The nature of the neural mechanisms involved in movement planning still remains widely unknown. We review in the present paper the state of our knowledge of the mechanisms whereby a visual input is transformed into a motor command. For the sake of generality, we consider the main problems that the nervous system has to solve to generate a movement, that is: target localization, definition of the initial state of the motor apparatus, and hand trajectory formation. For each of these problems three questions are addressed. First, what are the main results presented in the literature? Second, are these results compatible with each other? Third, which factors may account for the existence of incompatibilities between experimental observations or between theoritical models? This approach allows the explanation of some of the contradictions existing within the movement-generation literature. It also suggests that the search for general theories may be in vain, the central nervous system being able to use different strategies both in encoding the target location with respect to the body and in planning hand displacement. In our view, this conclusion may advance the field by both opening new lines of research and bringing some sterile controversies to an end.