Let $T: V \rightarrow V$ be a self adjoint linear map where $V$ is an inner product space.

Known fact: With respect to any orthonormal basis- the the matrix for $T$ is conjugate symmetric. When dealing with a real vectorspace just symmetric.

However is it true that if there exists some basis (not necessarily orthonormal) such that the matrix for $T$ is symmetric then $T$ is self adjoint?


1 Answer 1


No. Choose any diagonalizable operator $T \colon V \rightarrow V$ which is not self-adjoint. This means that there is a basis $\mathcal{B} = (v_1, \ldots, v_n)$ for $V$ consisting of eigenvectors of $T$ but the basis $\mathcal{B}$ is not orthonormal. Then $[T]_{\mathcal{B}}$ is diagonal and in particular, symmetric, but $T$ won't be self-adjoint. For example, you can take $V = \mathbb{R}^2$ with the standard inner-product and $T(x,y) = (2x + y, 3y)$.

In fact, $T \colon V \rightarrow V$ is diagonalizable with real eigenvalues if and only if there exists a basis $\mathcal{B}$ of $V$ such that $[T]_{\mathcal{B}}$ is Hermitian.


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.