The French movie "Life Is a Long Quiet River" (« La vie est un long fleuve tranquille », Chatiliez, 1988) compares the fates of two children, born in contrasting social classes and exchanged at birth, and plays with the unpredictable curves and bends of their developmental histories, which eventually bring them to acquire the skills and knowledge that they might have acquired in their original milieu. The ironic title strongly suggests that life is not a long quite river, and illustrates two long-time known principles of development. It shows that the development of a child or a person is not linear, and cannot be predicted; it also shows that many different ways can lead a person to develop a given skill or understanding (Vygotsky, 1929(Vygotsky, /1993. Admitting these principles has consequences for the study of human development: life-courses appear characterized not only by the regular and progressive establishment of regularities and continuities, but also, and mainly, by the moments in which these continuities are interrupted, reoriented, or challenged.Such moments are interesting for several reasons. Firstly, at a theoretical level, it is at points of bifurcation that the person or the organism has to develop new conduct. Secondly, at an empirical level, lives in contemporary societies are exposing people to interruptions to what appears continuous to them-for example through job reassignments, demands to engage in continuous education, changes in family composition, the introduction of new technologies at home or in the workplace, and so on and so forth. Thirdly, at a methodological level, these points constitute "natural change laboratories" in people's lives. In effect, experimental settings usually create tasks or strange situations to which people have to answer; in these situations, researchers examine people's answers or the processes that lead them to these. Ruptures in life produce equivalent tasks, or strange situations, calling for an answer-and thus processes whereby these answers, or adjustment are produced, can be observed; I call them here transitions.