Abstract
Word-final positions are frequently described as optionally salient, depending on the presence or the absence of bound morphology. In fact, they often incur disruptive phonological processes (such as deletion or assimilation), which are partially blocked in the presence of bound morphology.
Previous studies have shown that these effects are also active in the sublexicon (Post et al., 2008; Cilibrasi, 2015). Investigations of this phenomenon so far focused on monolingual speakers, and little is known about second-language learners of English.
This study investigates the perceptual salience of bound morphemes in L2 learners of English with Czech as L1, and it also inevitably attempts to find evidence either for the rule-based or the whole-word processing of words as regards the perceptual decomposition of inflected verbs into stems and affixes. To investigate this phenomenon, sixty participants at different proficiency levels (B1, B2, and C1) with Czech as L1 were tested using a minimal-pairs discrimination task with three types of minimal pairs (MPs): MPs containing non-words with potential morphological information, MPs containing non-words with no potential morphological information, and a control condition (see Cilibrasi, 2015).
Upon hearing a pair of non-words, the participants were asked to decide whether the two non-words sound different or the same. Reaction times and accuracy were recorded.
The results showed that L2 learners performed similarly to the monolingual speakers tested in Cilibrasi (2015). The reaction times reflected the presence of bound morphology, with non-words containing bound morphology taking longer to be discriminated at all language levels.
Interestingly, English proficiency did not generate any interaction: Intermediate participants performed similarly to upper-intermediate and advanced students. The data also confirmed the influence of phonotactic probabilities on reaction times, implying that there might be some frequency effects running parallel to morpheme stripping.
These results suggest that second-language learners of English with Czech as L1 behave in the same way as monolingual speakers when processing inflectional bound morphemes in English and that the strategy used during perception is the same from a relatively early language level. It is used implicitly by all subjects, possibly as a consequence of unconscious automatic processing.