Gender identity, interactionism, and politics: A reply to Rogers and Walsh