Decomposition in early stages of learning novel morphologically derived words: The impact of linear vs. non-linear structure

Upasana Nathaniel, Stav Eidelsztein, Kate Girsh Geskin, Brianna L. Yamasaki, Bracha Nir, Vedran Dronjic, James R. Booth, Tali Bitan

Research output: Contribution to journalArticlepeer-review


We examined whether morphological decomposition takes place in early stages of learning a novel language, and whether morphological structure (linear vs. non-linear) influences decomposition. Across four sessions, 41 native-Hebrew speakers learned morphologically derived words in a novel morpho-lexicon, with two complex conditions: linear and non-linear; and a third simple condition with monomorphemic words. Participants showed faster learning of trained words in the linear condition, and better generalization to untrained words for both complex conditions compared to the simple condition, with better performance for linear than non-linear morphology. Learning the root morpheme, which provides a concrete meaning, was better than learning template/suffix morphemes, which are more abstract. Overall, our results suggest that saliency of discrete units plays an important role in decomposition in early stages of learning derived words, even for speakers highly familiar with the non-linear structure in their L1.

Original languageEnglish
Article number105604
StatePublished - Nov 2023

Bibliographical note

Publisher Copyright:
© 2023 Elsevier B.V.


  • Artificial language
  • Decomposition
  • Derivational
  • Learning
  • Morphology
  • Second language acquisition

ASJC Scopus subject areas

  • Experimental and Cognitive Psychology
  • Language and Linguistics
  • Developmental and Educational Psychology
  • Linguistics and Language
  • Cognitive Neuroscience


Dive into the research topics of 'Decomposition in early stages of learning novel morphologically derived words: The impact of linear vs. non-linear structure'. Together they form a unique fingerprint.

Cite this