[The measurement of electrodermal activity]

Electrodermal activity (EDA) is an early physiological index and the subject of constant interest, in spite of the bad reputation attached to "lie detectors". This interest is expected to increase in the future, following the development of research related to the neurobiological aspect of emotions of which it is an index. Recent data provided by functional cerebral imaging has added to the significance of this index and should result in further interest.

The authors thus re-examined the various notions related to measuring EDA, and its practical aspect as well as its mechanisms. EDA should be useful both for authors wishing to use this variable and for readers wishing to form their own critical point of view.

The article first defines the various terms used to qualify EDA. Then, it analyses the mechanisms occurring at the sweat glands' level, showing that a distinct innervation of the sweat glands causes sweat to be released in the excretory channels, thereby allowing the recording of a negative surface potential in parallel to the lowering of skin conductance. Arguments are then pointed out to illustrate that the potential's positive phase following this first answer occurs in the case of high intensity stimulations. The study of the central command of sudation demonstrates that, several areas are involved and that different functions such as thermal regulation and motricity may interfere with emotive reactions. Difficulties regarding the mode of measurement of these answers as to their number and amplitude are also brought to light.

A particular interest of measuring EDA is its ability to highlight individual characteristic and unconscious emotional reactivity. Subjects who constitutionally present many spontaneous and therefore habitual EDA can indeed be opposed to subjects whose EDA reflexes are very few and hardly habitual. A theory suggests that for the first category, whose subjects are named labiles, emotional control may be at the origin of EDA. This characteristic brings to mind the case of antisocial subjects whose rate of EDA is also reduced, although for the latter a primitive drop in behavioral inhibition is involved. The production of EDA in response to non-conscious emotive stimulations can be objectified in the rare cases of prosopagnosia. These subjects who are unable to recognize familiar faces can produce EDA when presented faces with an emotional load. These cases contrast with the delusional denial of the Capgras syndrome where subjects do not present EDA, suggesting that the dysfunction of visual analysis occurs at a different level. There are other rare cases represented by cortical blindness where EDA shows that an unconscious emotional analysis is preserved. These subjects are known however to be capable of unconscious visual discriminations, which are possibly accompanied by EDA. This possibility of a "blind vision" is experimentally studied via subliminal vision testing (backward masking tests). These demonstrate that a rudimentary visual analysis is carried out in the subcortical circuits while taking into account the affective aspect of stimulations.

Present or future data should allow a greater comprehension of electrodermal signals, making it possible to overcome the difficulties related to their interpretation and facilitate their applications.