Software architecture description languages (ADLs) were a particularly active research area in the 1990s. In 2000, we published an extensive study of existing ADLs, which has served as a useful reference to software architecture researchers and practitioners. However, the field of software architecture and our understanding of it have undergone numerous changes in the past several years. In particular, the Unified Modeling Language (UML) has gained popularity and wide adoption, and as a result many of the ADLs we had studied have been pushed into obscurity. In this paper, we argue that the main reason behind this is that the early ADLs focused almost exclusively on the technological aspects of architecture, and mostly ignored the application domain and business contexts within which software systems, and development organizations, exist. Together, these three concerns—technology, domain, and business—constitute the three “lampposts” needed to appropriately “illuminate” sof...