La variabilité intra-individuelle par rapport au temps, souvent négligée dans les recherches contemporaines, a été centrale en psychologie dès les années 1900. Pierre Janet adopte souvent une perspective temporelle, pratiquant déjà l’étude longitudinale ; la biographie du sujet est au cœur de l’entretien clinique. Des recherches comme celles de Ninot et Costalat-Founeau (2011) s’inscrivent dans cette riche tradition. La confrontation épistémologique des travaux anciens et actuels permet d’éclairer des notions comme celles d’évolution temporelle ou « d’oscillation » qui jouent un rôle important en psychologie normale et pathologique. Un débat de près d’un siècle autour des troubles bipolaires témoigne de l’utilité de bien caractériser les périodicités de ces variations temporelles. Quelques imprécisions terminologiques compliquent parfois les discussions : les termes « temporel » et « dynamique » sont considérés synonymes, ce qui n’était pas le cas auparavant ; reconsidérer leur spécificité serait utile pour préciser certaines causalités. De ce point de vue, l’ancienne « psychologie dynamique expérimentale » du début du xx^e siècle présente encore un certain intérêt pour la recherche expérimentale actuelle.

Intra-individual variability across time, often neglected in current research, has been in focus of psychological research around 1900. Janet's “dynamical psychology” gave a great importance to biography and dealt with time evolution of psychological variables: he already used longitudinal studies. Researches like these of Ninot and Costalat-Founeau (2011) belong to this rich tradition. Some epistemological confrontations between ancient and current works may shed light on notions such as “time variation” or “oscillation”. Firstly, time variation was introduced in psychology by J.H. Jackson after Spencer's main ideas, under the notion of “evolutionism”. French psychology was immediately receptive to the idea. In this context, J.M. Baldwin started to investigate child psychology and coined the expression “genetic psychology” which would have a great diffusion. As soon as 1895, H. Beaunis wrote that genetic psychology had replaced all former approaches. Indeed, most of psychological phenomena would be studied as functions of time. Nevertheless, the study of time variation declined along with jacksonnism, but modern differential psychology has carried the tradition on: currently, time evolutions can be modeled by dynamic systems. As for the concept of “oscillation”, this is a core one in Janet's psychopathology. For Janet, oscillations of the mental level can be observed in all psychological troubles. That's why he criticized the “bipolar” diagnostic. Indeed, the bipolar debate has recently regained prominence in the research: Experience Sampling Methods (ESM) help to improve time variation collection done by bipolar subjects, while dynamic analysis can show fine structures in time patterns. Despite these advances the nosographical category of bipolar troubles reveals divergences in the clinicians points of view. Finally, the recent developments of time variation collection and analysis still bear some difficulties of its own. Currently it seems that “dynamic” has become a synonym of “temporal”. In the beginning of the 20th century, this was not the case. The two realms supported even different causalities. The distinction may still be conceptually relevant: for Janet, it is the grounding of psychotherapy. Indeed for him, psychotherapy targets time changes induced by dynamic interventions on the patient's forces. From these various points of view, the former “experimental dynamic psychology” advocated by Janet and his colleagues at the beginning of the 20th century still shows some relevance for current psychological research.