Phonological fusion in dichotic monitoring.

Phonological fusion (e.g., BACK + LACK perceived as BLACK) was examined in a dichotic monitoring task that required subjects to press a button whenever they thought a predesignated target (e.g., BLACK) occurred. Fusion rate was determined from incidental response to fusible pairs in this task. It was found that a voluntary strategy of selective attention to one ear reduced fusion rate. However, fusions could not be entirely eliminated using this strategy. A signal detection analysis indicated that strategy did not involve a response bias shift, although individual differences in fusion rates unimodally distributed) could be partly accounted for by differences in response bias and discriminatory ability. Phenomenological data suggested that a fusible stimulus pair may yield more than one percept on a single presentation, but not necessarily that same percept on repeated presentations. Reaction time data supported the view that phonological fusion involves parallel processing of fusible items in the language-dominant hemisphere.