Early problems associated with the development of computer systems led to the much cited 'software' crisis, which in turn has engendered an understanding of development that has sought a solution in terms of control and standardisation. However, systems still continue to disappoint and this paper proposes that two challenges provide one research avenue where greater understanding is required. The research avenue is that of developing adaptiveinformation systems. The challenges are those of complexityand change. A growing realisation of the former is noted through a discussion of theemergence of interpretivism as a method of understanding failure. A discussion of the poor, or lack of, understanding of the latter challenge leads to a conclusion that a different philosophy is required for developing information systems in modern environments