What Is Light?
From Maxwell's field equations to the quantum vacuum — questioning every assumption, deriving every result, accepting nothing on authority.
Prologue — The Genesis of a Question
On how curiosity, not curriculum, became the driver.
It began in Grade VIII — not with an answer, but with a discomfort. The textbooks described light as "electromagnetic radiation that enables vision." This is not a definition. It is a tautology dressed in jargon. It tells you what light does, not what light is.
By Grade IX, the discomfort had crystallised into action. I began fabricating lenses — not from polished glass, but from photopolymer drops. Different volumes. Different spherical profiles. Different refractive geometries. Each lens was a question cast in material form: what does curvature do to a wavefront, and why?
Scientists at the national laboratories in Dhanbad and Delhi did not dismiss these questions. They invited further discussion. They gave access to chemicals used in lens fabrication — optical-grade polymers, UV-curable resins. They showed me an electron microscope: a device that renders the wave nature of matter into an imaging instrument. Standing in front of that machine in Class IX, I understood that these are not merely complex devices — they are theorems made physical.
The work on novel lens designs and new approaches to microscope construction was selected at the state-level science exhibition, and eventually won at the national level. That recognition led to a science award presented by the President of India. The research also earned me a National Merit Scholarship that supported my education through the completion of college. But the awards were never the point. The point was that a question asked in Grade VIII — what is light, really? — had proven to be deep enough that serious scientists took it seriously, and rigorous enough that it survived contact with real laboratories, real chemicals, and real instruments. The question did not go away. It got larger.
This document is an attempt to reconstruct the theory of light from nothing — to start with the most primitive physical assumptions and derive everything. Every equation will be justified. Every "obvious" step will be questioned. Nothing is accepted on authority.
The central question that drives everything that follows: What is the minimal set of physical principles from which all optical phenomena — reflection, refraction, diffraction, interference, polarisation, photon emission, lens behaviour, and microscope resolution — can be derived, not postulated?
What Is Light? — Refusing the Textbook Answer
Strip everything. Start from what we can measure.
Before writing a single equation, we must decide what we actually know about light through direct measurement, independent of any theoretical framework:
Empirical Fact 1. Light carries energy. A photovoltaic cell converts incident light into measurable electric current. A focused beam can ablate material. Energy is frame-independent and real.
Empirical Fact 2. Light carries momentum. Radiation pressure has been measured since Lebedev (1900) and Nichols & Hull (1901). Solar sails are engineered around this fact.
Empirical Fact 3. Light exhibits interference and diffraction — phenomena characteristic of wave behaviour. Young's double-slit experiment (1801) is reproducible to arbitrary precision.
Empirical Fact 4. Light exhibits particle-like behaviour. The photoelectric effect (measured by Hertz, 1887; explained by Einstein, 1905) shows energy exchange in discrete quanta, $E = h\nu$.
Empirical Fact 5. Light propagates at a finite, invariant speed $c \approx 2.998 \times 10^8$ m/s in vacuum, independent of the motion of source or observer (Michelson–Morley, 1887).
These five facts are mutually incompatible under any single naïve model. A classical wave requires a medium (but no luminiferous aether exists). A classical particle cannot interfere with itself. How do we reconcile these? The answer will require us to abandon both "wave" and "particle" as fundamental categories and replace them with something deeper: the quantised field.
But we must earn that conclusion. Let us start with the classical theory — Maxwell's electrodynamics — and push it until it breaks.
The Electromagnetic Field: Maxwell's Equations in Full
Four equations. Two fields. The entire classical theory of light.
We begin with the four Maxwell equations in their differential form, written in SI units. These are not axioms we must accept blindly — each one is a precise mathematical encoding of an experimental observation.
3.1 Gauss's Law for Electricity
Here $\mathbf{E}$ is the electric field (a vector field on $\mathbb{R}^3$), $\rho$ is the volume charge density, and $\varepsilon_0 \approx 8.854 \times 10^{-12}$ F/m is the permittivity of free space. The divergence operator $\nabla \cdot$ measures the net outward flux per unit volume. Physically: electric charges are sources and sinks of the electric field.
In integral form (via the Divergence Theorem of Gauss-Ostrogradsky):
Why does this law have the specific form $1/\varepsilon_0$? Is $\varepsilon_0$ fundamental, or is it an artefact of our choice of units? In Gaussian units, $\varepsilon_0$ disappears entirely and Coulomb's law becomes $F = q_1 q_2 / r^2$. The constant $\varepsilon_0$ encodes nothing physical about light itself — it is a conversion factor between mechanical and electromagnetic units. This matters: it means $c$ is already hiding inside these equations.
