Formal framework for reasoning about programs are important not only for automated tools but also for programmers. Type systems have proven an enormously popular framework for validating static analyses of programs, as well as for documenting their interfaces for programmers. However, most type systems used in practice today fail to capture many essential aspect of p behavior dependencie of programs The considerabl c i pas twenty years int developing that captu e information I pape examin compa contrast connec mbe o highly influentia p ototy ica system capturing dependencies Specifically w classi e typ system a onceive fo l canonical example o pendenc system system information-flow di moda systems c (co)monadi e fec dependenc discipline linear system p e- easonin abou state esou Finally als esen calculus provide insight possibili fo a unified account for all of these systems.