Modeling Vowel Normalization and Sound Perception as Sequential Processes

This study constitutes the first attempt at combining vowel normalization procedures with the linguistic perception framework of Stochastic Optimality Theory [1] and the Gradual Learning Algorithm [2]. Virtual learners possessing different normalization procedures, and a control learner with no normalization, were trained to perceive Brazilian Portuguese and American English vowels. Our results show that learners equipped with normalization algorithms outperformed the control learners, obtaining accuracy scores up to 33% higher. Thus, this model in which normalization and sound perception are implemented as two sequential processes seems to be able to explain sound categorization adequately. That is, it improves the performance of a perception grammar when the training and testing sets have speakers with different ages and gender.