Dirac's Gamble

Dirac wanted the impossible-sounding thing that Chapter 1 pointed at: an equation gentle enough on time to keep probabilities positive, yet even-handed enough between time and space to satisfy relativity, that still reproduces relativity’s energy relation when you square it. The obstacle is arithmetic itself. Ordinary numbers commute – $3\times5$ equals $5\times3$ – and with numbers that obedient the equation simply cannot be built. Dirac’s gamble was to reach for a stranger kind of number, one where order matters, and the smallest such numbers that do the job turn out to be matrices: little grids of numbers. Accepting them forces the wavefunction to carry four components instead of one, and Dirac had no idea, at first, what the extra three were for.

Here is the trick in slow motion. Instead of an energy that is a square root, write it as a straight sum: some coefficient times the momentum, plus another coefficient times the mass. Now square that sum and insist you get back relativity’s energy relation. Multiply it out and, alongside the terms you want, unwanted cross-terms appear – products of the two coefficients. The only way for those cross-terms to cancel is if swapping the order of the coefficients flips their sign. No ordinary number does that; $a\times b$ is always $b\times a$. Dirac needed a number system where order is not sacred.

If the coefficients must be matrices, how big do the matrices have to be? Build it up dimension by dimension. In a world with only one direction of space, the smallest order-flipping objects that work are the $2\times2$ Pauli matrices – the very ones the quantum course introduced for spin. Two directions of space, still $2\times2$. But our three directions, plus mass, need four different objects that all mutually flip sign, and there are only three $2\times2$ ones to be had. You are forced up to $4\times4$ matrices – and the wavefunction they act on is dragged up with them, from one component to four. The three-ness of space forces the four-ness of the electron.

With the matrices found, the equation exists – and the very first thing it hands back is the cure Dirac was after. Its probability is built from all four components, added as squares, so it can never be negative. The dragon from Chapter 1 that turned probabilities negative is dead the instant the equation is written down. That alone would have justified the strange four-component wavefunction; that it also, later, explains spin and predicts antimatter is the windfall.

Tally the price and the prize. Dirac got his first-order, relativistic equation with honest probabilities – the prize. The price was a wavefunction that quietly quadrupled to four components with no stated meaning, and negative energies that refused to leave. The next chapter collects on the first mystery: slow the equation down and two of those four components reveal themselves as the electron’s spin – the property the quantum course had to assume – with its odd double strength, $g = 2$, thrown in for free.