Can simultaneous speech stimuli be classified in parallel?