Quantification techniques for lateral consonants using EMA
Date:
Abstract: Quantification of the degree of lateralization is crucial in the study of lateral consonant articulation. Ying et al. (2021) provides a quantification method for the lateralization of /l/ in Australian English using EMA data, relying on the difference in the z-dimensional position of a parasagittal sensor and a virtual point along the surface of the tongue based on the positions of three midsagittal sensors. This method estimates the midsagittal point for comparison based solely on the x-dimensional position of the parasagittal sensor, which may obscure the true degree of lateralization in cases where the tongue has significant upward or downward curvature. In this study we employ the “lateralization angle” estimation method, which takes the angle between the parasagittal sensor and the surface of the tongue at a certain point along the midsagittal line, where Euclidean distance in the x- and z-dimensions to the parasagittal sensor is the shortest. The two methods are compared by analyzing the same EMA data taken from Korean and Mandarin speakers, and it is found that the proposed method more effectively utilizes the 3D positional information provided by EMA, and is more robust in situations where significant tongue curvature is observed.