3.2 Gauss's Law for Magnetism
The magnetic field $\mathbf{B}$ is divergence-free everywhere. This is an empirical statement: no isolated magnetic monopole has ever been observed. Every field line of $\mathbf{B}$ closes upon itself. Topologically, $\mathbf{B}$ is a solenoidal vector field.
In integral form:
The net magnetic flux through any closed surface is identically zero. This constrains the topology of the field: $\mathbf{B}$ can always be written as the curl of a vector potential, $\mathbf{B} = \nabla \times \mathbf{A}$, since $\nabla \cdot (\nabla \times \mathbf{A}) = 0$ identically (a consequence of the antisymmetry of mixed partial derivatives).
3.3 Faraday's Law of Induction
A time-varying magnetic field generates a circulating electric field. The curl $\nabla \times \mathbf{E}$ measures the infinitesimal circulation of $\mathbf{E}$. The minus sign is Lenz's law: the induced field opposes the change that produces it. In integral form (via Stokes' theorem):
This is the principle behind every transformer, every generator, every induction cooktop. But for our purposes, it is half of the mechanism by which light propagates: a changing $\mathbf{B}$ creates $\mathbf{E}$.
3.4 The Ampère–Maxwell Law
Here $\mu_0 = 4\pi \times 10^{-7}$ H/m is the permeability of free space and $\mathbf{J}$ is the current density. The first term ($\mu_0 \mathbf{J}$) is Ampère's original law: currents produce circulating magnetic fields. The second term ($\mu_0 \varepsilon_0 \, \partial \mathbf{E}/\partial t$) is Maxwell's displacement current — the single most consequential addition in the history of physics.
Why did Maxwell add the displacement current? Without it, taking the divergence of Ampère's law gives $\nabla \cdot (\nabla \times \mathbf{B}) = \mu_0 \nabla \cdot \mathbf{J}$, but the left side is identically zero. This forces $\nabla \cdot \mathbf{J} = 0$, which contradicts charge conservation when $\partial \rho / \partial t \neq 0$. The continuity equation demands $\nabla \cdot \mathbf{J} = -\partial \rho / \partial t$. Maxwell's fix: add a term whose divergence supplies the missing $\partial \rho / \partial t$. Since $\nabla \cdot \mathbf{E} = \rho / \varepsilon_0$, the term $\varepsilon_0 \partial \mathbf{E} / \partial t$ has divergence $\partial \rho / \partial t$. Consistency, not aesthetics, forced Maxwell's hand. And from that consistency, light emerges.
Deriving the Wave Equation from First Principles
No hand-waving. Every step shown, every identity invoked.
We now derive the electromagnetic wave equation in free space (where $\rho = 0$, $\mathbf{J} = \mathbf{0}$). This is the derivation that proves light is an electromagnetic phenomenon.
4.1 Source-Free Maxwell Equations
In vacuum, Maxwell's equations reduce to:
4.2 The Full Derivation
An identical derivation (curl Ampère → substitute Faraday) yields:
Both $\mathbf{E}$ and $\mathbf{B}$ satisfy the same wave equation, coupled through Maxwell's equations. Neither field propagates alone — they regenerate each other. This mutual induction is light.
The Speed of Light as a Structural Constant
$c$ is not measured. It is calculated — and then confirmed.
From the wave equation, the propagation speed is:
This is arguably the most profound result in all of physics. Maxwell computed this value in 1865 and recognised it as the measured speed of light. He wrote: "The agreement of the results seems to show that light and magnetism are affections of the same substance." But we should ask a deeper question: why do $\mu_0$ and $\varepsilon_0$ conspire to produce exactly this speed? The answer requires special relativity: $c$ is not a property of electromagnetism. It is a property of spacetime itself. It is the speed at which causality propagates.
5.1 Plane Wave Solutions
The simplest solution to the wave equation is the monochromatic plane wave. Let $\mathbf{E}$ propagate in the $\hat{z}$-direction:
where $k = 2\pi/\lambda$ is the wavenumber, $\omega = 2\pi\nu$ is the angular frequency, and the dispersion relation is:
Substituting back into Faraday's law gives the companion magnetic field:
Key structural facts: $\mathbf{E} \perp \mathbf{B} \perp \hat{\mathbf{k}}$ (transversality); $|\mathbf{E}| = c|\mathbf{B}|$ (amplitude relation); $\mathbf{E}$ and $\mathbf{B}$ are in phase (no lag).
