Proceedings of the Annual Meeting of the Cognitive Science Society

Whether morphologically complex words are stored as a whole or decomposed into constituents has been well-investigatedexperimentally in Indo-European languages like English, Italian, Dutch and French. There is substantial evidence in theselanguages in favor of and architecture which allows both decomposition and storage. This study investigates how mor-phologically complex words involving two or more morphemes are represented in Turkish which, unlike Indo-Europeanlanguages, is renowned for its highly rich morphology. Applying a probabilistic tradeoff-based model of morphologicalstorage and computation (ODonnell 2015) to a corpus of Turkish word forms, we derive predictions about stored patternsin the language. We discuss these patterns and propose several for future experimental investigation.