Wave–Particle Duality

In 1803 Thomas Young split a sunbeam in two and watched bright and dark bands paint themselves on a screen — interference, the unforgeable signature of a wave. Case closed: light is a wave. A century later the case reopened in the strangest way. Shine faint violet light on a potassium plate and electrons pop out instantly, each carrying an energy set by the light's color — while a blinding red floodlight, delivering millions of times more energy, ejects nothing at all. Waves cannot behave like that. Lumps can. And then the insult ran the other way: beams of electrons — the very archetype of a particle — were caught diffracting off nickel crystals like X-rays, complete with interference peaks. This is the crack through which all of quantum mechanics entered. You already own both classical languages — trajectories, momentum, and energy on one side; wavelength, frequency, and superposition on the other. Wave–particle duality is the discovery that nature refuses to file photons and electrons under either heading, and the whole dictionary connecting the two languages is a single constant: $E = h\nu$ and $\lambda = h/p$ , with Planck's $h$ appearing in both. Master this topic and you gain the conceptual key to everything that follows — the wavefunction, the Born rule, the uncertainty principle. Every one of them is this chapter's puzzle, formalized.

Key takeaways

Every quantum entity — photon, electron, neutron, even a 2000-atom molecule — shows wave-like interference and particle-like, localized detection; the experiment decides which face you see.
Light's particle face: energy arrives in photons of $E = h\nu$ , proven by the photoelectric threshold $\nu_0 = W/h$ and by Compton recoil with momentum $p = h/\lambda$ .
Matter's wave face: de Broglie's $\lambda = h/p$ , confirmed when Davisson and Germer diffracted 54 eV electrons off nickel in 1927.
The two faces are one object: a complex amplitude $\psi$ propagates and interferes like a wave, while $|\psi|^2$ gives the probability of a particle-like detection (Born rule).
Complementarity: any setup that reveals which path a quantum took erases the interference — amplitudes add only for indistinguishable alternatives.
Key idea: there is one quantum object, described by one amplitude — "wave" is how it travels, "particle" is how it arrives.

The core insight

A quantum entity is described by a single complex amplitude $\psi$ that propagates and superposes like a wave; the probability of any localized, particle-like detection is $|\psi|^2$ . "Wave" and "particle" are not two kinds of object but two complementary aspects of one amplitude, selected by the experiment.

Where this lives today

Duality stopped being a debate and became an industry. Every electron microscope is de Broglie's equation with a lens system: at $300\ \mathrm{kV}$ the electron wavelength is about $2\ \mathrm{pm}$ , thousands of times shorter than visible light, which is why cryo-electron microscopy can resolve individual protein atoms — the 2017 Nobel Prize in Chemistry rests on equation (4). Every CMOS camera sensor and solar cell runs on the internal cousin of Einstein's photoelectric logic: one photon in, one electron–hole pair created inside the silicon, threshold set by the band gap rather than the work function — no electron ever leaves the material, but the one-quantum-per-electron bookkeeping is Einstein's. At the frontier, the question "how big can a wave be?" is being answered by force. In 2019 the Arndt group in Vienna interfered molecules beyond $25{,}000$ atomic mass units — objects of two thousand atoms passing through a grating as a single amplitude. Atom interferometers now use matter waves as gravimeters precise enough to map underground water, and the current generation of experiments is pushing toward interfering nanoparticles, hunting for any mass scale at which superposition might quietly fail. So far, nature has refused to blink.

The physical picture

Fire electrons at a double slit one per minute, so slowly that no two are ever in flight together. Each electron announces itself on the far screen as a single sharp dot — as particle-like as a bullet hole. No smear, no half-hits. But let the dots accumulate overnight and something no bullet could do emerges: the dots avoid certain bands entirely and pile up in others, building exactly the fringe pattern Young saw with light. No individual electron is a fringe. The fringe lives in the statistics — in where each lone electron is likely to land — and a probability pattern with dark bands requires something that traveled through both slits and interfered with itself.