5.2 Complex Notation and the Phasor Representation
For computational convenience we write:
where $\tilde{E}_0 = E_0 e^{i\phi_0} \in \mathbb{C}$, and the physical field is $\mathbf{E} = \text{Re}(\tilde{\mathbf{E}})$. This is not merely notation — it reflects the underlying $U(1)$ symmetry of the electromagnetic field, which will become the gauge symmetry of quantum electrodynamics.
Energy, Momentum, and the Poynting Vector
Light carries real, measurable quantities through empty space.
6.1 Energy Density of the Electromagnetic Field
The total electromagnetic energy stored per unit volume is:
For a plane wave with $|\mathbf{B}| = |\mathbf{E}|/c$ and $c = 1/\sqrt{\mu_0\varepsilon_0}$:
The electric and magnetic contributions are exactly equal. This is not a coincidence — it reflects the Lorentz invariance of the electromagnetic field. In any other frame, the partition changes, but the total energy density (as part of the stress-energy tensor $T^{\mu\nu}$) transforms covariantly.
6.2 The Poynting Vector: Energy Flux
$\mathbf{S}$ gives the energy flow per unit area per unit time (units: W/m²). For a plane wave propagating in $\hat{\mathbf{z}}$:
The time-averaged intensity (what we actually measure with a detector) is:
6.3 Poynting's Theorem — Energy Conservation
6.4 Radiation Pressure and Momentum
The electromagnetic field carries momentum density:
For a plane wave fully absorbed by a surface, the radiation pressure is:
For perfect reflection, $P_{\text{rad}} = 2I/c$ (momentum reversal doubles the impulse). This is the physical principle behind solar sails and optical trapping ("optical tweezers").
The Failure of Classical Theory — Toward Quantisation
Where Maxwell's beautiful framework cracks.
The classical theory derived above is extraordinarily successful. But by 1900, three experimental results had exposed fatal contradictions:
7.1 The Ultraviolet Catastrophe
The Rayleigh–Jeans law, derived from classical statistical mechanics applied to electromagnetic modes in a cavity, predicts the spectral energy density:
This diverges as $\nu \to \infty$: the total energy $\int_0^\infty u(\nu)\,d\nu \to \infty$. A blackbody at any finite temperature should radiate infinite energy — absurd.
Planck's resolution (1900): assume energy exchange between matter and radiation occurs in discrete quanta $E = nh\nu$, where $n \in \mathbb{N}_0$. The average energy of a mode at frequency $\nu$ becomes:
yielding the Planck distribution:
7.2 The Photoelectric Effect
Classically, light of any frequency should eject electrons from a metal surface if the intensity is high enough (energy accumulates over time). Experimentally:
(a) Below a threshold frequency $\nu_0$, no electrons are ejected regardless of intensity.
(b) Above $\nu_0$, electron kinetic energy increases linearly with frequency, not intensity.
Einstein's explanation (1905): light arrives as discrete quanta (photons), each carrying energy $E = h\nu$. The maximum kinetic energy of an ejected electron is:
where $\phi = h\nu_0$ is the work function of the metal. This is not a classical wave phenomenon. The energy is localised in quanta.
7.3 Compton Scattering
When X-rays scatter off electrons (Compton, 1923), the wavelength shift depends on the scattering angle $\theta$:
This is derived from relativistic energy-momentum conservation treating the photon as a particle with energy $E = h\nu = pc$ and momentum $p = h/\lambda$. No classical wave theory can produce this result.
The Photon: Quantised Excitation of the EM Field
Neither wave nor particle. Something fundamentally new.
8.1 Canonical Quantisation of the Free Field
The classical electromagnetic field in a cubic cavity of volume $V = L^3$ can be decomposed into normal modes (Fourier expansion). Each mode of wavevector $\mathbf{k}$ and polarisation $\lambda$ behaves as a simple harmonic oscillator with Hamiltonian:
Quantise by promoting the canonical variables to operators satisfying $[\hat{q}, \hat{p}] = i\hbar$, and define creation/annihilation operators:
These satisfy the bosonic commutation relations:
The quantised Hamiltonian becomes:
where $\hat{n}_{\mathbf{k},\lambda} = \hat{a}^\dagger \hat{a}$ is the number operator with eigenvalues $n = 0, 1, 2, \ldots$
A photon is a single quantum of excitation of a mode $(\mathbf{k}, \lambda)$ of the electromagnetic field. The state $|n_{\mathbf{k},\lambda}\rangle$ contains $n$ photons, each with energy $\hbar\omega_k$ and momentum $\hbar\mathbf{k}$. The photon is not "a particle of light" in the classical sense — it is an eigenvalue of the number operator acting on a Fock state of the quantised field.
