An on-going and major problem faced by information systems developers and business users alike is reaching a clear and consensual understanding of the system by both groups. This can be difficult because the businesses are (typically) process driven, while the systems are (increasingly) object-oriented. Enter modeling. Modeling is one way of presenting complex information in a way that enhances or eases understanding. But, even models can be extremely complex, and the underlying tools and modeling languages are not any less complex. This chapter investigates the possibility that modeling languages can be simplified by considering that not all of the “words” in the language are used all of the time. If theoretical (maximum) represents all the words in a modeling language, then this chapter suggests that there might exist a more use-based (we name it practical complexity) subset that represents an easier to learn and use subcomponent of the language.