically, coincidence detection is suppressed entirely: when the photons overlap in all relevant modes, both always exit together through the same port.
Standard quantum optics gives a complete predictive account of this at its own level: unitary mode mixing, bosonic commutation relations, interference of the both-transmitted and both-reflected amplitudes, and a mode-overlap dependence that matches experiment in detail. Nothing in that account is vague. What HGUT seeks is a layer beneath it — an underlying collective mechanism from which the bosonic mode algebra and its occupation structure emerge as the effective description of identical transverse excitations of the Graviton Mesh. This chapter reproduces the HOM effect exactly within that effective description, states precisely which ingredient HGUT proposes to ground, and marks the grounding itself at its earned level.
The chapter’s anchor. For one photon in each input of a balanced lossless beam splitter,
so the one-photon-in-each-arm output sector is absent: P(1,1) = 0, P(2,0) = P(0,2) = ½.
[ESTABLISHED — exact within the two-mode occupation algebra].
HGUT proposes that this algebra is the effective quantum description of identical transverse Mesh excitations; the derivation of the algebra from Harrison-Knot dynamics is the open microscopic gate.
[OPEN — emergence of the symmetric occupation algebra].
The derivation proceeds in five steps: the beam splitter as a medium transformation; the corrected two-alternative amplitude accounting; the two-mode occupation calculation; the quantitative overlap law governing imperfect indistinguishability; and the occupation-budget reading that connects the result to the locked measurement architecture. The chapter closes by stating what has and has not been established.
Beam Splitter as a Medium Transformation
A 50:50 beam splitter is a region of the medium that mixes two incoming modes into two outgoing modes with equal magnitude and a relative phase shift. Physically, it is a material element — itself a structured Harrison-Knot configuration embedded in the Mesh — whose geometry couples the two input modes to the two output modes through linear field equations. The coupling acts as a unitary transformation on the mode amplitudes:
The transformation is unitary and conserves total intensity, |ψ_A|² + |ψ_B|² = |ψ₁|² + |ψ₂|². The factor of i represents a π/2 phase shift on reflection. The phase is not arbitrary: unitarity of a symmetric splitter with equal-magnitude transmission and reflection requires it. In the amplitude language used below,
The transformation is linear in the mode amplitudes, and that linearity is what allows the two-quantum input to be analyzed alternative by alternative.
The Input: Two Occupation Quanta, Not Two Pellets
The input must be described with more care than an earlier draft of this chapter described it. Two single-photon excitations enter the opposite input modes with overlapping temporal, spectral, polarization, and spatial mode profiles. No fixed classical optical phase between the two sources is required. Independent single-photon states do not possess a definite classical phase relative to one another, and the HOM dip does not depend on one — this is precisely what distinguishes it from first-order (one-photon) interference, and it is why the dip survives with photons from fully independent sources. What HOM requires is two-photon indistinguishability: overlap of the two excitations in every mode degree of freedom that could in principle label which is which.
In occupation language the input is
two occupation quanta (Integer B of the Wave–Particle Duality chapter) carried by a shared transverse field configuration. In the HGUT reading, the photons are not compact pellets making independent port choices; they are two units of excitation whose field alternatives combine coherently at the splitter.
One honesty note before the calculation, because an earlier draft claimed more here than it had earned. The symmetric two-photon sector — the fact that the joint state of two identical photons lives in the exchange-symmetric part of the two-mode space, represented by commuting bosonic creation operators — is experimentally established and is the effective description this chapter computes in. HGUT proposes that this algebra emerges from the occupation structure of identical transverse Mesh excitations: identical configurations of one substrate carry no label that could antisymmetrize their exchange. That is an ontological proposal, not a completed derivation; a continuous field alone does not automatically supply quantized occupations, symmetric multi-quantum sectors, or photon-counting statistics. The reduction is marked open here and again in the closing ledger. [OPEN — emergence of the symmetric occupation algebra from Harrison-Knot dynamics].
The Two Alternatives That Interfere — Corrected Accounting
Which alternatives interfere to empty the coincidence sector? An earlier draft of this chapter reversed this accounting, and the correction is worth stating in full because the corrected version is more instructive than the error.
For a coincidence event — one quantum in each output — there are exactly two indistinguishable alternatives:
both photons transmitted and both photons reflected.
(If photon 1 is transmitted it reaches one output and photon 2, transmitted, reaches the other; if both are reflected they swap. Either way, one quantum per port.) The coincidence amplitude is therefore
The both-reflected alternative carries two factors of i — that is where i² = −1 enters — and it cancels the both-transmitted alternative exactly.