8.2 Zero-Point Energy and the Quantum Vacuum
Even in the ground state $|0\rangle$ (no photons), the field has energy:
This sum diverges — it is the vacuum energy of the electromagnetic field. It gives rise to measurable effects: the Casimir force (measured by Lamoreaux, 1997), the Lamb shift (measured by Lamb & Retherford, 1947), and spontaneous emission. The vacuum is not empty. It seethes with zero-point fluctuations of every mode of the field.
The vacuum energy density, when summed over all modes up to the Planck scale, exceeds the observed cosmological constant by approximately $10^{120}$ orders of magnitude. This is the worst prediction in the history of physics. Either our understanding of vacuum energy is fundamentally wrong, or something cancels it with extraordinary precision. This is an open problem.
Optics from First Principles — Refraction, Lenses, and the Experiments
Connecting field theory to curved surfaces and photopolymer drops.
The lens experiments in Grade IX were not "science fair projects." They were attempts to answer a precise question: if I change the radius of curvature of a liquid photopolymer drop — keeping the refractive index fixed but varying the geometry — how does the focal length respond? And if I change the polymer concentration (altering $n$) while holding the geometry fixed, what happens? These two degrees of freedom — $n$ and $R$ — are the complete parameter space of a thin spherical lens. Every lens ever made is a point in this space.
9.1 Refraction from Maxwell's Equations: Snell's Law Derived
When an electromagnetic wave crosses a boundary between two dielectric media with refractive indices $n_1$ and $n_2$, the boundary conditions on $\mathbf{E}$ and $\mathbf{B}$ (continuity of tangential components) enforce:
9.2 The Refractive Index as a Material Response
The refractive index $n$ is defined as $n = c/v_{\text{phase}}$, but what causes it? In a dielectric, the electric field of the incoming wave polarises the medium — it displaces the electron clouds relative to the nuclei. Each atom becomes an oscillating dipole. The superposition of the incident wave and the radiated field from these dipoles produces a net wave that travels at $v_{\text{phase}} < c$.
The Lorentz oscillator model gives, for a medium with $N$ oscillators per unit volume:
where $\omega_{0j}$ are resonant frequencies, $\gamma_j$ are damping coefficients, and $f_j$ are oscillator strengths. This is a complex function — the real part gives refraction, the imaginary part gives absorption. Dispersion (the dependence of $n$ on $\omega$) arises naturally: different frequencies interact differently with the electronic structure of the material.
9.3 The Lensmaker's Equation
For a thin lens with surfaces of radii $R_1$ and $R_2$, refractive index $n_L$ embedded in a medium of index $n_m$:
For a single refracting surface (relevant to a liquid drop where one side is curved and the other is approximately flat, $R_2 \to \infty$):
This is exactly what the photopolymer experiments tested. A drop of liquid photopolymer on a flat glass slide forms a plano-convex lens. By varying the drop volume, the radius of curvature $R$ changes (smaller drops → smaller $R$ → shorter $f$). By varying the polymer concentration, the refractive index $n_L$ changes. The two parameters are decoupled. The experiments were, in effect, a physical scan of the $(n, R)$ parameter space of the lensmaker's equation — without knowing the equation existed.
9.4 Fermat's Principle and the Calculus of Variations
All of geometrical optics — every ray-tracing law, every lens equation — can be derived from a single variational principle:
Light travels along the path that makes the optical path length stationary. The integral $\int n \, ds$ is the optical path length. The Euler–Lagrange equation for this functional yields the ray equation:
In a homogeneous medium ($\nabla n = 0$), rays are straight lines. At an interface where $n$ is discontinuous, the variational condition reproduces Snell's law. In a continuously varying medium (like a gradient-index lens), rays curve — exactly as they do in the gravitational lensing of general relativity, where the effective refractive index is determined by the spacetime metric.
