Human futuristics, as a study of future cultural alternatives, limitations and choices, will not be another branch of "science" in the traditional sense. Its function will differ from that of science in many respects. First, future cultures are not existing "objects" to be observed, analysed and explained. Second, future cultures cannot be predicted by extrapolating the past pattern of change, the past rate of change or even past rate of acceleration of change. There are too many unprecedented innovations taking place which render extrapolations invalid. Third, culture changes do not just happen. People make them happen. Therefore culture changes are subjeot to people's goals, imagination, will and choice.In this third sense human futuristics may be considered more like engineering than science. But here again, the philosophy and the principle of human futuristics differ from those of engineering. First, engineering is mainly a matter of designing an object for a specifically given purpose, while in human futuristics the cultural goals are not apriori given. Goals are generated by people. Goals are empirically built inductively by the opinions of the people rather than apiori given. Engineers work from a given goal downward, while human futuristics needs to operate upward Srom grtm-roots @ople.Second, cultural goals are lielei ogeneous because people's needs vai y from individual to individual, and from group to group. Therefore human futuristics deal with diverse-goal system while an engineering