Sound correspondences in the English-Spanish bilingual lexicon

While it has been recognized that L2 word learning is facilitated for cognates (De Groot & Keijzer 2000), approaches to cognate acquisition have focused on the similarity of L1- L2 forms, overlooking regular patterns in differences between items. For example, English phone [dʒ] regularly corresponds to the Spanish phone [x]: agent [eɪdʒɛnt]- agente [axente], voyage [vɔɪədʒ]- viaje [bjaxe]. The current studies test whether L1 English, L2 Spanish learners acquire and utilize regular cognate sound correspondences. Experiment 1 compared accuracy for cognate forms that include or do not include regular correspondences. Subjects learned the English names of 20 monsters. Afterward, they saw each monster's image and heard its name in English, then recalled and produced the monster's (cognate) name in Spanish. Results revealed higher accuracy for items containing regular cognate correspondences. Subjects with higher proficiency showed greater differences in accuracy between regular and irregular items. In Experiment 2, subjects heard a novel word in either English or Spanish and invented a plausible cognate in the other language. Their modifications to the word forms were analyzed. Analyses revealed that subjects’ modifications were not random, but rather demonstrated convergence on dominant modification strategies. Higher proficiency correlated with greater convergence on dominant strategies. Together, these results demonstrate that L1 English, L2 Spanish learners have knowledge of regular cognate correspondences and can utilize correspondences to learn or invent new cognate forms. Furthermore, because this knowledge is acquired gradually by the L2 learner, cognate processing is not consistent across proficiency levels or between individual learners.