The role of morphology and training modality in the acquisition of complex novel words

<p>This thesis investigates the role of morphemic knowledge in learning novel words in L1 and L2 speakers of English. In the first two experimental chapters (Chapters 2 and 3), a written training paradigm was used to examine the generalisation of written trained morphemes to a new morphemic context through a post-training recognition task where the participants were presented with items consisting of trained and untrained morphemic units. Moreover, we investigated whether morphological family size (i.e., large vs. small) impacts the learning of written complex novel words (e.g., <em>torbiph</em>, <em>farsherp</em>). The results showed that after written training participants found it harder to reject items containing a trained morphemic constituent. Likewise, novel words containing a trained constituent with large than small family size were harder to reject as untrained novel words. These findings suggest that participants learned and parsed complex novel words through written exposure and were able to recognize the trained morphemic units in new morphemic contexts without previously encountering the morphemic units in isolation. In addition, morphological family size facilitated learning of complex novel words. In the third experimental chapter (Chapter 4), we used an oral training paradigm to examine whether participants generate orthographic expectancies of complex novel words (e.g., <em>vishes</em>, <em>vishing</em>, <em>vished</em>) without prior print exposure. Half of the novel words had predictable spelling from phonology (e.g., <em>vish</em>, <em>nesh</em>) and the other half had unpredictable spelling (e.g., koyb, vaype). Following training, participants’ eye fixations on novel word stems embedded in sentences were monitored. The results of the spoken training showed that participants were successful in learning complex novel words which facilitated reading times of the trained novel stems in the post-training phase. However, the training effects were not modulated by differences in spelling predictability in the L2 group, suggesting that L2 compared to L1 English speakers did not integrate orthographic expectancies in their reading. In sum, this thesis provides important insights into the mechanisms by which L1 and L2 speakers of English acquire morphologically complex novel words, showing that spoken and written vocabulary acquisition goes beyond the learning of exact orthographic forms and instead point to a more complex interplay between the members of the same morphological family, with morphological family size having a clearly facilitatory role in this process.</p>