Word storage and computation
Antonio Fábregas, Martina Penke · 2020
The goal of this chapter is to assess some representative morphological theories with respect to their compatibility with results coming from psycholinguistic experiments. We will concentrate mainly on the question of whether inflected words are computationally built by the addition of discrete units, called ‘morphemes’, or have to be treated as undecomposable wholes which relate to each other through connections with other words. The issue is complicated by two problems: the nature of morphemes as abstract units that anchor phonological and semantic information or symbolic Saussurean units, and the problem of whether inflected forms are decomposable at some level of analysis or must be stored as fully built forms.