Figure

One dot at a time, a wave pattern emerges

Now try to catch nature in the act. Put a detector at the slits so you know which one each electron used. The moment the paths become distinguishable — even in principle, even if the tagging disturbs nothing — the fringes vanish and you get two boring single-slit blobs. Wave behavior and which-path knowledge are mutually exclusive. Bohr called this complementarity, and it is not a technological limitation; it is bookkeeping built into the amplitude algebra you will meet below. Hold on to the double slit as your compass. Everything in this article — Einstein's photon, Compton's recoiling X-rays, de Broglie's matter wave, the Born rule — is a commentary on that single experiment.

The particle face of light

Between 1887 and 1902, Hertz and Lenard established that ultraviolet light knocks electrons out of clean metals — and that the details make no classical sense: electron energy depends on frequency only, emission is instantaneous even for feeble light, and below a threshold frequency nothing comes out at any intensity. A spread-out wave, which should reward brightness and demand soak time, can do none of this. Einstein's 1905 resolution: light of frequency $\nu$ is a hail of discrete quanta — photons — each carrying

E = h\nu = \hbar\omega \qquad (1)

where: $h = 6.626\times10^{-34}\ \mathrm{J\,s}$ — Planck's constant; $\nu$ — frequency; $\omega = 2\pi\nu$ — angular frequency; $\hbar = h/2\pi = 1.055\times10^{-34}\ \mathrm{J\,s}$ . Read the structure: energy scales with frequency, and intensity appears nowhere — brightness only sets how many photons arrive, while color sets what each one is worth. That one design choice explains every photoelectric puzzle at a stroke — and the minuscule size of $h$ (a desk lamp pours out some $10^{20}$ photons a second) is why the graininess hid for a century. The full machinery — the escape-cost bookkeeping $K_{\max} = h\nu - W$ , the threshold $\nu_0 = W/h$ , the stopping-potential meter, and Millikan's universal $h/e$ line — is built and calibrated in the Photoelectric Effect and the Photon article. Duality borrows only the verdict: light's energy arrives in indivisible, frequency-priced packets. But a particle carries momentum as well as energy, and the photon's momentum was caught red-handed in 1923: Compton fired X-rays at graphite and found the scattered radiation shifted to longer wavelength, by an amount depending only on the scattering angle — each photon recoiling off an electron like a billiard ball. A wave would re-radiate at the frequency it arrived with; a particle must pay momentum and energy into the collision. The photon's momentum is

p = \frac{h}{\lambda} = \hbar k \qquad (2)

where: $p$ — momentum of one photon; $\lambda$ — its wavelength; $k = 2\pi/\lambda$ — the wavenumber. Notice the pairing with (1): energy goes with frequency, momentum goes with wavelength, and the same $h$ prices both. For light the two statements are equivalent (since $E = pc$ and $\nu\lambda = c$ ); for matter, as you are about to see, the momentum form is the one that survives.

Derivation AThe Compton shift from a photon-electron collision

We derive the signature formula of photon momentum: the wavelength shift of an X-ray photon scattering through angle $\theta$ off a free electron at rest. The strategy is pure collision mechanics — conserve relativistic energy and momentum, then eliminate the electron's recoil variables, which nobody measures. Relativity is not optional — though not because this recoil is fast (at these X-ray energies the electron leaves at only $\sim 0.05c$ ): the photon is massless, $E = pc$ is the only honest mechanics for it, and the formula must survive up to gamma-ray energies where the recoil genuinely is relativistic. Watch every trace of the X-ray's intensity and duration vanish from the answer — only geometry survives.

Step 1. Write the conservation laws. Incoming photon: energy $hc/\lambda$ , momentum $h/\lambda$ . Outgoing photon: $hc/\lambda'$ , momentum $h/\lambda'$ at angle $\theta$ . The electron (rest energy $m_ec^2$ ) recoils with momentum $p_e$ and total energy $E_e$ . Energy: $hc/\lambda + m_ec^2 = hc/\lambda' + E_e$ . Why: X-ray photon energies (tens of keV) dwarf atomic binding (eV), so treating the electron as free and at rest is a rounding error.

