It is demonstrated that when concepts elicited from the subject are analyzed in a repertory grid together with those supplied by the experimenter, then the grid no longer preserves its advantages in reflecting individual psychological organization. However, the necessary experimental modifications and checks are themselves avenues of understanding, and in particular of the interplay between the conceptual worlds of subject and observer. The repertory grid’s basic asymmetry reflects psychological contrasts between elements and constructs; it follows that concepts supplied by the experimenter take their place amongst the elements.