Phase Revisited: Interference & Physical Meaning
Track: Foundations · Difficulty: Beginner · Est: 14 min
Phase Revisited: Interference & Physical Meaning
Overview
Earlier we learned that a qubit state has complex amplitudes, and that global phase is unobservable while relative phase can matter.
This page answers the natural next question: how can phase matter if probabilities come from magnitudes like and ?
The answer is interference: when multiple amplitude contributions combine, their phases can reinforce or cancel, redistributing probability across outcomes.
We will explain this without time evolution or gates—only by comparing different measurements and how the same state looks in different bases.
Intuition
Think about adding arrows (vectors) in the plane.
- Two arrows pointing the same way add to a longer arrow.
- Two arrows pointing opposite ways can cancel.
Quantum amplitudes behave like arrows too—except they live in the complex plane.
Probabilities are squares of magnitudes, but the magnitude you square can come from a sum of amplitudes. If the summed amplitudes partially cancel, the probability can decrease; if they reinforce, it can increase.
That is the core meaning of phase: it controls how amplitudes add.
Formal Description
Probability is the square of an overlap
For a pure state and a measurement outcome state ,
Now suppose itself can be written as a superposition of computational basis states, such as
with complex coefficients and .
If
then the overlap becomes
This is a sum of complex numbers. The relative phases of and matter because they affect whether this sum is large or small.
Relative phase can change probabilities in a different basis
In the computational basis, the outcomes correspond directly to and , so probabilities depend only on and .
But in a different basis, each outcome overlap involves sums like , and phase can influence those sums.
Worked Example
Compare the two states
Same probabilities in the computational basis
For both states:
- amplitude of has magnitude
- amplitude of has magnitude
So measuring in the computational basis gives for both.
Different probabilities in the basis
Define
Notice that is exactly , and is exactly .
So:
- If the state is , measuring in the basis gives “” with probability 1.
- If the state is , the same measurement gives “” with probability 1.
This is interference in its simplest form: a relative minus sign changed nothing in one measurement basis, but made outcomes deterministic in a different basis.
Turtle Tip
If phase feels invisible, change the question. Phase often does not affect probabilities in the basis you wrote the state in, but it shows up when you measure in a different basis where outcomes depend on sums of amplitudes.
Common Pitfalls
Don’t conclude “phase doesn’t matter” just because and depend only on magnitudes in the computational basis.
Also, don’t confuse global phase with relative phase. Multiplying the whole state by changes no overlaps’ magnitudes; changing the relative phase between components can change interference.
Quick Check
- Why can a relative minus sign be invisible in one measurement basis but decisive in another?
- In , where can cancellation occur?
What’s Next
Interference is the mechanism by which quantum computations shape probabilities. Next we’ll consolidate the state picture by comparing pure and mixed states side-by-side, including how the Bloch sphere generalizes to the Bloch ball for mixed states.
