Intermediate level of analysis: features or segments?

In this commentary, we have focused on the phonetic representation used in perception en route to the phonological representation in the lexicon. Under the model that we envisage, the acoustic–phonetic analysis by the perceptual system extracts information from the signal which is then translated into phonological features. Thus, no intermediate segmental analysis is made prior to access of the phonological representation. Instead, it is argued that featural information is directly mapped onto the lexicon.