The deep question: why does light obey a variational principle at all? In quantum mechanics, Fermat's principle is the geometrical-optics limit of the path integral. Every photon takes all paths simultaneously; the classical path is the one where the phases constructively interfere (stationary phase approximation). The variational principle is not fundamental — it is an emergent consequence of quantum superposition.
The Electron Microscope — Beyond the Diffraction Limit
When light is not enough: de Broglie, diffraction, and the matter-wave revolution.
Seeing the electron microscope at the national laboratory was a turning point. Here was a device that achieves nanometre-scale resolution — not by using better lenses or brighter light, but by abandoning light entirely and using electrons as the imaging wave. The principles of optics are preserved (focussing, aberration, diffraction), but the wavelength changes by five orders of magnitude. The question became: what sets the resolution limit, and why?
10.1 The Abbe Diffraction Limit
The resolving power of any wave-based imaging system is fundamentally limited by diffraction. For a lens of numerical aperture $\text{NA} = n\sin\alpha$, the minimum resolvable distance is:
For visible light ($\lambda \approx 550$ nm) with the best oil-immersion objective ($\text{NA} \approx 1.4$):
This is a hard physical limit. No improvement in lens quality can overcome it — it is a consequence of the wave nature of light itself, encoded in the Fourier transform relationship between the aperture function and the point spread function.
10.2 The de Broglie Hypothesis: Matter Waves
De Broglie (1924) proposed that every particle of momentum $p$ has an associated wavelength:
For an electron accelerated through $V = 100$ kV:
An electron microscope is, at its core, an optical instrument. It has "lenses" (electromagnetic coils that bend electron trajectories via the Lorentz force $\mathbf{F} = e(\mathbf{E} + \mathbf{v} \times \mathbf{B})$). It has apertures, focal planes, and aberrations — all described by the same formalism developed for light optics. The theory of light, pushed to its logical end, reveals that light is just one instance of a more general principle: wave-based imaging. The electron microscope proves that the concepts of optics transcend the electromagnetic spectrum.
Open Questions and Forward Directions
What we still don't know.
Every derivation in this document rests on assumptions that can be questioned:
1. Why $U(1)$? The gauge symmetry of electrodynamics is $U(1)$ — the simplest continuous symmetry group. Why not $SU(2)$, $SU(3)$, or something else? The Standard Model answers this partially (electromagnetism is the unbroken $U(1)_{\text{EM}}$ subgroup after electroweak symmetry breaking), but the deeper question — why these gauge groups? — remains open.
2. Why is $\alpha \approx 1/137$? The fine-structure constant $\alpha = e^2/(4\pi\varepsilon_0 \hbar c) \approx 1/137.036$ determines the strength of all electromagnetic interactions. It is dimensionless and therefore cannot depend on our choice of units. Its value appears to be a fundamental property of the universe. No theory predicts it from first principles. Feynman called it "one of the greatest damn mysteries of physics."
3. The vacuum energy problem. As noted in §8.2, the predicted vacuum energy density exceeds observation by $\sim 10^{120}$. Either our method of summing zero-point energies is wrong, or an unknown cancellation mechanism exists.
4. Quantum gravity and the photon. Does the photon acquire a (tiny) mass due to quantum gravitational effects? Current experimental bounds place $m_\gamma < 10^{-18}$ eV/$c^2$. A nonzero photon mass would modify Maxwell's equations (Proca equations), break gauge invariance, and have profound implications for the structure of the vacuum.
5. The measurement problem. When a photon's wavefunction "collapses" upon detection, what physically happens? The Copenhagen interpretation, many-worlds, decoherence — all attempt to answer this, none fully satisfactorily. The photon, as the simplest quantum system, remains the primary arena for this debate.
This document is not finished. It cannot be. The questions that began with photopolymer drops in a school laboratory — why does curvature bend light? what is light? why does it have exactly these properties? — have no final answer. They have only deeper questions. The purpose of this framework is not to arrive, but to know precisely how far we have come, and to see clearly where the road still leads.
The Unsolved Universe: Where Light Holds the Missing Links
The deepest mysteries of the cosmos are, at their core, questions about light.
There is a pattern hiding in plain sight. The greatest unsolved problems in physics — dark energy, dark matter, quantum gravity, the matter-antimatter asymmetry, the nature of the Big Bang, the black hole information paradox — are not disconnected puzzles. They are all, in precise and demonstrable ways, questions about the behaviour, propagation, and quantum structure of light. Until we resolve the open questions about photons and the electromagnetic field laid out in this document, these cosmological mysteries will remain unsolved. The theory of light is not a finished chapter in physics. It is the thread that, once pulled correctly, unravels the rest.
