English has become the dominant donor language for many languages, including Croatian. Its prestigious status reduces the likelihood of borrowed words to adapt to a recipient language. As a result, some English loanwords occur in an unadapted form. Recent computational linguistic resources have given the necessary corpus-based data on the frequency and use of English loanwords in Croatian. This paper investigates the strategies employed by 116 students of the Faculty of Maritime Studies, University of Rijeka when asked to translate 392 most frequent, corpus-derived English loanwords into Croatian. The results were then compared with the available corpus-based data. The results show that single-word Croatian equivalents were preferred over adapted forms of English loanwords and multi-word expressions. When no such equivalent was available, unadapted English forms were used more frequently compared to adapted forms and multi-word expressions. The co-existence of loanwords and their native equivalents is reflected in responses to loanwords that have and those that do not have single-word equivalents. The results highlight the need for creating semantically precise single-word native equivalents, at the same time illustrating the resistance to accept novel native words.

No abstract

Verb constructions with a reflexive marker are termed “reflexive constructions”. Reflexive constructions in language acquisition research have been studied mainly within a formal theoretical framework while focusing on developmental differences in comprehension of syntactically bound or free pronominal elements. The present study aimed to determine the acquisition pathway of reflexive constructions in Croatian by examining the errors that children produce in early stages of acquisition. Correct and erroneous reflexive verb constructions were extracted from the spontaneous language production of three children previously recorded and transcribed for the Croatian Corpus of Child Language (Kovačević 2003). Errors were classified as omissions or overgeneralizations and further analyzed with respect to the type of reflexive construction, the complexity of the verb’s argument structure and consistency between the reflexive marker and verb. The results showed that children initially omitted the reflexive marker, then gradually introduced it into their production, occasionally overextending its use and thus producing overgeneralization errors. With age they became more successful in producing reflexive constructions. Consistent use of a reflexive marker alongside the verb in child-directed speech influenced the error rates in different types of reflexive constructions, while potential influence of the complexity of argument structure needs to be studied further.