JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

The Three Principles: Open Mapping, Closed Graph, Uniform Boundedness

Completeness pays off here. These three theorems — all powered by the Baire category theorem — say that for operators between Banach spaces, surjective means open, a closed graph forces continuity, and pointwise bounds become uniform bounds.

Open mapping and closed graph

Let X and Y be Banach spaces and T: X -> Y a bounded surjective operator. The open mapping theorem says T maps open sets to open sets. The striking corollary is the bounded inverse theorem: if T is also a bijection, then T⁻¹ is automatically bounded. Invertibility of a continuous linear bijection between Banach spaces is free continuity of the inverse — something utterly false without completeness.

The closed graph theorem is its twin. The graph of T is { (x, Tx) }. For a linear operator between Banach spaces, T is bounded if and only if its graph is closed. In practice this halves the work of proving continuity: instead of estimating ||Tx||, you only check that whenever x_n -> x and Tx_n -> z, the limit forces z = Tx. You get to assume Tx_n already converges.

Uniform boundedness

The uniform boundedness principle (Banach–Steinhaus) closes the trio. Let (T_α) be a family of bounded operators from a Banach space X to a normed space Y. If for each fixed x the set { ||T_α x|| } is bounded, then the operator norms { ||T_α|| } are bounded uniformly in α. Pointwise control upgrades, for free, to uniform control. Again the proof is a Baire-category argument inside the complete space X.

Application: a pointwise limit of operators is bounded.
Let X, Y be Banach, T_n ∈ B(X, Y), and suppose
    T x := lim_n T_n x  exists for every x.
Claim: T is a bounded linear operator.

Linearity. Limits respect linear combinations:
   T(a x + b y) = lim (a T_n x + b T_n y) = a T x + b T y.

Boundedness. For each fixed x, (T_n x) converges, hence is
bounded, so sup_n ||T_n x|| < ∞. By uniform boundedness,
   M := sup_n ||T_n|| < ∞.
Then for every x:
   ||T x|| = lim_n ||T_n x|| ≤ sup_n ||T_n|| · ||x|| ≤ M ||x||.
So ||T|| ≤ M < ∞: T is bounded.   ∎

Note the leap: each T_n having finite norm is obvious, but
that the *common* bound M is finite — uniform over n — is
exactly what Banach–Steinhaus supplies.
Banach–Steinhaus in action: from pointwise convergence alone you get a uniform norm bound, hence the limit operator is bounded.

These three principles, together with Hahn–Banach from guide 4, are the classical pillars of linear functional analysis. Notice the recurring lesson of the whole track: completeness is what converts a soft pointwise or algebraic hypothesis into a hard quantitative bound. That conversion — from existence to estimate — is the deep reason Banach spaces are the right setting.