The questions I try to answer in the pages that follow were raised after talks I gave at the National Association for Research in Science Teaching (NARST) meeting in Atlanta, GA (April1990) and at an American Association for the Advancement of Science (AAAS) symposium in Washington, D.C. (February 1991). Given the time limitation at both occasions, I could address only a small selection. While reviewing the whole collection at a later date, I found that the material could be roughly divided into three subject areas . I begin with the specifically epistemological ones, then consider those that concern the problem of social interaction, and I end with some implications the constructivist orientation might have for teachers and the philosophy of instruction. Because the answers I give are not derived from an established dogma butspringfrom my subjective point of view, the reader will find a certain amount of overlapping between the three sections. I would like to claim that this is inevitable because, in my experience, once one shifts to the constructivist orientation, everything one thinks and does changes in a way that seems remarkably similar and coherent. Let me emphasize a point I have made in many of my papers: constructivism, as far as I am concerned, is one possible way of thinking. It is a model, and models, no matter how useful they may prove, must never be claimed to be "true."