Special relativity is usually taught as a response to the Michelson-Morley experiment. Physicists looked for the ether, failed to find it, and Einstein explained why. It is a tidy story and mostly wrong. Einstein mentions the null ether-drift results once, in passing, as a secondary consideration. What the paper actually opens with is something much stranger: a complaint about an asymmetry in how electromagnetic theory describes a situation where nothing physical has changed.

The Problem

Take a magnet and a conductor. Move the magnet toward the stationary conductor, and Maxwell's theory says the changing magnetic field creates an electric field, which drives a current. Now hold the magnet still and move the conductor instead. Now there is no electric field; the charges in the conductor experience a magnetic force because they are moving through the field. Different mechanism, different intermediate quantities, entirely different account of what is happening.

The current is identical. It depends only on the relative motion.

This bothered Einstein in a way it apparently bothered nobody else. The theory gives the right answer twice by two unrelated routes, which suggests the distinction between the two cases is not real. But Maxwell's equations single out a preferred frame — the one in which the ether sits still — and it is that preferred frame that makes the two descriptions different. If no experiment can tell you which object is "really" moving, the theory is carrying structure that does not correspond to anything.

Background Science

The ether was not a fringe idea. Light is a wave, waves propagate in a medium, therefore something must be waving. Maxwell's equations produce a specific speed for light, c, and a speed has to be measured relative to something. The ether was that something.

The trouble is that motion through the ether should be detectable, and it never was. Michelson and Morley found no effect in 1887. Lorentz and FitzGerald responded by proposing that objects contract along their direction of motion by exactly enough to hide the effect — a real physical compression caused by moving through the ether. Lorentz worked out the full transformation equations, the same ones that appear in Einstein's paper. He had the mathematics before Einstein did.

What he did not have was a reason. In Lorentz's version, the contraction is a dynamical accident, and there is still a true rest frame that we are simply prevented from finding.

What He Did

Einstein discarded the machinery and started from two postulates. First, the laws of physics take the same form in all inertial frames. Second, light travels at c regardless of the motion of its source.

Taken together these are close to absurd. If I chase a light beam at 99% of c, the second postulate says it still recedes from me at c. That cannot be true of any ordinary object, and it forces a conclusion about time and space rather than about light.

Einstein's method was to ask what "simultaneous" means operationally. Two events at different places are simultaneous if light signals from them, sent under specified conditions, arrive together — there is no other way to establish it. Follow that definition through, and observers in relative motion disagree about which events are simultaneous. Not because their clocks are faulty, but because simultaneity is not a property of the events.

Everything else drops out of that. Moving clocks run slow, moving objects contract, velocities add nonlinearly. The Lorentz transformations reappear, now as consequences of what measurement means rather than as a description of how matter deforms in a headwind.

The ether does not get disproven. It gets dismissed in a sentence as superfluous, which is a considerably more aggressive move.

Why It Mattered

The equations were not new. Lorentz had them, and Poincaré was close to the interpretation. What changed was that the same mathematics stopped describing distortions of objects moving through a medium and started describing the structure of space and time. Nothing needed to contract because there was no preferred frame to contract relative to.

This also removed the asymmetry that started the paper. Electric and magnetic fields turn out to be the same object viewed from different frames. Whether you attribute the current to an electric field or a magnetic force is a question about your coordinates, not about the physics.

Some Thoughts

The thing I find most instructive here is the nature of the initial dissatisfaction. Nothing was broken. Maxwell's equations gave correct predictions in both the magnet case and the conductor case, and any working physicist could compute the answer either way. The complaint was aesthetic: the theory offered two stories where the world offered one situation.

That is a hard signal to take seriously, because "the formalism is inelegant" is usually not a reason to overhaul the foundations of physics. Most of the time it means you have not understood the formalism. Distinguishing genuine structural redundancy from your own confusion seems to me the whole difficulty, and I do not think there is a rule for it.

It is also worth noting that the paper contains no new experimental result and no new equation. Lorentz had the transformations, Poincaré had much of the interpretive apparatus, and the null ether results were eighteen years old. Einstein's contribution was deciding which of these things to stop believing. The paper's title is about the electrodynamics of moving bodies, and it opens by pointing at a place where the existing theory described the same event twice — that was, apparently, enough to go on.