In this article, we discuss recognition of social dancing from a real image sequence and an acoustic signal of music. Since social dancing is somewhat complicated movements by multiple humans, it is difficult to recover detailed description of each dancer's posture by conventional approach based on matching of an articulated human body model to each frame of an image sequence. Assuming to use the result of recognition process for annotating video of social dancing, we focus on the bare minimum of motion description enough to recognize each motion element corresponding to a symbol that we employ to annotate the movement. We show that the motion description can be acquired from a real image sequence and an acoustic signal by simple familiar techniques of computer vision and music information processing.