Abstract
Two groups of 35 paid graduate students were presented nonsense words grouped as (a) orthographically and phonologically illegal and unpronounceable; (b) orthographically and pronologically illegal and pronounceable; and (c) orthographically and phonologically legal. The first group was directed to rate each item for pronounceability. The second group was to guess what each word was before it had been slightly distorted. The results of the first group's ratings appear to indicate that there may be alternatives to a phonemic recoding explanation. The second study indicates that a graphemic encoding explanation may be more appropriate. This study concludes that the coding processes employed may be contingent upon the tasks required in varying situations.