The Postulates of Quantum Mechanics
Quantum mechanics is not derived from classical mechanics; it rests on its own postulates, which state what a physical state is, which quantities can be measured, what the possible outcomes of a measurement are and with what probabilities they occur, how a measurement alters the system, and how composite systems are described. This chapter collects the conventional postulates in the state-vector formulation, together with the canonical quantization rules and the notion of uncertainty on which Heisenberg's principle is built. The mathematical arena — Hilbert spaces and the operators acting on them — is developed in Hilbert Spaces. Standard modern treatments of this material are found in [Sakurai:2017] [CohenTannoudji:1977] [Griffiths:2018]. The source for this chapter is an early outline: its completed content is ported in full below, and the headings it reserves without content are marked as pending rather than filled in with material the source does not contain.
The conventional postulates
The state postulate
Let \(\mathcal{H}\) be a Hilbert space over \(\C\) and let \(\ket{\psi}\in\mathcal{H}\). The relation
is an equivalence relation. We say that an equivalence class is a ray in \(\mathcal{H}\).
A quantum physical state is represented by a ray of normalized vectors, that is, vectors satisfying
If no ambiguity is possible, we shall say that the state of a system is one of its representatives with vanishing global phase; that is, the state will be \(\ket{\psi}\).
The observable postulate
Let \(O\in\mathcal{L}(\mathcal{H})\) be a linear operator on \(\mathcal{H}\).
To every measurable quantity there corresponds a Hermitian operator \(O\).
We shall call such an operator an observable. It must be noted that not every Hermitian operator is an observable.
The measurement postulate
Consider the spectral decomposition of an observable \(O\),
The possible results of a measurement of \(O\) are its eigenvalues \(\lambda_i\).
The probability postulate: the Born rule
Consider a quantum system in the state \(\ket{\psi}\), and let \(O\) be an observable with eigenvalues \(\lambda_i\).
The probability of obtaining \(\lambda_i\) in a measurement of \(O\) is given by
We must note that the probability is determined on the whole ray of \(\ket{\psi}\), since \(\abs{\ee^{\ii\phi}}=1\). Moreover, by the theory of probability, it must hold that
Expanding \(\ket{\psi}\) in the eigenbasis \(\set{\ket{\lambda_i}}_{i=1}^{\dim\mathcal{H}}\) as \(\ket{\psi}=\sum_{j}a_j\ket{\lambda_j}\), we have
so the probability of obtaining \(\lambda_i\) in a measurement is the squared modulus of the \(i\)-th component of \(\ket{\psi}\) in the eigenbasis.
The collapse postulate
Consider a quantum system in the state \(\ket{\psi}\), and let \(O\) be an observable with eigenvalues \(\lambda_i\).
If a measurement of \(O\) yields the eigenvalue \(\lambda_i\), then immediately after the measurement the state of the system is \(\ket{\lambda_i}\).
The time-evolution postulate
The source records for this postulate only the stationary eigenvalue problem of the Hamiltonian,
[Heading reserved in the source outline; the statement of the time-dependent Schrödinger equation governing the evolution of quantum states is pending port from a completed source.]
The composite-system postulate
Consider a quantum system composed of \(N\) individual systems whose states are rays in the Hilbert spaces \(\mathcal{H}_{1},\ldots,\mathcal{H}_{N}\).
The state of the composite system is a ray in the Hilbert space
Density-matrix postulates
[Heading reserved in the source outline; the reformulation of the postulates in terms of density matrices, covering statistical mixtures as well as pure states, is pending port from a completed source.]
On Planck's constant
[Heading reserved in the source outline; the discussion of Planck's constant, its role as the quantum of action, and its measured value is pending port from a completed source.]
Pictures of time evolution
The Schrödinger picture
[Heading reserved in the source outline; no content to port.]
The Heisenberg picture
[Heading reserved in the source outline; no content to port.]
Canonical quantization rules
From the Poisson bracket to the commutator
[Heading reserved in the source outline, together with its subheadings on conjugate coordinates and conserved quantities; the rule mapping classical Poisson brackets to quantum commutators, and with it the canonical commutation relations between conjugate coordinates, is pending port from a completed source.]
From Poisson brackets to commutators: statement of the canonical quantization rule and derivation of the canonical commutation relations between conjugate coordinates, which are used throughout the following chapters.
Linear momentum as a differential operator
In the coordinate representation, the components of the linear momentum act on wave functions as the differential operators
Justification of the coordinate-representation rule \(p_i=-\ii\hbar\,\pp_i\) from the canonical commutation relations; the source asserts the rule without derivation.
The Heisenberg uncertainty principle
Mean value of an observable
Consider an observable \(A\). We define the mean (or expected) value of the operator in the state \(\ket{\psi}\) as
Uncertainty
In the same context as above, we define the uncertainty of the observable \(A\) in the state \(\ket{\psi}\) as
Let us observe the following. We have
where we used the normalization Equation (77.1). Hence the uncertainty of \(A\) is given by
The generalized uncertainty principle
[Heading reserved in the source outline.]
Statement and derivation of the generalized uncertainty relation bounding the product of the uncertainties of two non-commuting observables by the mean value of their commutator.