The current terminology for animal groups mainly stands on affinitive interactions among organisms. This method is not appropriate to analyze evolutionary process, because firstly it treats the consequences, not the process, of group formation; secondly, the criterion may heavily affected by the peculiarity of human perception. The differences among consequence patterns does not always reflect the differences among process patterns and the important factors during the evolutionary process is not always clear for human perception. In order to eliminate such “bias, ” it seems to be necessary to focus on the ƒactors affecting the group formations, i.e., predation pressure, contraction of disease, foraging efficiency, resource defense. Different factors would produce differently structured groups, so that animal groups can be recognized in pluralistic ways. Groups should not be limited just to a “social” one: anti-predation group, foraging group and so on, exist with similar reality. The merits and demerits of such new terminology is discussed, focusing on complicated grouping patterns in non-human primates. Some perspectives for future studies are also examined.