Keep the lips to free the larynx: Comments on de Boer's articulatory model (2010)
Résumé
De Boer (2010) used a simplified adaptation of Mermelstein's articulatory model - without lips - to investigate "the effect of larynx position on the articulatory abilities of a human-like vocal tract." He found "an optimal larynx height at which the largest range of signals [are] produced". Our replication of de Boer's simulations confirmed the peak in the F2 range function when "the vertical and horizontal parts of the tract are approximately equally long". However, simulations show that the diminished F2 range at high larynx positions stems from an insufficiently low F2 in the /u/ region. Without an independent lip closure paired with the velo-dorsal constriction, the model cannot create the two cavities necessary to produce simultaneously low F1 and F2. A second series of simulations with a proper lip section essentially restored the F2 space across a large range of larynx heights. These experiments confirm that larynx height does not inordinately restrict the acoustic F1-F2 capacities of a mammalian vocal tract, provided that lips are appropriately included in the simulations. De Boer, B. (2010). Investigating the acoustic effect of the descended larynx with articulatory models. Journal of Phonetics, 38(4), 679-686.