Stress distinction in German: simulating kinematic parameters of tongue-tip gestures.
Citation
Mooshammer, C. & Fuchs, S. (2002) Stress distinction in German: simulating kinematic parameters of tongue-tip gestures., Journal of Phonetics, vol. 30, , pp. 337-355,
Abstract
Levels of stress are not only distinguished by varying fundamental
frequency contours but also by changes of supralaryngeal parameters,
e.g., unstressed syllables exhibit reduced movement amplitudes and
durations compared with stressed syllables. To investigate the effect
of deaccentuation on apical gestures in /tVt/ sequences with all
vowels of German, we recorded lingual movements of five speakers
by means of EMMA. Movement paths of recorded stressed items
were manipulated to simulate kinematic parameters of recorded
unstressed items in three different ways: truncation, rescaling and
combined truncation and rescaling. We assumed that the simulation
type that generated parameters most similar to recorded unstressed
items can be interpreted in terms of a generalized motor program for
deaccentuation. The following parameters of simulated movements
were compared with measured unstressed items: movement durations,
peak velocities, distances, interval between velocity peaks in percent
of syllable duration, symmetry of velocity profiles and number of
acceleration peaks between velocity peaks. Combined simulations
resembled most closely the kinematic parameters of unstressed items
but could not generate the smaller amplitudes of unstressed syllables
with lax vowels, since durational reduction of lax vowels due to
deaccentuation was very small, i.e., the spatial reduction was not
proportional to the temporal reduction for lax items. Therefore, it
can be concluded that with the method used here no single
parameter or pattern could be found whose manipulation results in
the kinematic characteristics of unstressed syllables, which speaks
against the concept of a generalized motor program for
deaccentuation.