This talk revisits some very old questions raised by the morphosyntactic marking of `short' and `long' WH-movement in Irish. It develops a new analysis which claims to integrate some old and famously recalcitrant data, at the same time as it simplifies (in some sense) the treatment of the most basic cases.