Robinson Hall B, #434
November 16, 2018, 10:00 AM to 12:00 PM
This dissertation presents a series of production and perception studies arguing that dissimilation patterns found in native languages can also inform the acquisition of a second language. I argue that dissimilarity, defined in terms of the Obligatory Contour Principle, can have an effect on L2 regardless of language experience, supporting the notion that this principle is universally present in L2 interlanguage. Specifically, the constraint OCP-place has been extensively documented in native languages with some researchers claiming that the constraint is a statistical universal. However, there have only been a few studies that explored the effect of OCP-place on L2 production and perception with no consensus regarding the source of this constraint.
First, the dissertation presents a corpus analysis of the production of L2 from several L1 backgrounds. Speakers demonstrated a preference for OCP-place abiding forms as production errors increased with OCP violations. The study took measures to explore the source of the constraint. The target structure (CC onset clusters) were absent from the speakers’ L1 to help limit the influence of native phonotactics. The L2 experience of the speakers was also considered and found to have no effect on production errors. The findings support earlier claims that OCP is a universal constraint.
The dissertation further tests OCP-place by narrowing the scope of the language experience of the learners. In a perception study, the listeners were chosen from L1s that differ in how strongly the language abides by OCP-place. I show that listeners from three L1 backgrounds, English, Arabic and Mandarin, prefer nonwords with no OCP-place violation. The study also shows that listeners have a broad bias towards OCP, yet gradient OCP effects can be learned from L1 exposure.
Finally, I argue that adjacent environments (CC onset clusters) are influenced by OCP-place. Participants from L1 backgrounds with no onset clusters, Arabic and Mandarin, perceived an illusionary vowel between homorganic consonants. I show that perception of non-native clusters is hindered by OCP. I also argue that OCP-place effects are present regardless of proximity as listeners are influenced by OCP in both adjacent (CC onsets) as well as non-adjacent sCVC words.
Results from the three experiments presented in this dissertation support the following: 1) OCP-place is a constraint available to L2 learners. It informs their production as well as their judgment of novel words; 2) L1 and L2 listeners alike have a broad universal perceptual bias against similarity that is specialized to fine-grained constraints that are language-specific. The findings support earlier conclusions that OCP-place’s typological reoccurrence is a result of a universal preference for dissimilarity.