I showed that the simulation's error rate depends on the speaking rate (number of time steps per syllable).- Dell (2004), a pag.162