This article deals with the empirical implications of the choice of the “actionnist” model for the explanation of common feelings of justice. If the actionnist model implies that we should consider the reasons people could set out to justify their moral standpoints as part of the explanation that the sociologist will himself be able to give of them, it raises the following issue: are these reasons to be reconstructed or, on the contrary, to be revealed by people themselves? Two arguments are discussed. One deals with the interpretation of the results of experiments whose procedure aims to encourage people to be impartial. The second one deals with the sense of the reconstruction of people’s reasons in the actionnist model and its compatibility with an analysis of actual reasons.