The bunched outcomes (both quanta in one port) are reached by alternatives containing one transmission and one reflection each. The unnormalized path sum for either bunched channel is
After normalization of the two-boson Fock state, (c†)²|0⟩ = √2 |2,0⟩, the amplitude of each bunched output is i/√2 and its probability is ½. The alternatives therefore survive because their amplitudes add, not because they are exempt from exchange symmetry. Both facts matter: the coincidence sector dies by destructive interference of TT against RR; the bunched sector lives by constructive addition of TR and RT.
The Two-Mode Occupation Derivation
The cleanest derivation uses the mode creation operators directly. Using operators does not commit HGUT to a particle ontology; it accurately represents the established occupation-number structure that any deeper theory must reproduce. Let a†, b† create one excitation in the two inputs and c†, d† in the two outputs. With the convention of the beam-splitter transformation above,
Transforming the input |1,1⟩ab = a†b†|0⟩:
= ½ (c† + i d†)(i c† + d†) |0⟩
= ½ [ i(c†)² + c†d† + i²d†c† + i(d†)² ] |0⟩
= ½ [ i(c†)² + (c†d† − d†c†) + i(d†)² ] |0⟩
= (i/2) [ (c†)² + (d†)² ] |0⟩
= (i/√2) ( |2,0⟩cd + |0,2⟩cd )
where the cross terms carry opposite signs from the two factors of i distributed across the two alternatives, and their difference vanishes because bosonic creation operators commute:
This is the coincidence-cancellation result in operator form: the commutativity of the creation operators is the algebraic expression of the symmetric bosonic occupation sector, and it is what deletes the |1,1⟩ output term. HGUT proposes that this symmetric sector ultimately reflects the structural identity of transverse excitations of one medium; deriving that connection remains the open microscopic gate.
[ESTABLISHED — exact within the two-mode algebra]. One downgrade from an earlier draft belongs here rather than in a footnote: only the absence of the coincidence sector is deterministic in the ideal state. Which output arm receives both quanta remains genuinely probabilistic, ½ each. The medium picture does not make the HOM experiment deterministic; it makes the cancellation structural.
The classical particle prediction was P_coinc = ½; the quantum result is 0; experiment, with photons overlapped in all modes, agrees with 0 to the limit of source and detector imperfections, with visibilities approaching unity and exceeding 99% in selected state-of-the-art implementations.
The Overlap Law: Imperfect Indistinguishability, Quantitatively
Real experiments operate at finite indistinguishability, and the correct statement of what degrades the dip is quantitative, not metaphorical. Let |f⟩ and |g⟩ denote the full mode states (temporal, spectral, polarization, spatial) of the two input photons, and let
be their total mode overlap. Then
[ESTABLISHED]. This equation assumes normalized pure single-photon inputs, a balanced lossless beam splitter, and ideal detection apart from the mode mismatch represented by γ. For mixed single-photon states, the relevant overlap is expressed through the density operators; for imperfect sources, multiphoton contamination, loss, and detector response introduce additional corrections.
The limits read off directly: |γ| = 1 gives P_coinc = 0 (perfect dip); |γ| = 0 gives P_coinc = ½ (classical value); intermediate overlap gives intermediate suppression. For normalized pure spectral amplitudes, when a relative delay τ is the only mismatch,
which is the measured HOM dip shape: coincidences at the classical level for large |τ|, falling to the minimum as the wave packets are brought into overlap.
One conceptual correction from an earlier draft: it is notthat “exchange symmetry scales continuously with similarity.” Exchange symmetry does not gradually turn off. What changes continuously is the mode overlap γ — the degree to which the two alternatives are indistinguishable in principle — and with it the weight of the interfering (indistinguishable) part of the amplitude relative to the non-interfering (distinguishable) part. Distinguishability in anydegree of freedom — polarization, frequency, arrival time, spatial mode, or a which-path marker elsewhere — enters through the same single number |γ|. This is the HOM analogue of the visibility–distinguishability boundary of the double-slit chapter, and it is what makes the dip depth a precision diagnostic of source indistinguishability in modern photonics.
The Occupation-Budget Reading
The result connects directly to the locked measurement architecture, and the connection is the modern content of the chapter.
The transverse field alternatives interfere;
the total occupation count remains two.
The beam splitter destroys neither occupation quantum. It redistributes the available two-quantum output sectors: the transformation |1,1⟩ → (i/√2)(|2,0⟩ + |0,2⟩) removes the one-and-one sector from the menu before any detector acts. Detection then proceeds by the architecture of the Wave–Particle Duality chapter, once per quantum: two occupation quanta are claimed, both in the same output arm. This does not require both quanta to be absorbed at the same microscopic site — a number-resolving detector may register two local events within the one output channel; what the interference forbids is a completed event in each arm. The division of labor is the familiar one, now at two-quantum order: the field layer (mode transformation, interference of alternatives) fixes the exposure sectors and their weights; the occupation layer (two indivisible quanta, claimed once each) fixes what a detector can register. HOM is what that division looks like when the field layer deletes a sector outright.
