The goal of this research is to test a model of speech production in which the speaker operates under two basis constraints: (1) a listener- oriented requirement to produce an intelligible signal with (2) a mechanism that has dynamical properties that limit kinematic performance. Certain strategies are used to produce speech under these constraints. In order to achieve an acceptable level of intelligibility, the speaker makes a global situation-dependent modification of clarity by using values of kinematic parameters that will produce sufficiently-distinctive acoustic cues while minimizing the expenditure of articulatory effort. This global modification of clarity influences the kinematic properties of the movements of individual articulators, and also the timing of the movements of articulators with respect to one another. Under some circumstances, a combination of phonetic context, kinematic performance limits and inherent variability can modify the displacement of an articulator so as t threaten the distinctiveness of certain acoustic cues. To maintain clarity in such cases, speech motor programming includes a strategy of """"""""motor equivalence"""""""" that makes subtle adjustments in articulatory displacements to help keep acoustic variation within perceptually-acceptable bounds. To test hypotheses based on this overview, we will make kinematic and acoustic measures in four inter-related studies, all performed on the same 10 subjects using a multiple single-subject design, and we will obtain intelligibility measures of the subjects speech from an independent group of listeners. The four studies are: (1) The clarity constraints vs. economy of effort: speaker strategies under varying clarity demands. (2) Clarity vs. economy of effort; the relative timing of articulatory movements. (3) Kinematic performance limits on speech articulations. (4) Interarticulator coordination in the production of acoustic/phonetic goals: A motor equivalence strategy. Data from all of these experiments will be derived from the acoustic signal and the movement of points on the lips tongue and mandible, as transduced by an ElectroMagnetic Midsagittal Articulometer (EMMA) system. Spectral and temporal measures will be extracted from the acoustic signal and articulatory kinematic measures (displacements, velocities, accelerations, durations) will be extracted from the movement signals. Measures of """"""""effort"""""""" will be calculated from the kinematic parameters. These measures will be compared with one another and with results of tests of intelligibility and prototypicality of phonetic tokens. The results from these highly-integrated experiments should help understand how the constraints of intelligibility and articulatory performance limits influence the strategies used in adjusting clarity vs. effort in normal speech production. The addition to the literature of these data, which include measures of their variability, should be of great help in designing and interpreting studies of disordered speech production.
Showing the most recent 10 out of 19 publications
