# Connections between classical mechanics and quantum mechanics [duplicate]

I've been studying quantum mechanics and classical mechanics for a little while now, and I still don't feel as though I fully understand the motivation for some of our choices in Heisenberg mechanics. For example, it clearly isn't a coincidence that the classical observables (functions of coordinates and their conjugate momenta) and the quantum observables (Hermitian operators) seem to form analogous Lie algebras with the Poisson bracket and commutator respectively. But it isn't clear to me why this is true. Is there some deep meaning contained in this statement? Or is it more indicative of the fact that in constructing a quantum model of the universe we took substantial inspiration from our intuition and previous study of classical mechanics?

Along these same lines, what motivates the move from classical functions on phase space to Hermitian operators? I understand why operators corresponding to observables must be self-adjoint (the eigenvalues must be real), but I don't understand what motivates the move to operators in general. Why would we expect that operators on a Hilbert space would give physical predictions? Part of my confusion here may also come from the fact that it isn't entirely clear to me what exactly these operators do in all cases. For example, I get that $\langle \psi | \hat{x} | \psi \rangle$ corresponds to the expected position of a particle in state $|\psi\rangle$, but it's much less obvious what the $\hat{x}$ operator does to a state in general. In some cases (such as $J_\pm$ when considering angular momentum), it's clear what the operator does to a state (raises or lowers eigenstates of $J_z$), but in all these cases the operator is non-Hermitian. Perhaps the answer to this question is simply that the model gives accurate predictions and so we use it, but I'm wondering if there's a better way to think about these things.

## marked as duplicate by ACuriousMind♦, BMS, joshphysics, Rob Jeffries, JamalSDec 27 '14 at 18:50

• A mathematician called Van der Waerden was a friend of Heisenberg and a historian of science. I am not in a position to write an answer but his works give a first hand account of development and motivation of quantum mechanics – Cheeku Dec 27 '14 at 2:20
• Did you look at Schroedinger's papers? They are not easy to read, but at least "Quantisierung als Eignenwertproblem" is pretty straight forward. In that paper he proposes the Schroedinger equation. The main motivation seems to be that the eigenvalue problem replaces an explicit quantization condition. Given the confusion at the time it's likely that the founders of quantum mechanics drew motivation from many different kinds of arguments. – CuriousOne Dec 27 '14 at 3:16
• If you want some non commutative features (that are nothing else but counterfactuals, as we see it in the Young' double split experiment), you do need to see observables as operators. – sure Dec 27 '14 at 11:30
• Why didn't you answer the question? It is a beautiful question. Indeed, as you say, I learnt the problem of finding the eigenvalues and eigenstates of a matrix, as a way of treatment of discrete values for physical quantities. – Sofia Dec 27 '14 at 11:33
• related: physics.stackexchange.com/q/46015 and links therein – glS Dec 27 '14 at 12:09

"Is there some deep meaning contained in this statement? Or is it more indicative of the fact that in constructing a quantum model of the universe we took substantial inspiration from our intuition and previous study of classical mechanics?"

Both variants are true.

When developing the special relativity, Einstein didn't begin from zero, but from the classical mechanics supplemented with the fact that light velocity is the same in any inertial frame. We didn't invent the concepts of space, time, velocity, etc. They were in use already.

With the QM, we didn't invent position, linear momentum, angular momentum, etc. We just came to the conclusion that some objects behave like waves, and for their description we need an equation describing the evolution of a wave (in fact the Schrodinger eq. is more resembling the heat equation, just the diffusion coefficient is imaginary - these things were discusses in connection with other questions). Also, we concluded that some physical properties don't admit continuous values, but discrete values.

So, we developed the QM by introducing modifications in the classical mechanics - the Thomson atom, the Bohr model of atom, the to-day model based on the Schrodinger equation.

Heisenberg developed first a matrix calculus. Why matrixes? First of all because some physical quantities have discrete values that can be obtained from a problem with discrete eigenfunctions and eigenvalues (you should have learnt of calculus of eigenvalues and eigenfunctions of matrices). So, a quantity, e.g. projection of angular momentum, can have some value $q$, when the system is in a certain situation, (state). By the way, when Schrodinger developed his equation, he has had adhered to the idea of statistical nature of QM. Thus, a quantum state allows for some observable a precise value, but for other observables (which may be position) only statistical predictions.