The kaon result asked for a place in the theory
The 1964 discovery of CP violation showed that a weak process could distinguish matter from its mirror-transformed antimatter counterpart. By 1973, Makoto Kobayashi and Toshihide Maskawa asked how that difference could fit inside the emerging renormalizable weak-interaction theory. A simple extension from two quark generations to three could retain a complex phase with observable consequences. It opened a route to the asymmetry before all six quarks had been discovered.
CP combines the exchange of particles with antiparticles and a reversal of spatial directions. If a theory respects CP, the corresponding transformed processes have matching probabilities. The kaon observation established a failure of that match in nature. The next problem was structural: where could the difference live among the theory’s interactions, without adding an arbitrary separate rule for every decay?
Weak decays do not pair quark flavours one by one
Quarks come in flavours, the name for species such as up, down and strange. They are arranged in generations with one up-type and one down-type quark in each. A charged weak interaction can change an up-type quark into a down-type quark, but its preferred partner is a quantum mixture of the down-type flavours. A mixing matrix records the amplitudes for those alternatives.
An amplitude is the quantum quantity whose combination determines probabilities. When several routes lead to the same final state, amplitudes can interfere. Their relative phases determine whether they reinforce or cancel. A phase can therefore affect a decay rate, but some phases are merely choices in how the quark fields are named mathematically. Those conventional choices must be removed before claiming a new physical effect.
With two generations, the available redefinitions remove the complex phases from the quark-mixing matrix, leaving one real mixing angle. In that minimal setting, the matrix supplies no CP-violating phase. Adding a third generation changes the counting: after the removable choices are used, one physically significant phase can remain.
Sources: [2]
A surviving phase can change particle and antiparticle patterns
A physical phase does not guarantee that every decay has a visible asymmetry. The process must contain the required competing amplitudes, and their sizes and other phases matter. Quark masses and mixing values also matter; special choices can remove CP violation even in a three-generation theory. Kobayashi and Maskawa provided the structural room for the effect, not a universal asymmetry in every reaction.
In neutral-meson experiments, a particle can reach a final state directly or after first mixing into its antiparticle. The two routes can interfere. Comparing time-dependent decay patterns for differently tagged initial states then exposes combinations of the mixing parameters. This approach turns an otherwise abstract phase into a difference between measured distributions.
The description acquired the name Cabibbo–Kobayashi–Maskawa, or CKM, mixing. Cabibbo’s earlier quark-mixing angle forms part of the historical foundation. The third generation gave the enlarged matrix enough structure to carry a CP-violating phase while keeping the weak interaction organized within the same framework.
A phase matters through a comparison of routes. Measuring one isolated amplitude’s overall phase gives no observable result, just as changing a common clock zero leaves time intervals unchanged. Relative phases between contributions to the same outcome can alter an interference pattern. Field redefinitions remove the arbitrary shared conventions; the CKM phase that survives them can affect those relative comparisons. This is why the counting of removable phases has physical consequences.
B factories tested the phase in a different particle family
In 2001, the BABAR experiment at SLAC and Belle at KEK reported CP violation in neutral B meson decays. Their asymmetric electron–positron colliders produced B pairs. One decay helped tag the other meson’s flavour, while the separation of the decay points supplied information about the time difference. The measured time asymmetry tested the CKM explanation beyond the kaon system.
The two collaborations used independent detectors and samples. Their early central values were different, with substantial statistical uncertainties, but both established the effect and supported the emerging three-generation picture. Subsequent measurements developed a much more detailed test of the framework.
Kobayashi and Maskawa shared half of the 2008 Nobel Prize for the origin of broken symmetry that predicts at least three quark families. The essential idea remains a careful distinction between a phase we can rename away and one that changes a measurable interference pattern. A third generation made the second kind possible.