Abstract
In the present study we assume –following Foucault− that science is a discourse. With this we mean that science is capable of normalizing practices. But, as Derrida has shown, every discourse can be deconstructed. In this paper, we show that it is possible to de‐normalize the scientific discourse using the same elements that form science. Science, of course, is not a monolithic discourse, but composed of multiple ways of observing –e.g. disciplines. In this article we present the case of a trend in psychology which studies love relationships and couple election. By analyzing the implicit assumptions of this trend we show what has been left out and it’s not visible in the scientific presentation of results. That is, we present what has been excluded in order to establish normality.