Let us make this claim precise, mystery by mystery.
12.1 Dark Energy and the Vacuum Catastrophe
The accelerating expansion of the universe, discovered via Type Ia supernovae observations (Riess et al. 1998; Perlmutter et al. 1999), is attributed to a "dark energy" with an effective equation of state $w = P/\rho \approx -1$. The simplest candidate is the cosmological constant $\Lambda$, which enters Einstein's field equation as:
The observed value of $\Lambda$ corresponds to an energy density of approximately:
But recall from Section 8.2: the quantum vacuum of the electromagnetic field has a zero-point energy. Every mode $(\mathbf{k}, \lambda)$ contributes $\frac{1}{2}\hbar\omega_k$ to the ground state. Summing over all modes up to a UV cutoff $\Lambda_{\text{UV}}$, the vacuum energy density scales as:
If we set the cutoff at the Planck scale, $k_{\text{max}} \sim \ell_P^{-1} = \sqrt{c^3/\hbar G}$:
This is the vacuum catastrophe — the worst quantitative disagreement between theory and observation in all of science. The calculation that produces this absurd number is a direct consequence of the quantised electromagnetic field, the same field theory derived in Sections 3 through 8 of this document. Dark energy is not a separate mystery from light. It is a mystery about light — about what happens when we sum the zero-point fluctuations of every photon mode in the universe. Solve the vacuum energy problem, and you solve dark energy. The dots are the same dots.
12.2 Dark Matter and the Photon Coupling Question
Approximately 27% of the energy content of the universe is in the form of dark matter — gravitationally interacting, but electromagnetically invisible. The evidence is overwhelming: galaxy rotation curves, gravitational lensing, CMB anisotropies, large-scale structure formation. But the defining property of dark matter is a negative one: it does not emit, absorb, or scatter photons.
In the language of quantum field theory, dark matter has zero electromagnetic coupling. If dark matter particles carry no electric charge $q$, their interaction vertex with the photon field vanishes:
But this raises a profound question. In the Standard Model, all known massive particles carry some form of gauge charge. The electron couples to $U(1)_{\text{EM}}$; quarks couple to both $U(1)_{\text{EM}}$ and $SU(3)_C$. Even the neutrino, though electrically neutral, participates in weak $SU(2)_L$ interactions. What kind of particle has mass, participates in gravity, but has exactly zero coupling to the photon field?
The search for dark matter is, at a fundamental level, an investigation into the completeness of the theory of light. We are asking: does the photon couple to everything that has energy? General relativity says yes — all energy gravitates. But the photon field says no — only charged matter radiates. This asymmetry between gravity and electromagnetism, between the universal coupling of $g_{\mu\nu}$ and the selective coupling of $A_\mu$, is one of the deepest structural puzzles in physics. Dark matter is the empirical proof that the photon does not see the entire universe. Understanding why requires understanding the photon's place in the full gauge structure of nature.
12.3 The Matter-Antimatter Asymmetry and CP Violation in Photon Interactions
The observable universe contains approximately $10^{80}$ baryons and essentially zero primordial antibaryons. Yet the Standard Model is nearly symmetric between matter and antimatter. The Sakharov conditions (1967) require three ingredients for baryogenesis: baryon number violation, C and CP violation, and departure from thermal equilibrium.
The known sources of CP violation in the Standard Model — the CKM phase in the quark sector and the PMNS phase in the neutrino sector — are quantitatively insufficient to explain the observed asymmetry by many orders of magnitude. But consider: every measurement of CP violation is ultimately an electromagnetic measurement. We detect the decay products through their interactions with photons — scintillation, Cherenkov radiation, ionisation tracks, calorimetric energy deposition.
The electromagnetic field is CP-even. This means pure QED cannot distinguish matter from antimatter. The photon treats an electron and a positron identically up to charge sign. Yet the universe has chosen one over the other. Either there exist interactions beyond QED that violate CP more strongly than anything we have measured, or our understanding of how the photon field behaves at extreme energies (near the GUT or Planck scale) is incomplete.
The asymmetry between matter and antimatter is a question about symmetry breaking — and electromagnetism is the primary tool through which we probe symmetries. If light cannot distinguish matter from antimatter, then the mechanism that created the asymmetry must operate in a regime where the photon as we know it did not yet exist: the epoch before electroweak symmetry breaking, when the $U(1)_{\text{EM}}$ photon had not yet separated from the $SU(2)_L \times U(1)_Y$ gauge bosons. The matter-antimatter mystery leads us directly to the question: what was the photon before it became the photon?
