Croatian (mor)phonotactic word-medial consonant clusters in the early lexicon

Abstract This study investigates the emergence of (mor)phonological consonant clusters in L1 acquisition. Following the Strong Morphonotactic Hypothesis (SMH), distribution and preferability of word-medial consonant clusters in the corpus of three children acquiring Croatian were explored. VCCV and VCCCV clusters were extracted from the Croatian Corpus of Child Language. Subsequently, all word-medial clusters were tagged as exclusively phonotactic, exclusively morphonotactic, and ones that occurred in both contexts. The results partially corroborated SMH, strengthening the premise that morphological richness of a language moderates the interaction between morphology and phonology in early language acquisition. The data showed a clear predominance of morphonotactic clusters in one out of three subcorpus. Developmentally, an increase in all three cluster types was observed and generally all clusters were predominantly less preferred. Early emerging phonotactic clusters appeared to be mostly preferred, while the first morphonotactic clusters comprised less preferred combinations, with a gradual increase in preferability. Individual differences in cluster distribution and preferability blurred the role of the importance of morphonotactic information in L1 acquisition, suggesting that children employ different paths to acquire a language, but also suggesting that further exploration of this hypothesis in different languages and using different approaches is needed.