Suspended affixation in Turkish

Abstract

This study investigates Suspended Affixation (SA) and its environment. Lewis (1967) is credited with the term, but observations of SA can be found in earlier Turkish grammar books (Emre, 1945; Gencan, 1966). It is a morphological ellipsis process observed in constructions that contain conjuncts. Overt suffixes on the rightmost conjunct in a conjunction are interpreted in the other conjuncts. I provide both empirical and theoretical observations that paint a broader picture for SA. I argue for an ellipsis analysis in SA and propose analyses for the suffixes ile/=(y)lA and -(y)Ip. My findings indicate that the environment of SA greatly impacts its acceptability. The conjoiner veya ‘or’ has pragmatic implicatures that can hinder SA. On the other hand SA in -(y)Ip constructions might require changes in the information structure. In the literature on Turkish SA, Kabak (2007) provides constraints and Orgun (1995), Kornfilt (2012), Broadwell (2008) provide a lexical sharing analysis. The literature on other languages adopts an ellipsis approach (Guseva & Weisser, 2018; Erschler, 2018). My analysis is more in line with the ellipsis approach, yet it abides by the proposed constraints in the literature. I conducted three experiments. The first (214 participants) investigated if SA of derivational suffixes was acceptable and the second (160 participants) investigated if different amounts of SA changed processing difficulty. Both experiments investigated the effect of a conjoiner choice between ve ‘and’ and veya ‘or’. The third experiment showed that SA can be a testing ground for sentence processing in Turkish. The experiment used an environment dependent on SA and showed effects of Reanalysis and how reanalyzed readings were accessible in further tasks. All the experiment results jointly indicated that the SA environment was more crucial than solely identifying suspendable affixes.