Step 2. Eliminate the recoil direction. Momentum conservation as vectors gives $\vec p_e = \vec p - \vec p\,'$ , so $p_e^2 = \left(\tfrac{h}{\lambda}\right)^2 + \left(\tfrac{h}{\lambda'}\right)^2 - \tfrac{2h^2}{\lambda\lambda'}\cos\theta$ . Why: the law of cosines on the momentum triangle removes the electron's unmeasured direction in one stroke.

Step 3. Bring in the relativistic energy-momentum relation. The electron obeys $E_e^2 = (p_ec)^2 + (m_ec^2)^2$ . Substitute $E_e$ from Step 1 and $p_e^2$ from Step 2. Why: this is the honest energy-momentum link at any speed — $\tfrac12 mv^2$ would wreck the bookkeeping exactly when the recoil is fastest.

Step 4. Grind and cancel. Expanding the squares, the $(hc/\lambda)^2$ and $(hc/\lambda')^2$ terms cancel against their partners in $p_e^2c^2$ , and what remains rearranges to $m_ec^2\!\left(\tfrac{hc}{\lambda} - \tfrac{hc}{\lambda'}\right) = \tfrac{h^2c^2}{\lambda\lambda'}(1 - \cos\theta)$ . Dividing through by $hc \cdot m_ec^2/\lambda\lambda'$ :

\Delta\lambda = \lambda' - \lambda = \frac{h}{m_ec}\,(1 - \cos\theta) \qquad (3)

Why: nothing about the light survives except its scattering angle — no intensity, no exposure time, no material constant beyond the electron mass.

Step 5. Read the scale. The prefactor $h/m_ec = 2.426\ \mathrm{pm}$ is the electron's Compton wavelength: the maximum shift (at $\theta = 180°$ ) is twice that, about 5 pm. Why: a picometre shift is invisible against visible light's 500 nm but percent-level for 70 pm X-rays — the effect must be comparable to what you can resolve. *The elimination reproduces $\Delta\lambda = (h/m_ec)(1-\cos\theta)$ symbolically from the full conservation system, and the Compton wavelength evaluates to $h/m_ec = 2.426\times10^{-12}\ \mathrm{m}$ from CODATA constants.*

Figure

Compton scattering — a photon plays billiards with an electron

The measured shifts matched (3) at every angle. Between the photoelectric threshold (energy in lumps) and Compton recoil (momentum in lumps), the particle face of light was closed to appeal.

The wave face of matter

In 1924 Louis de Broglie asked the symmetric question: if waves of light carry momentum like particles, do particles of matter carry a wavelength like waves? His proposal, made in a doctoral thesis before any evidence existed:

\lambda_{\mathrm{dB}} = \frac{h}{p} = \frac{h}{mv}, \qquad p = \hbar k \qquad (4)

where: $\lambda_{\mathrm{dB}}$ — the de Broglie wavelength of a particle of momentum $p$ ; $m$ , $v$ — mass and speed (non-relativistic); $k = 2\pi/\lambda$ — wavenumber. Read the reciprocal: the heavier or faster the particle, the shorter its wavelength — which is why matter's wave face hides at everyday scales and surfaces only when $p$ is small enough to stretch $\lambda$ to atomic dimensions. Note also which constant appears: the same $h$ as in equation (1). One number runs the dictionary in both directions — light acquiring momentum, matter acquiring wavelength.

Go deeper · Derivation B · 1 minDe Broglie's inversion — from photon momentum to matter wavelength

Derivation BDe Broglie's inversion — from photon momentum to matter wavelength

This is the shortest consequential derivation in physics: three lines that end with a hypothesis rather than a theorem. We retrace how the photon's energy-momentum relations suggest a wavelength for matter — honest about where deduction stops and audacity begins. Watch the direction of the logic: for light, $p = h/\lambda$ was derived; for matter, $\lambda = h/p$ was postulated, and only Davisson and Germer's nickel crystal made it physics.

