Semantic transparency and doublet formation: the case of Hebrew location nouns

Research output: Contribution to journalArticlepeer-review


This study examines the correlation between derivational paradigms and morphological variation and change. I will examine a case study of Hebrew location nouns formation. Semitic morphology relies highly on non-concatenative morphology, where words are formed in patterns. Some Hebrew location nouns that are formed in one pattern, receive an additional form in another pattern with no change of their meaning. In contrast, there are location nouns, which are also formed in the same pattern, but do not have morphological doublets. Previous studies accounted for this change and proposed phonological and semantic criteria that trigger it. However, such explanation only account for why the change occurs, but not for cases where there is no doublet formation. I argue that morphological change is highly motivated in cases where the forms that undergo a change are part of a derivational paradigm. Specifically, I will show that only location nouns that are derivationally related to a verbal counterpart, such that the semantic relation between them is highly transparent, can undergo such change and have doublets. In contrast, words that are not part of such a paradigm are less likely to undergo change. The study highlights the important role of semantic transparency and derivational paradigms in morphological variation and change, showing that properties of words are not the only criteria that are taken into consideration, but also their relations with other words within a derivational paradigm.

Original languageEnglish
StateAccepted/In press - 2024

Bibliographical note

Publisher Copyright:
© The Author(s) 2024.


  • Derivation
  • Derivational paradigms
  • Doublets
  • Location nouns
  • Semantic transparency
  • Variation


Dive into the research topics of 'Semantic transparency and doublet formation: the case of Hebrew location nouns'. Together they form a unique fingerprint.

Cite this