In a proof of the Gelfand-Mazur theorem, I read that $$ \lim_{\lambda\to\infty}\|(a-\lambda 1)^{-1}\|=0 $$ (where it is assumed for the sake of contradiction that the inverses in the norm exist for all $\lambda\in\mathbb{C}$)

How can this be proven?


For $\lvert\lambda\rvert > \lVert a\rVert$, we can write

$$(a-\lambda 1)^{-1} = -\lambda^{-1} (1 - \lambda^{-1}a)^{-1} = -\lambda^{-1}\sum_{\nu = 0}^\infty \lambda^{-\nu} a^\nu,$$

and the triangle inequality together with the submultiplicativity of the norm yields

$$\lVert (a-\lambda 1)^{-1}\rVert \leqslant \frac{1}{\lvert\lambda\rvert} \sum_{\nu=0}^\infty \biggl(\frac{\lVert a\rVert}{\lvert\lambda\rvert}\biggr)^\nu = \frac{1}{\lvert \lambda\rvert - \lVert a\rVert}.$$

This part is true in any Banach algebra, for $\lvert\lambda\rvert > \lVert a\rVert$, the inverse of $a-\lambda 1$ exists, and its norm is bounded by $\dfrac{1}{\lvert\lambda\rvert - \lVert a\rVert}$.

And thus, if an $a$ had empty spectrum, that would give rise to a non-constant bounded entire holomorphic function.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.