Human language constitutes a set of patterns along with the rules for their construction, combinations and transfer, and in this sense it can be easily made an object of pattern and symmetry studies. It is possible to investigate the patterning both within one language and across languages. Similarly to nature, in language and across languages, perfect symmetry is much less common than symmetry-breaking. How symmetrical different languages could be in regard to each other is seen as the reflection of the universal patterning of human brain, yet the sources of symmetry distortions are traced in the language evolution (Aitchinson, 1996). Most studies of symmetry and patterning in language relate to syntax (e.g., Stonham & Yiu, 2003). In particular, much attention has been given to symmetry-breaking, asymmetries and antisymmetries in syntactical constructions (e.g., syntax antisymmetry theory by Kayne, 1994; and dynamic antisymmetry theory by Moro, 2000). Many studies of phonological systems and phonological laws also operate with the notion of symmetry (e.g., Johnson & Hume, 2003). However, although various systems of prosody and intonation description have been put forward (e.g., Ladd, 1996), they do not normally consider symmetrical/asymmetrical patterning of the units. Since prosody and intonation are based on alternating acoustic patterns, it is also possible to address them from the viewpoint of symmetry theory. This paper considers asymmetries in the perception of prosodic parameters by Japanese and Russian subjects. The parameters employed in this study are the height of the pitch peak, which was earlier shown to be relevant for the perception of sentence type in short one-word stimuli (Makarova, 2001a) as well as pitch peak alignment (i.e. the exact location of the F0 maximum regarding the onset/offset of the accented syllable). Some earlier studies allow to hypothesize that the pitch peak alignment may contribute to the perception of sentence type in identification of short utterances (Makarova, 2001b). It is not clear, however, whether this feature can be symmetrical or asymmetrical across different languages.