Step 1. Start from the photon. A massless photon obeys $E = pc$ , from the relativistic relation $E^2 = (pc)^2 + (mc^2)^2$ with $m = 0$ . Why: relativity links the photon's energy and momentum with no free parameters.

Step 2. Insert the Planck relation. $pc = h\nu = hc/\lambda$ , so $p = h/\lambda$ . Why: combining Einstein's $E = h\nu$ with $\nu = c/\lambda$ turns the energy quantum into a momentum quantum — the same statement Compton's collisions verify directly.

Step 3. Invert the reading. De Broglie's bold step: assert that $\lambda = h/p$ holds for every particle, massive or not. Why: if nature pairs wave and particle quantities for light, symmetry suggests it does so for matter — a hypothesis, to be settled by experiment, and it was: electron diffraction appeared within three years. * $p = h/\lambda$ follows directly from combining $E = pc$ and $E = hc/\lambda$ ; the inversion to matter wavelengths is confirmed experimentally in de Broglie Matter Waves.*

Turning (4) into laboratory numbers — the electron-gun formula, the Davisson–Germer nickel peak that confirmed it to a percent, G. P. Thomson's foil rings, thermal neutrons — is the business of the next article, de Broglie Matter Waves, which owns that craft. The headline suffices here: electrons of a few tens of eV have wavelengths near $0.1\ \mathrm{nm}$ , the spacing of atoms in crystals, and when Davisson and Germer scattered them off nickel in 1927 the diffraction peaks sat exactly where $\lambda = h/p$ demanded. Matter diffracts. Since then neutrons, atoms, and molecules beyond 25,000 atomic mass units have shown the same fringes.

One amplitude, two faces

What kind of object interferes like a wave but arrives like a particle? Quantum mechanics answers with the wavefunction $\psi(\mathbf{r},t)$ : a complex amplitude that propagates and superposes like any wave, while

P(\mathbf{r},t) = |\psi(\mathbf{r},t)|^2 \qquad (5)

where: $P$ — probability density for a particle-like detection at $\mathbf{r}$ ; $\psi$ — the complex probability amplitude (Born rule). Notice the division of labor: $\psi$ propagates, superposes, and carries phase — but is never directly observed; $|\psi|^2$ is what the detector statistics actually measure. Squaring throws the phase away, unless two amplitudes were added before the square was taken — and that single reservation is where all of interference lives.

Go deeper · Derivation C · 1 minWhy adding amplitudes — not probabilities — is the whole mystery

Derivation CWhy adding amplitudes — not probabilities — is the whole mystery

Duality's entire content compresses into one algebraic contrast: amplitudes first, square second — versus square first, add second. We work out both bookkeeping rules for a two-route experiment and isolate the term that distinguishes them. Nothing beyond complex arithmetic enters, which is the point: every "how can one electron use both slits?" paradox becomes a statement about when the cross term survives. Watch what tagging a path does to the algebra — nothing mystical, just a sum moved outside a modulus.

Step 1. Two indistinguishable routes. If a detection can occur via route 1 or route 2 and nothing anywhere records which, the amplitudes superpose: $\psi = \psi_1 + \psi_2$ , and $P = |\psi_1 + \psi_2|^2$ . Why: superposition is the wave-like layer of the theory.

Step 2. Expand the square. Writing each amplitude with its phase, $|\psi_1 + \psi_2|^2 = |\psi_1|^2 + |\psi_2|^2 + 2\,\mathrm{Re}(\psi_1^*\psi_2)$ :

P_{\text{indist}} = |\psi_1|^2 + |\psi_2|^2 + 2|\psi_1||\psi_2|\cos\phi \qquad (6)

Why: the cross term is the only place phase survives the squaring, and it is precisely the fringes — for equal amplitudes $P$ swings between $4|\psi_1|^2$ and zero as $\cos^2(\phi/2)$ .

