Larger lexicons enable representation of fine-grained phonological similarity structure: Evidence from English L2 speakers' sound similarity judgments of word pairs

Native English speakers are sensitive to the small-world structure and community structure of the phonological similarity network of English words. In this study we investigated whether L2 speakers of English are sensitive to the overall similarity structure of the phonological lexicon, and whether this sensitivity is modulated by the size of their L2 vocabulary. Participants with diverse L1s (with English as their L2) completed a phonological similarity rating task where they listened to pairs of English words and provided similarity judgments for word pairs of varying path lengths and community membership. Path length in the phonological network represented the number of steps needed to traverse from one word to another word in the network. Word pairs were either from the same phonological community or different communities. English vocabulary knowledge was assessed using the LexTALE (Lemhöfer & Broersma, 2012). Results indicated that participants with higher LexTALE scores showed greater sensitivity to both community membership of word pairs as well as phonological distance between words at shorter path lengths. Computational simulations of the task with phonological networks depicting various levels of L2 proficiency qualitatively align with the observed behavioral results. The simulations suggest that larger network sizes provide more degrees of freedom for representing subtle patterns of similarity relations among word-forms. These findings have implications for understanding how expansion of the phonological mental lexicon enables learners to represent fine-grained, internal structure of phonological similarity relations among words.