Substantial empirical evidence is available in support of both Herzberg's duality theory of job satisfaction and more conventional uniscalar explanations. This article breaks the deadlock by emphasizing the importance for the development of theory of the fact that different methods of measurement do consistently yield different results.