Why Distinguishable Inputs Show No Dip
If the two inputs are made distinguishable in any degree of freedom, the dip disappears and the coincidence rate returns to ½. In the corrected accounting this is immediate: the TT and RR alternatives can only interfere if nothing in the final joint state (including any marker system elsewhere) records which alternative occurred. A polarization tag, a frequency offset, an arrival-time separation, or a correlated marker degree of freedom makes the alternatives distinguishable in principle; the interfering fraction of the amplitude is reduced by |γ|²; and at |γ| = 0 the two alternatives contribute as independent probabilities, |t²|² + |r²|² = ¼ + ¼ = ½ — the classical value, recovered from within the quantum description. The same logic as the which-path result of the double-slit chapter applies: distinguishability, however implemented, is what removes interference; no completed click at the splitter is required.
What Classical Models Can and Cannot Do
The comparison with classical physics must be drawn carefully, because an earlier draft drew it too broadly.
Independent classical particles predict a coincidence probability of ½: each particle independently chooses a port, four equally likely outcomes, two of them coincidences. This prediction fails decisively against the observed dip.
Classical waves are a subtler comparison. Two phase-locked classical fields interfering at a beam splitter can produce output-intensity suppression, and certain classical-correlation arrangements produce HOM-like dips of limited visibility. What classical fields do not reproduce is the complete, phase-insensitive, two-single-photon counting behavior of the ideal HOM experiment: a dip that requires no phase relation between independent sources and appears in coincidence statistics of individual detection events.
Under the standard phase-averaged coincidence definition used to test HOM interference, classical optical inputs are subject to a 50% visibility boundary. Exceeding that boundary establishes the nonclassical character of the measured coincidence suppression, subject to the stated source and detector assumptions. The discriminating content of HOM is therefore the quantum occupation statistics, not field cancellation as such — and that is precisely why the open gate for HGUT is the occupation algebra, not the wave mechanics. HGUT must reproduce the counting statistics, and it does so in this chapter by computing in the effective occupation algebra whose microscopic origin it still owes.
Connection to HGUT Multi-Mode Dynamics
An earlier draft of this chapter cited the two-knot interaction simulations of Volume III — non-additive energies E(R) ≠ 2E_single, orientation-dependent interaction structure, phase-linked corridors — as “direct numerical evidence” for the non-factorizable configuration responsible for HOM interference. That claim is withdrawn. The distinction that modern HGUT maintains is exactly the collision-artifacts distinction of the matter volumes: Harrison Knots are localized topological matter configurations; photons are transported transverse excitations of the Mesh. Simulations of two interactingknots demonstrate that the medium supports coupled, non-additive, phase-linked configurations — a suggestive structural property of the substrate — but they do not derive photonic bosonic algebra, beam-splitter mode transformations, the |1,1⟩ → |2,0⟩ + |0,2⟩ map, or coincidence suppression. They are evidence about the medium, not a derivation of the HOM effect.
The honest statement of the program obligation is this: HGUT must derive the symmetric multi-occupation sector for transverse photon excitations from the underlying collective dynamics — plausibly from the superfluid base state that already supplies the single-quantum occupation ladder (Integer B), extended to the multi-quantum sector. Until that reduction is performed, the two-mode algebra used in this chapter is an effective description adopted from quantum optics, exactly reproduced but not yet grounded.
[OPEN — transverse multi-occupation reduction].
What This Chapter Has Established
- The beam splitter is a linear unitary mixing region of the medium with amplitudes t = 1/√2, r = i/√2; the π/2 reflection phase is forced by unitarity for a balanced symmetric splitter. [ESTABLISHED].
- For coincidence outputs, the interfering alternatives are both-transmitted and both-reflected; A_coinc = t² + r² = 0. The bunched alternatives (one transmission, one reflection each) survive because their amplitudes add. [ESTABLISHED — corrected accounting].
- In the two-mode occupation algebra, |1,1⟩ → (i/√2)(|2,0⟩ + |0,2⟩), so P(1,1) = 0 and P(2,0) = P(0,2) = ½; the cancellation follows from the symmetric bosonic occupation algebra and the commutativity of the creation operators. [ESTABLISHED — exact within the algebra].
- For normalized pure single-photon inputs at a balanced lossless splitter, imperfect indistinguishability obeys the overlap law P_coinc = ½(1 − |γ|²), with the delay-dependent dip shape derived above. [ESTABLISHED].