Step 3. Tag the paths. Let any system — a detector, a stray photon, one atom of the apparatus — end up in different states depending on the route. The routes are now distinguishable in principle, and the statistics become $P_{\text{dist}} = |\psi_1|^2 + |\psi_2|^2$ : probabilities add, the cross term is gone. Why: amplitudes for distinguishable final states of the world never interfere — the record blocks the cross term whether or not anyone reads it.

Step 4. State complementarity as arithmetic. Fringes ⇔ cross term ⇔ indistinguishable alternatives; which-path knowledge ⇔ no cross term ⇔ no fringes; partial distinguishability ⇔ partial visibility. Why: Bohr's principle is not philosophy bolted on — it is the difference between Steps 2 and 3. *The expansion $|\psi_1+\psi_2|^2 = |\psi_1|^2 + |\psi_2|^2 + 2\mathrm{Re}(\psi_1^*\psi_2)$ and the equal-amplitude form $4|\psi_0|^2\cos^2(\phi/2)$ follow directly from complex arithmetic.*

That completes the dictionary — four entries, one constant, and a rule for when to add what:

equation	connects	reach for it when
$E = h\nu$	color → energy	photon budgets: thresholds, spectra, detectors
$p = h/\lambda$	wavelength → momentum	light striking matter: Compton recoil, radiation pressure
$\lambda_{\mathrm{dB}} = h/p$	momentum → wavelength	matter diffracting: electron microscopes, crystal peaks
$P = \lvert\psi_1 + \psi_2\rvert^2$	amplitudes → statistics	indistinguishable paths — expect fringes; tagged paths — add $\lvert\psi\rvert^2$ 's instead

The fringe geometry itself — where the bright bands land, how the spacing $\lambda L/d$ follows from the phase in (6) — is derived in the Double-Slit Experiment article; the wavefunction and the Born rule get their own full treatments just ahead.

Limiting cases

Duality is governed by one comparison: the de Broglie wavelength against the feature size $d$ of the apparatus. The classical particle and classical wave pictures are both recovered, in opposite extremes:

Case	Result	Meaning
macroscopic object	$\lambda_{\mathrm{dB}} = h/p \ll$ any aperture	diffraction unobservable — classical trajectories emerge; a baseball has $\lambda \sim 10^{-34}\ \mathrm{m}$
$\lambda_{\mathrm{dB}} \sim$ lattice spacing	strong diffraction	the electron/neutron regime, $\sim 0.1\ \mathrm{nm}$ — where Davisson–Germer lived
many photons per mode	classical electromagnetic wave	graininess $h\nu$ is a negligible fraction of the energy; Maxwell recovered
which-path known	$P = \lvert\psi_1\rvert^2 + \lvert\psi_2\rvert^2$	cross term gone: particle statistics, no fringes

Common trap“The electron switches: sometimes it's a wave, sometimes a particle”

Nothing switches, splits, or shape-shifts. The electron is always one quantum object described by one amplitude $\psi$ . The amplitude is not a material wave — the electron is never physically smeared into a blob — and the dot on the screen is not the wave "collapsing into" a particle that secretly chose a slit. The wave lives in the probabilities; the particle lives in the detections. Interference appears whenever the alternatives are indistinguishable, and disappears the instant which-path information exists anywhere in the universe — even if nothing touched the electron. If you find yourself asking "but which slit did it really go through?", the honest quantum answer is that the question presumes a trajectory the theory does not contain.

Worked examples

1 · The potassium plate (confidence builder)

Violet light of wavelength $400\ \mathrm{nm}$ falls on potassium, work function $W = 2.30\ \mathrm{eV}$ . Find the maximum kinetic energy of the photoelectrons, the stopping potential, and the longest wavelength that ejects any electron at all.

Step 1. Photon energy. Using $hc = 1240\ \mathrm{eV\,nm}$ : $E = hc/\lambda = 1240/400 = 3.10\ \mathrm{eV}$ .

Step 2. Einstein's equation (built in the Photoelectric Effect article): $K_{\max} = 3.10 - 2.30 = \boxed{0.80\ \mathrm{eV}}$ .

