Automaticity: A new framework for dyslexia research?