- The HOM dip requires two-photon mode overlap, not classical phase locking between sources. [ESTABLISHED].
- The occupation-budget reading: the field layer deletes the one-and-one sector; two quanta are then claimed, once each, in one arm. [CANDIDATE — HGUT interpretation, consistent with the locked measurement architecture].
What This Chapter Has Not Established
The bosonic algebra itself.
The symmetric two-photon sector and commuting creation operators are used as the established effective description. HGUT proposes that this algebra emerges from the occupation structure of identical transverse Mesh excitations — identical configurations of one substrate carrying no exchange label — but that microscopic reduction has not been performed. HGUT does not treat bosonic statistics as the final ontology; neither has it yet replaced the postulate with a derivation. [OPEN — the chapter’s decisive gate].
Higher-order multi-photon interference.
Three or more identical excitations into multi-port mixing regions (boson sampling and related linear-optical interference) require the same analysis at higher occupation order; not treated here. [OPEN].
Fermionic anti-bunching.
The anti-HOM dip for fermions requires the spinor structure of the Dirac volume; the bosonic case treated here uses only the envelope dynamics of this volume. [OPEN — deferred].
Detector-level dynamics at two-quantum order.
The single-quantum first-passage architecture is locked; its extension to number-resolving detection of two quanta in one channel is assumed consistent, not derived. [OPEN].
Status Ledger
| Claim | Status |
|---|---|
| Beam-splitter unitary with π/2 reflection phase | [ESTABLISHED] |
| Coincidence cancellation A_coinc = t² + r² = 0 (TT against RR) | [ESTABLISHED — corrected accounting] |
| |1,1⟩ → (i/√2)(|2,0⟩+|0,2⟩); P(1,1) = 0 | [ESTABLISHED — exact within the two-mode algebra] |
| Overlap law P_coinc = ½(1−|γ|²) and dip shape under the stated pure-input, balanced-lossless assumptions | [ESTABLISHED] |
| No classical-phase requirement between sources | [ESTABLISHED] |
| Occupation-budget reading (sector deletion at field layer; two quanta claimed once each) | [CANDIDATE — HGUT interpretation] |
| Emergence of the symmetric occupation algebra from Harrison-Knot / superfluid dynamics | [OPEN — decisive gate] |
| Two-knot simulations as evidence for photon HOM structure | [WITHDRAWN] — knot results concern matter configurations; suggestive of medium coupling only |
| Multi-photon interference; fermionic anti-bunching; two-quantum detector dynamics | [OPEN] |
Summary
The Hong–Ou–Mandel effect is often presented as evidence that nature requires particles with special statistics. Read within HGUT, it is evidence about what the statistics are: the interference of indistinguishable two-quantum alternatives of a shared field configuration.
Two occupation quanta enter opposite ports of a balanced splitter with overlapping mode profiles — no classical phase relation required. The coincidence sector collects two alternatives, both-transmitted and both-reflected; the two reflections contribute i² = −1, and the alternatives cancel exactly. The bunched sectors, reached by one transmission and one reflection each, survive because their amplitudes add. In the occupation algebra the same fact reads |1,1⟩ → (i/√2)(|2,0⟩ + |0,2⟩): the one-and-one output sector is deleted before any detector acts, and the two quanta are then claimed, once each, in whichever single arm the probabilistic outcome selects — ½ each, and only theabsence of coincidences is deterministic. Distinguishability in any degree of freedom re-populates the coincidence sector through the overlap law P_coinc = ½(1 − |γ|²), down to the classical ½ at zero overlap.
Two quanta in. Two quanta out, together.
Zero coincidences — in the ideal limit of full mode overlap.
Standard quantum optics and HGUT agree on every prediction above; they differ in what they take as primitive. Quantum optics takes the bosonic mode algebra as given and computes; HGUT proposes that the algebra is the effective description of identical transverse excitations of the Graviton Mesh — identity of configuration standing behind symmetry of exchange — and accepts the obligation that proposal creates:
HGUT proposes that the bosonic two-mode occupation algebra is the effective quantum description of identical transverse excitations of the Graviton Mesh. The HOM dip is reproduced exactly within that algebra; deriving the algebra from the underlying Harrison-Knot dynamics remains the open microscopic gate.
The next chapter turns to the Aharonov–Bohm effect, where a gauge-invariant phase holonomy records background electromagnetic structure even along paths where the local field strength vanishes. HGUT examines whether the electromagnetic connection admits a physical realization in the transverse displacement of the medium. That proposed realization must preserve gauge redundancy: gauge-equivalent descriptions must produce identical observables.