Reading Proof Sketches

A proof sketch compresses an argument while preserving the mechanism that connects its hypotheses to its conclusion. Reading one well means locating the steps whose validity depends on domains, topology, or a separate theorem. An omitted algebraic expansion and an omitted compactness argument are different kinds of omission: the second can carry most of the mathematical content.

Helpful background. Functional Analysis Essentials supplies weak convergence and compactness. Quadratic Forms explains the energy-form language used in the worked example.

Separate the conclusion from the mechanism

Before filling in details, restate the conclusion with its quantifiers. Does the argument prove existence of a state, uniqueness of that state, or convergence of a specified approximation? Is convergence for each fixed initial vector or uniformly over a class? Does a spectral value exist, or does a normalizable eigenvector exist?

Then identify the mechanism. A positivity argument orders quantities. A compactness argument produces convergent subsequences. A uniqueness theorem identifies possible limits. A norm estimate controls an approximation error. None can automatically replace the others.

For example, a variational calculation may approach the bottom of a spectrum arbitrarily closely without producing any ground-state vector. To infer existence, the proof must explain what prevents a minimizing sequence from escaping or concentrating in a way that destroys normalization in its limit.

Read imported results as mathematical inputs

When a sketch says “by the spectral theorem,” determine which part is used: a projection-valued measure, functional calculus, a domain characterization, or a spectral estimate. The theorem’s name is not a substitute for matching its hypotheses to the current operator.

A useful working note has three clauses: input, operation, and output. For instance: a self-adjoint $H$ is the input; bounded spectral calculus is the operation; $e^{-itH/\hbar}$ is a unitary operator on the entire Hilbert space. Differentiability of its orbit is a further conclusion requiring domain information; it does not follow just from writing an exponential.

The Stone theorem treatment supplies that precise result. The reader’s task is to recognize whether the sketch needs continuity, differentiability, or both, and to import the corresponding statement.

A plausible variational sketch with a missing step

Consider this proposed reasoning: choose normalized vectors with energies tending to the infimum, pass to a weakly convergent subsequence, and use lower semicontinuity to obtain a normalized minimizer.

The normalization claim is unsupported. Let $H$ be multiplication by $x$ on $L^2(0,1)$ and set

$$\psi_n(x)=\sqrt n\,\mathbf1_{(0,1/n)}(x).$$

These vectors satisfy

$$\|\psi_n\|=1, \qquad \langle\psi_n,H\psi_n\rangle=\frac1{2n}\longrightarrow0.$$

For every $f\in L^2(0,1)$, Cauchy–Schwarz gives

$$|\langle f,\psi_n\rangle| \leq\left(\int_0^{1/n}|f(x)|^2dx\right)^{1/2} \longrightarrow0.$$

Thus $\psi_n$ converges weakly to zero. The norm has been lost in the weak limit. Since $x>0$ almost everywhere, $H\psi=0$ implies $\psi=0$; there is no normalized zero-energy eigenvector. Lower semicontinuity gives a valid statement about the zero vector, but that vector is not an admissible normalized trial state.

This example identifies the exact missing implication. The problem is not that variational reasoning is unreliable; the problem is that weak convergence alone does not preserve the constraint.

A compactness hypothesis repairs the argument

Suppose instead that a densely defined closed form $q$ is bounded below and its form domain embeds compactly into the Hilbert space. Shift $q$ by $c\|\psi\|^2$ so that

$$\|\psi\|_q^2=q[\psi]+c\|\psi\|^2$$

is a positive form norm equivalent to the usual closed-form norm. A normalized minimizing sequence is bounded in this norm. A subsequence converges weakly in the form Hilbert space, and compactness gives strong convergence in the original Hilbert space to the same limit.

Strong convergence now preserves normalization. Weak lower semicontinuity of the shifted form norm, together with convergence of the ordinary norm, shows that the limit attains the infimum. The form representation theorem and the variational equation identify it as a ground-state eigenvector of the associated operator.

This is still a sketch: it imports closed-form representation and compact-embedding results. But it says exactly where compactness is used and which conclusion it supplies. Checking those two imported results is a finite, well-defined reading task.

Match confidence to what has actually been justified

Label a step in your notes as an explicit calculation, an application of a stated theorem, or an unresolved requirement. “Take the limit” should specify the topology and the operation that passes to the limit. “Choose an eigenbasis” should identify the assumption guaranteeing one. “Integrate by parts” should identify the function space and boundary contribution.

A failed proof step does not by itself disprove the theorem; another argument may establish it. Conversely, a correct calculation in one example does not establish the quantified theorem. Keeping those distinctions explicit helps a reader criticize an argument without discarding useful intuition.

Exercise: what does lower semicontinuity preserve?

A sequence satisfies $\|\psi_n\|=1$ and $\psi_n\rightharpoonup\psi$. What can be concluded about $\|\psi\|$? What additional convergence would recover normalization?

Solution

Weak lower semicontinuity gives $\|\psi\|\leq\liminf_n\|\psi_n\|=1$, not equality. Norm convergence gives $|\|\psi_n\|-\|\psi\||\leq\|\psi_n-\psi\|\to0$, hence $\|\psi\|=1$. A compactness theorem is one way to obtain that stronger convergence along a subsequence; it must apply to the actual bounded class containing the sequence.

References

• B. C. Hall, Quantum Theory for Mathematicians, Graduate Texts in Mathematics 267, Springer, 2013 — the relationship between operator statements and their proof hypotheses.
• G. Teschl, Mathematical Methods in Quantum Mechanics: With Applications to Schrödinger Operators, 2nd ed., Graduate Studies in Mathematics 157, American Mathematical Society, 2014, Chapters 2 and 4 — closed forms, spectral minima, and variational arguments.