Evaluation and organization, it turns out, are somewhat contradictory. The ideal member of the self-evaluating organization is best conceived of as a person committed to specific modes of problem-solving. Evaluation should lead not only to finding better policy programs to accomplish objectives but also to altering objectives themselves. Evaluative man must learn to live with contradictions. The internal conflicts generated by these earlier modes of evaluation have received abundant documentation in the administrative literature. Evaluation cannot ordinarily proceed, then, by determining how well the unknown objectives of a program are being achieved at whatever cost. A critical aspect of evaluation is which level within the organization might conceivably use it. The popularity of decentralization further deepens the potential conflict between evaluators and program managers. Impatience grows at the operating level because evaluators have difficulty in meeting programmatic needs at the appropriate time.