Ontologies have been suggested as a means to ensure interoperability of Ambient Intelligence (AmI) applications as well as effectiveness within different sensory environments. We introduce a new logical language specifically tailored to the task of ontology design for AmI applications. This language has known advantages of description logics, namely a precisely defined semantics and a brief, intuitive syntax; and it facilitates formalising knowledge about contexts: a unified syntax for representing containment relations makes it easy to specify that a context is spatially or temporally contained in another, that two contexts share a certain group of agents or objects, or that an application is in a state that requires it to adapt its behaviour.