[Automatic information processing, the frontal system and blunted affect. From clinical dimensions to cognitive processes toward a psychobiological explanation of temperament].

Several theorists have drawn a distinction between automatic and attentional or controlled processing. Hasher and Zacks (1979), were the very first to argue that effortful processes are reduced under conditions of stress including depression. They suggested that, in these conditions, no such deficit occurs in automatic processing. Then, Weingartner and co-workers provided some experiments which seemed to support such an interpretation of the cognitive dysfunction in depressed patients. However, some recent data do not fit with this well admitted theoretical framework. The purpose of our article is to try to clarify this issue both from a theoretical and from a methodological point of view. First, we make a critical review of the most recent results in three fields of experimentation related to the "automatic versus controlled" topic: 1) The classical neuropsychology of memory which manipulates the level of effort required to perform the tasks. Confusion arises when theories at the process level are tested with reference to data collected at the task level. The transparency assumption could be false: Impairment in an effort-demanding task could be due to a defect in automatic processes and it is possible to hypothesize that the more automatic processes are deficient, the more controlled processes are saturated and the effort demanding task impaired. The emergence of controlled processes could depend on the level of automaticity and the regulation of automatic processes could be determinant for the ability of the subject to make associations. 2) The recent studies on implicit memory in depression.(ABSTRACT TRUNCATED AT 250 WORDS)