12.4 Black Holes, the Information Paradox, and Hawking Radiation
A black hole is defined by its event horizon — the surface beyond which no signal, including light, can escape. The event horizon is fundamentally an electromagnetic boundary: it is the locus where outgoing null geodesics (the paths light follows in curved spacetime) become trapped. The Schwarzschild radius is:
Note that $c$ appears in the denominator. The event horizon is defined in terms of the speed of light. A black hole is, in the most literal sense, a region of spacetime where the theory of light predicts its own limitation: there exist trajectories from which photons cannot reach future null infinity.
Hawking's calculation (1974) showed that quantum effects near the horizon cause black holes to radiate thermally at temperature:
This radiation is predominantly photons (for large black holes). Every fundamental constant of the photon appears: $\hbar$ (quantisation), $c$ (propagation), $k_B$ (thermality). The Hawking process is a prediction that emerges from applying exactly the quantum field theory of Section 8 — quantised electromagnetic modes in curved spacetime — to the vicinity of an event horizon.
The information paradox asks: when a black hole evaporates completely via Hawking radiation, what happens to the quantum information of everything that fell in? If the radiation is perfectly thermal (as Hawking's semiclassical calculation suggests), then information is destroyed, violating unitarity — a foundational axiom of quantum mechanics. The resolution, whatever it may be, must be expressible as a statement about photons: about the quantum state of the electromagnetic field near a horizon, about whether subtle correlations (entanglement) are encoded in the outgoing Hawking photons. The information paradox is not merely adjacent to the theory of light. It is a paradox within the theory of light, applied at the most extreme gravitational boundary nature provides.
12.5 Quantum Gravity: Where Spacetime and the Photon Meet
General relativity describes gravity as the curvature of spacetime. Quantum field theory describes photons as excitations of the electromagnetic field on a fixed spacetime background. Neither theory knows how to accommodate the other at the Planck scale:
At distances approaching $\ell_P$, the smooth spacetime manifold on which Maxwell's equations are defined is expected to break down. The very notion of "propagation" of a photon becomes ill-defined if spacetime itself is subject to quantum fluctuations. Does the concept of a light cone survive at the Planck scale? Is $c$ still invariant when the metric itself fluctuates? Can the photon acquire an effective mass from quantum gravitational corrections?
where $E_P = \sqrt{\hbar c^5 / G} \approx 1.22 \times 10^{19}$ GeV is the Planck energy and $\eta$ is a dimensionless parameter that different quantum gravity models predict differently. Loop quantum gravity suggests $\eta \sim \mathcal{O}(1)$; string theory models often predict $\eta = 0$ at leading order. The correction term $\eta E^3 / E_P$ would cause an energy-dependent speed of light — a violation of Lorentz invariance at Planck-suppressed order.
Remarkably, this is testable. Gamma-ray bursts (GRBs) emit photons across a vast energy range ($\sim$keV to $\sim$TeV) from cosmological distances ($z \sim 1$ to $z \sim 8$). If high-energy photons travel at a speed slightly different from low-energy photons, they arrive at slightly different times. Current observations from Fermi-LAT constrain:
Quantum gravity is the most fundamental unsolved problem in theoretical physics. And the most sensitive experimental probe we have is the photon. Because photons travel at $c$ and arrive from cosmological distances, even Planck-scale corrections accumulate into measurable time delays over billions of light-years. The photon is not just a messenger that passively reports on spacetime; it is the primary diagnostic tool for testing whether spacetime itself is quantised. If quantum gravity modifies the propagation of light, however slightly, photons from the distant universe will carry that signature. The theory of light is the experimental frontier of quantum gravity.
12.6 The Cosmic Microwave Background: A Photograph in Light
The CMB is the oldest light in the universe — photons that last scattered at the surface of last scattering, approximately 380,000 years after the Big Bang, at a redshift $z \approx 1100$. Every cosmological parameter we measure — $H_0$, $\Omega_m$, $\Omega_\Lambda$, $\Omega_b$, $n_s$, $\sigma_8$ — is extracted from the angular power spectrum of temperature and polarisation fluctuations in this primordial light:
where $C_\ell^{TT}$ is the angular power spectrum at multipole $\ell$, $P(k)$ is the primordial power spectrum, and $\Delta_{T,\ell}(k)$ is the transfer function encoding the physics of photon-baryon oscillations before recombination.
