Phonological transfer of voicing and devoicing rules: evidence from L1 Dutch and L2 English conversational speech

This paper examines the productivity of voicing and devoicing rules in Dutch–English interlanguage. In Dutch, but not in English, coda obstruents can be subject to final devoicing and various voice assimilation processes, depending on the context. The analysis of a corpus of casual Dutch and English conversational speech of native Dutch speakers revealed that there are significant differences in the extent to which (de)voicing processes which were produced with a very high frequency in the native language (L1) are transferred into the second language (L2). These differences are explained in light of universal principles of markedness. The degree of transfer of intra-word processes is compared to that of cross-word assimilation processes. Zsiga [Zsiga, E.C., 2003. Articulatory timing in a second language, Studies in Second Language Acquisition 25, 399–432] predicts that no language learner should show a preference for transfer of cross-word timing patterns over intra-word patterns. It is argued that this hypothesis is not confirmed when learners have received explicit instruction on the absence of intra-word processes in the L2.