A set of re-synthesized stimuli with modified parameters of pitch peak height and alignment was prepared and employed in the perceptual experiment in which 40 Japanese and 40 Russian subjects (male and female, aged between 20 and 30) performed a sentence type identification task. Re-synthesis was performed on a one-word utterance ‘banany’ [bananas],  read as a declarative by one female speaker of standard (Moscow) Russian who was 29 at the time of the recording. The sound file was sampled at 16 KHz. Re-synthesis of the original stimuli and their subsequent manipulations were performed with Praat software programme ( PSOLA analysis and resynthesis mode). Two heights of the F0 (pitch) peak chosen for manipulations were 320 Hz (referred to as ‘high peak’) and 270 Hz (‘low peak’). These heights were determined by the threshold of ‘declarative’ vs ‘non-declarative’ sentence type judgments by Japanese and Russian listeners established in earlier experiments (Makarova 1999a, 2000a, c, d). Fourteen identical manipulations of the pitch peak alignment were subsequently performed at both pitch heights, which yielded the total of 28 experimental stimuli. The peak was shifted within the accented and postaccented syllables and in about 20ms intervals (2 F0 periods). All the resulting sound files were copied 5 times each with 2.5s silent intervals between them and randomized to obtain the sound record for the experimental session. See Fig. 1 for the examples of parameter manipulations.