In their comment on K. Rothermund and D. Wentura, A. G. Greenwald, B. A. Nosek, M. R. Banaji, and K. C. Klauer agreed that salience asymmetries can be a source of Implicit Association Test (IAT) effects. The authors applaud this conclusion but point to problems with the other points that Greenwald et al. made. The authors have difficulties understanding the nominal feature account that Greenwald et al. put forward and have doubts about the usefulness of their broad conception of the concept association. The authors also argue that existing evidence concerning the construct validity of the IAT does not allow one to discriminate between the association and the salience accounts. In addition, the new studies that were presented by Greenwald et al. do not provide insights into what the IAT measures because they are either irrelevant for a decision between the different accounts or contain methodological problems that prevent a meaningful interpretation in terms of the models.