The acoustic peaks in the CMB power spectrum arise from standing sound waves in the photon-baryon fluid. The physics governing these oscillations is electromagnetic at its core: it is the radiation pressure of photons (computed via the Poynting vector of Section 6) that provides the restoring force against gravitational compression. The first peak at $\ell \approx 220$ tells us the universe is spatially flat. The second and third peaks encode the baryon density and dark matter density. Every single one of these measurements is a measurement of how photons behaved in the early universe.
The CMB also contains anomalies that remain unexplained: the hemispherical asymmetry, the Cold Spot, the alignment of low multipoles (the "axis of evil"), and a possible tension in $H_0$ between CMB-derived values and local measurements (the "Hubble tension," $H_0^{\text{CMB}} \approx 67.4$ vs $H_0^{\text{local}} \approx 73.0$ km/s/Mpc). These discrepancies may indicate new physics. But whatever that new physics is, it must be compatible with the photon transport theory encoded in the Boltzmann equations for photon distribution functions. New physics, whatever form it takes, must pass through the theory of light to reach our detectors.
12.7 The Origin of the Universe: Light at $t = 0$
At the earliest moments after the Big Bang, the universe was a plasma of quarks, leptons, and gauge bosons at temperatures above the electroweak scale ($T \gtrsim 100$ GeV). At these energies, the $U(1)_{\text{EM}}$ photon did not exist as a distinct entity. It was part of the unified electroweak gauge field $SU(2)_L \times U(1)_Y$. The "photon" as we know it — the massless eigenstate of the electromagnetic field — only emerged after electroweak symmetry breaking, when the Higgs field acquired its vacuum expectation value:
where $\theta_W \approx 28.7°$ is the Weinberg angle, $B_\mu$ is the $U(1)_Y$ gauge field, and $W_\mu^3$ is the neutral component of $SU(2)_L$. Before this transition, there was no "light" in the sense we have been discussing — only a more symmetric structure from which light was born.
If we push further back — to the GUT scale ($\sim 10^{16}$ GeV) or the Planck scale ($\sim 10^{19}$ GeV) — even the electroweak structure dissolves. Grand Unified Theories propose that $SU(3)_C \times SU(2)_L \times U(1)_Y$ unifies into a single gauge group (such as $SU(5)$, $SO(10)$, or $E_6$). In such a theory, the photon is just one component of a much larger gauge field. Quarks and leptons become interchangeable. The distinction between strong, weak, and electromagnetic forces disappears. Understanding the origin of the photon — how it emerged as a distinct, massless gauge boson from a higher symmetry — is equivalent to understanding the origin of the forces of nature. And the origin of the forces is the origin of the universe as we know it.
12.8 Connecting the Dots
Let us now step back and make the pattern explicit. Every unsolved mystery above connects to the theory of light through a specific, identifiable mechanism:
This is not a rhetorical observation. It is a structural fact about how physics is organised. The electromagnetic field is the one field that participates in every measurement, at every energy scale, from every distance. It is the carrier of astronomical observation (telescopes), the probe of atomic structure (spectroscopy), the medium of information transfer (fibre optics, radio), and the theoretical foundation of both classical and quantum physics. Every other field — strong, weak, gravitational — is observed through its effects on photons or on charged particles that subsequently emit photons.
The photon is the universal intermediary between the universe and the observer. Until we complete the theory of the photon — resolving its vacuum energy, understanding its place in grand unification, determining whether its masslessness is exact, clarifying its quantum behaviour near horizons and singularities — the great mysteries of the cosmos will remain precisely that: mysteries seen dimly, through the glass of an incomplete theory of light.
The questions that began in a school laboratory have not shrunk as the theory grew. They have expanded. The photopolymer drops tested the lensmaker's equation; the lensmaker's equation follows from Snell's law; Snell's law follows from Maxwell's boundary conditions; Maxwell's equations follow from gauge invariance; gauge invariance connects to electroweak unification; electroweak unification connects to the origin of the universe itself. The thread is continuous. Pull on one end — a curved lens, a bent beam of light — and you are pulling on the fabric of every unsolved question in fundamental physics. This is not metaphor. It is the actual logical structure of the theory.
This is a living document. Every equation is an invitation to go deeper.
Every derivation is a challenge: can you find a simpler path to the same truth?