Meta-methodology: clearing the underbrush