Step 3. Stopping potential: $eV_0 = K_{\max} \Rightarrow V_0 = 0.80\ \mathrm{V}$ .

Step 4. Threshold wavelength: $\lambda_0 = hc/W = 1240/2.30 = 539\ \mathrm{nm}$ . Anything redder — including a megawatt of $633\ \mathrm{nm}$ laser light — ejects nothing, while the feeblest glimmer at $400\ \mathrm{nm}$ ejects electrons instantly.

2 · An X-ray plays billiards (exam level)

A molybdenum X-ray photon of wavelength $71.0\ \mathrm{pm}$ Compton-scatters through $90°$ off a nearly free electron. Find the scattered wavelength, both photon energies, and the electron's recoil energy.

Step 1. Shift from equation (3) at $\theta = 90°$ : $\Delta\lambda = \dfrac{h}{m_ec}(1 - \cos 90°) = 2.43\ \mathrm{pm} \times 1 = 2.43\ \mathrm{pm}$ — one full Compton wavelength.

Step 2. Scattered wavelength: $\lambda' = 71.0 + 2.43 = 73.43\ \mathrm{pm} \approx \boxed{73.4\ \mathrm{pm}}$ .

Step 3. Photon energies via $E = hc/\lambda = 1240\ \mathrm{eV\,nm}/\lambda$ : in, $E = 1240/0.0710 = 17.463\ \mathrm{keV}$ ; out, $E' = 1240/0.07343 = 16.886\ \mathrm{keV}$ — carrying the unrounded $\lambda'$ ; round it to $73.4$ first and the subtraction in Step 4 stops closing.

Step 4. The electron pockets the difference: $K_e = E - E' = 0.577\ \mathrm{keV} \approx \boxed{0.58\ \mathrm{keV}}$ — 3.3% of the incident energy, and some fifty times carbon's $11.3\ \mathrm{eV}$ outer-electron binding (the incident photon is over a thousand times it), justifying the "free electron" assumption after the fact.

Step 5. Sanity check: the shift is independent of $\lambda$ — a 500 nm photon suffers the same 2.43 pm, fractionally $5\times10^{-6}$ , unmeasurable; that is why Compton needed X-rays.

Origins · 1 minAlbert Einstein and how this was discovered

Albert Einstein

1879–1955 · Bern patent office, 1905

Einstein's photon paper — the one he called revolutionary, unlike relativity — took Planck's quantization out of the cavity walls and put it into light itself. The proposal was so radical that when Millikan spent a decade trying to disprove the photoelectric equation, his own data confirmed it perfectly; he measured $h$ to better than a percent while still calling the underlying idea "reckless." Einstein's 1921 Nobel Prize cited the photoelectric law, not relativity. He never stopped worrying about what the quantum meant: "All these fifty years of pondering have not brought me any closer to answering the question, What are light quanta?"

History

1905

— Einstein proposes the light quantum:

E = h\nu

taken as physically real, explaining the photoelectric threshold.

1916

— Millikan's stopping-potential measurements confirm

K_{\max} = h\nu - W

and yield

h

, against his own expectations.

1923

— Compton shows X-ray photons recoil off electrons carrying momentum

p = h/\lambda

: particles of light, caught colliding.

1924

— De Broglie's thesis inverts the logic: every particle of momentum

p

has a wavelength

\lambda = h/p

1927

— Davisson & Germer (nickel) and G. P. Thomson (thin foils) observe electron diffraction; the same year Bohr states complementarity.

