Decomposition is a dominant design strategy because it enables complex problems to be broken up into more manageable modules. However, although it is well known that complex systems are rarely fully decomposable, much of the decomposition literature is framed around reordering or clustering processes that optimize an objective function to yield a module assignment. As illustrated in this study, these approaches overlook the fact that decoupling partially decomposeable modules can require significant additional design work, with associated consequences that introduce considerable information to the design space. This paper draws on detailed empirical evidence from a NASA space robotics field experiment to elaborate mechanisms through which the processes of decomposing can add information and associated descriptive complexity to the problem space. Contrary to widely held expectations, we show that complexity can increase substantially when natural system modules are fully decoupled from one another to support parallel design. We explain this phenomenon through two mechanisms: interface creation and functional allocation. These findings have implications for the ongoing discussion of optimal module identification as part of the decomposition process. We contend that the sometimes-significant costs of later stages of design decomposition are not adequately considered in existing methods. With this work we lay a foundation for valuing these performance, schedule and complexity costs earlier in the decomposition process.