Is not the problem with reports of this sort the assumptions on which they are made—not least the assumption that though circumstances change dramatically, behaviour will not change at all?