Special Section IN WHAT SENSE ARE POSITIVE AND NEGATIVE AFFECT INDEPENDENT? A Reply to Tellegen, Watson, and Clark

304 Copyright © 1999 American Psychological Society VOL. 10, NO. 4, JULY 1999 Abstract—We agree with most of the claims made by Tellegen, Watson, and Clark: Pleasant and unpleasant affect (i.e., happiness and unhappiness) are strongly negatively correlated; random and nonrandom error need to be taken into account before interpreting correlations among affect scales; scales of positive affect (PA) and negative affect (NA) embedded within the affective circumplex are orthogonal in theory, as are pleasure and arousal; and two-dimensional views of the affective space are more valid and useful than oneor threedimensional models. Finally, and most important, we agree that PA and NA, as typically operationalized, are not independent, despite what may be the case theoretically. In this brief rejoinder, we offer some reasons why measurement problems likely prevent PA and NA from being measured orthogonally, why the factor analytic work reported in the target article is unlikely to shed new light on the perplexing problem of how best to characterize and measure affect, and why Pleasantness and Arousal tend to be operationally superior to their PA and NA counterparts.