Word Recognition: Do We Need Phonological Representations?

Under what format(s) are spoken words memorized by the brain? Are word forms stored as abstract phonological representations? Or rather, are they stored as detailed acoustic-phonetic representations? (For example as a set of acoustic exemplars associated with each word). We present a series of experiments whose results point to the existence of prelexical phonological processes in word recognition and suggest that spoken words are accessed using a phonological code.