Going further

Texts. Feynman Lectures Vol. III, Ch. 1 (the definitive double-slit discussion) · Griffiths & Schroeter, Introduction to Quantum Mechanics, 3rd ed., Ch. 1 · Townsend, A Modern Approach to Quantum Mechanics, Ch. 1 · Cohen-Tannoudji, Diu & Laloë, Vol. 1, Ch. 1 · Sakurai & Napolitano, Modern Quantum Mechanics, 2nd ed., Ch. 1. Key numerical anchors used throughout: the electron-wavelength coefficient $h/\sqrt{2m_e e}\approx1.226\,\mathrm{nm\,V^{1/2}}$ , giving the Davisson–Germer value $54\,\mathrm{V}\to0.167\,\mathrm{nm}$ ; the Compton wavelength $h/m_e c=2.426\times10^{-12}\,\mathrm{m}$ from relativistic energy-momentum conservation; the CODATA constants $h=6.626\times10^{-34}\,\mathrm{J\,s}$ and $\hbar=1.055\times10^{-34}\,\mathrm{J\,s}$ ; and the worked Compton result (71.0 pm at 90°: $\lambda' = 73.4\,\mathrm{pm}$ , electron recoil 0.58 keV).

Prerequisites. Breakdown of classical physics · blackbody radiation and the Planck quantum · photoelectric effect and the photon. Leads to. de Broglie matter waves · double-slit experiment · the wavefunction · Born statistical interpretation.

Does the electron interfere with itself, or with other electrons?

With itself — and "itself" means its amplitude, not its body. Double-slit experiments have been run at intensities so low that only one particle is ever in flight (one electron per minute, in Tonomura's version), and the fringes build up anyway — so interference between different electrons cannot be the explanation. What interferes are the two amplitudes for one electron's indistinguishable routes; each detection is a single dot, and the pattern is the statistics of many single-particle events.

Is the wave a real physical thing or just our ignorance about where the particle is?

Neither classical option survives. The amplitude cannot be mere ignorance about a definite trajectory: ignorance obeys $P = P_1 + P_2$ , and the dark fringes — places where opening a second slit reduces the count to zero — are impossible under that rule. But $\psi$ is not a material wave either: it is complex-valued, unobservable in a single event, and for two particles it lives in six-dimensional configuration space. Safe statement: $\psi$ is the object whose squared modulus gives detection statistics — its status is taken up in The Wavefunction and Born Statistical Interpretation.

Why does the same constant h appear in both E = hν and λ = h/p?

Because both are entries of one dictionary between the particle language (energy, momentum) and the wave language (frequency, wavelength), and one conversion constant must run both entries or the dictionary is inconsistent — for a photon, $E = h\nu$ plus $E = pc$ plus $\nu\lambda = c$ forces $p = h/\lambda$ . De Broglie's inversion extends the same pairing to matter. In the formalism ahead the two relations merge into $E = \hbar\omega$ , $p = \hbar k$ : Planck's constant is the exchange rate between dynamics and wave geometry.

If which-path information destroys interference, how gentle can the "measurement" be and still destroy it?

Gentleness is irrelevant; distinguishability is everything. The cross term in (6) dies as soon as any degree of freedom anywhere ends up in different states for the two routes — a detector click, a scattered photon, a flipped spin — and no energy transfer to the electron is required; conversely, a violent interaction that leaves no record preserves fringes. Partial records give partial visibility. And if the record is erased before detection, interference returns — the quantum eraser, treated with the entanglement machinery much later in the track.

Where exactly does the classical world begin, if duality never turns off?

There is no borderline where $\lambda = h/p$ stops holding — only a scale where its consequences become unobservable. A dust grain of $10^{-15}$ kg at 1 mm/s has $\lambda \sim 7\times10^{-16}$ m, smaller than a proton: no instrument can reveal such a wave, so trajectories become a flawless approximation. Decoherence finishes the story — air molecules and photons continuously tag macroscopic paths, killing cross terms faster than any experiment can look. Classicality is duality with its interference term buried, not repealed.

Compton's formula assumed a free electron at rest — real electrons are bound and moving. Why does the experiment work?

Scale separation. A 17 keV X-ray photon dwarfs the few-eV binding of carbon's outer electrons, so they respond as if free — the correction is parts per thousand. Their initial motion smears the shifted line rather than moving it; that smearing (the Compton profile) is itself used to map electron momenta in solids. Tightly bound inner electrons are different: the whole atom recoils, $m_e$ in (3) becomes the atomic mass, and the shift shrinks a thousandfold — exactly why Compton saw an unshifted line alongside the shifted one.