My point exactly! Moreover, such a metonymic shift (from the state of the hair itself to the individual's psychological state) may well account for the variable constituent grouping (i.e., "bad hair" day vs. bad "hair day"). PAT