Activity oriented approaches were once introduced and regarded as promising future alternatives to what was then the conventional culture of transport demand modelling methods. Over the years, the methods developed, and the research area does now constitute an active and independent tradition with an internationally widespread network of researchers involved. The interest for activity modelling has mostly been concentrated to a group of scientists, who share also an interest in (and conviction of) travel decision mechanisms that within classical economic theory would be regarded as non-rational. During the same time, the role of travel forecasting models generally has developed to be strongly linked to economic appraisal, cost-benefit techniques and welfare theory. In these contexts, rationality and utility maximisation are key concepts. Thus, the gap between "traditionalists" and "activity modellers" has widened. Both groups claim the need for "consistent" modelling systems, and the assumptions with which the models need to be consistent seem to drift apart. This is a discussion paper, based on analyses of available literature. Thus, it will not introduce any new research results, but rather open for a discussion of the latent and explicit conflicts between the two "approaches". It will be claimed that some differences between the approaches are in fact less significant than has been indicated in literature, while some other (common) problems may be larger than what has been admitted.