## Richard Feynman on Erwin Schrödinger

I thought it is interesting to see what the great Nobel Laureate physicist Richard Feynman said about Erwin Schrödinger’s attempts to discover the famous Schrödinger equation in quantum mechanics (see quote below). It has been my experience in reading physics that this sort of “heuristic” reasoning is part of doing physics. It is a very creative (sometimes not logical!) art with mathematics in attempting to understand the physical world. Dirac did it too when he obtained his Dirac equation for the electron by pretending he could take the square root of the Klein-Gordon operator (which is second order in time). Creativity is a very big part of physics.

“When Schrödinger first wrote it down [his equation], he gave a kind of derivation based on some heuristic arguments and some brilliant intuitive guesses. Some of the arguments he used were even false, but that does not matter; the only important thing is that the ultimate equation gives a correct description of nature.” — Richard P. Feynman (Source: The Feynman Lectures on Physics, Vol. III, Chapter 16, 1965.)

## Principles of quantum theory

One beautiful summer morning I spent a couple hours in a park reflecting on what I know about quantum mechanics and thought to sketch it out from memory. (A good brain exercise to recapture things you learned and admire.) This note is an edited summary of my handwritten draft (without too much math).

Being a big subject, I will stick to some basic ideas (or principles) of quantum theory that may be worth noting.

Two key concepts are that of a `state’ and that of an ‘observable’.

The former describes the state of the system under study. The observable is a thing we measure. So for example, an electron can be in the ground state of an atom – which means that it is in `orbital’ of lowest energy. Then we have other states that it can be in at higher energies.

The observable is a quantity distinct from a state and one that we measure. Such as for example measuring a particle’s energy, its mass, position, momentum, velocity, charge, spin, angular momentum, etc.

QM gives us principles / interpretations by how states and observables can be mathematically described and how they relate to one another. So here is the first principle.

**Principle 1. The state of a system is described by a function (or vector) ψ. The probability density associated with it is given by |ψ|².**

This vector is usually a mathematical function of space, time (sometime momentum) variables.

For example, f(x) = exp(-x^2) is one such example. You can also have wave examples such as g(x) = exp(-x^2) sin(x) which looks like a localized wave (a packet) that captures both being a particle (localized) and a wave (due to the wave nature of sin(x)). This wave nature of the function allows it to interfere constructively or destructively with other similar functions — so you can have interference! In actual QM these wavefunctions involve more variables that one x variable, but I used one variable to illustrate.

**Principle 2. Each measurable quantity (called an ‘observable’) in an experiment is represented by a matrix A. (A Hermitian matrix or operator.)**

For example, energy is represented by the Hamiltonian matrix H, which gives the energy of a system under study. The system could be the hydrogen atom. In many or most situations, the Hamiltonian is the sum of the kinetic energy plus the potential energy (H = K.E. + V).

For simplicity, I will treat a measurable quantity and its associated matrix on equal footing.

From matrix algebra, a matrix is a rectangular array of numbers – like, say, a square array of 3 by 3 numbers, like this one I grabbed from the net:

Turns out you can multiply such things and do some algebra with them.

Two basic facts about these matrices is:

(1) they generally do not have the commutative property (so AB and BA aren’t always equal), unlike real or complex numbers,

(2) each matrix A comes with `magic’ numbers associated to it called eigenvalues of A.

For example the matrix

(called diagonal because it has all zeros above and below the diagonal) has eigenvalues 1, 4, -3. (When a matrix is not diagonal we have a procedure for finding them. Study Matrix or Linear Algebra!)

**Principle 3. The possible measurements of a quantity will be its eigenvalues. **

For example, the possible energy levels of an electron in the hydrogen atom are eigenvalues of the associated Hamiltonian matrix!

**Principle 4. When you measure a quantity A when the system is in the state ψ the system `collapses’ into an eigenstate f of the matrix A.**

Therefore the system makes a transition from state **ψ** to state **f **(when A is measured).

So mathematically we write

**Af = af**

which means that **f** is an eignstate (or eigenvector) of **A** with eigenvalue **a**.

So if **A** represents energy then `**a’** would be energy measurement when the system is in state **f**.

For a general state **ψ** we cannot say that **A ψ = aψ. **This eigenvalue equation is only true for eigenstates, not general states.

**Principle 5. Each state ψ of the system can be expressed as a superposition sum of the eigenstates of the measurable quantity (or matrix) A.**

So if ** f, g, h, … **are the eigenstates of

**A**, then any other state

**ψ**of the system can be expressed as a superposition (or linear combination) of them:

**ψ = bf + cg + dh + …
**

where b, c, d, … are (complex) numbers. Further, **|c|^2** = probability **ψ** will `collapse’ into the eigenstate **g** when measurement of **A** is performed.

These principles illustrate the *indeterministic* nature of quantum theory, because when measurement of **A** is made, the system can collapse into any one of its many eigenstates (of the matrix **A**) with various probabilities. So even if you had the ‘exact same’ setup initially there is no guarantee that you would see your system state change into the same state each time. That’s non-causality! (Quite unlike Newtonian mechanics.)

**Principle 6. (Follow-up to Principles 4 and 5.) When measurement of A in the state ψ is performed, the probability that the system will collapse into the eigenstate vector φ is the dot product of Aψ and φ.**

The latter dot product is usually written using the Dirac notation as **<φ|A|ψ>**. In the notation above, this would be same as **|c|^2**.

Next to the basic eigenvalues of A, there’s also it’s `average’ value or expectation value in a given state. That’s like taking the weighted average of tests in a class – with weights assigned to each eigenstate based on the superposition (as in the weights b, c, d, … in the above superposition for ψ). So we have:

**Principle 7. The expected or average value of quantity A in the state described by ψ is <ψ|A|ψ>.**

In our notation above where **ψ = bf + cg + dh + …,** this expected value is

**<ψ|A|ψ>** = |b|^2 times ** (eigenvalue of f)** + |c|^2 times (eigenvalue of g) + …

which you can see it being the weighed average of the possible outcomes of A, namely from the eigenvalues, each being weighted according to its corresponding probabilities **|b|^2, |c|^2, … . **

In other words if you carry out measurements of A many many times and calculate the average of the values you get, you get this value.

**Principle 8. There are some key complementary measurement observables. (Classic example: Heisenberg relation QP – PQ = ih.)**

This means that if you have two quantities P and Q that you could measure, if you measure P first and then Q, you will not get the same result as when you do Q first and then P. (In Newton’s mechanics, you could at least in theory measure P and Q simultaneously to any precision in any order.)

Example, position and momentum are complementary in this respect — which is what leads to the Heisenberg Uncertainty Principle, that you cannot measure both the position and momentum of a subatomic particle with arbitrary precision. I.e., there will be an inherent uncertainty in the measurements. Trying to be very accurate with one means you lose accuracy with they other all the more.

From Principle 8 you can show that if you have an eigenstate of the position observable Q, it will not be an eigenstate for P but will be a superposition for P.

So `collapsed’ states could still be superpositions! (Specifically, a collapsed state for Q will be a superposition, uncollapsed, state for P.)

That’s enough for now. There are of course other principles (and some of the above are interlinked), like the Schrodinger equation or the Dirac equation, which tell us what law the state ψ must obey, but I shall leave them out. The above should give an idea of the fundamental principles on which the theory is based.

Have fun,

Samuel Prime