The impact of segmental familiarity on incidental suprasegmental category learning

Previous research on tone category learning has suggested that the learners' familiarity with the segmental composition of tokens can impact learning. However, the impact of segmental familiarity on tone category learning has seldom been tested. Also, the learning of novel tone categories in natural speech has been studied primarily under explicit learning paradigms. In the current study, we expand the use of incidental learning paradigms from synthesized sound category learning to natural speech to examine the impact of segmental familiarity on tone category learning. In one online session, three groups of native English speakers from multiple countries learned four Thai tone categories presented in tokens with familiar or unfamiliar vowel segments. Results demonstrated that incidental learning is a quick and effective method for studying novel sound category acquisition, with learners in each group achieving up to 100% identification accuracy at test. Using this paradigm, we observed a small impact of segmental familiarity on learning, suggesting that its influence may be highly granular, depending on the type and complexity of segmental features in the stimuli.